mark arguments of case-lambda as variable, mark case-lambda as keyword (#14386)

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

.envrc

@@ -1,7 +1,8 @@
 watch_file shell.nix
+watch_file default.nix
 watch_file flake.lock
 watch_file rust-toolchain.toml
 
 # try to use flakes, if it fails use normal nix (ie. shell.nix)
 use flake || use nix
-eval "$shellHook"
+eval "$shellHook"

.gitattributes

@@ -8,4 +8,5 @@
 *.md       text diff=markdown
 
 book/theme/highlight.js linguist-vendored
+runtime/queries/**/*.scm linguist-language=Tree-sitter-Query
 Cargo.lock text

.github/ISSUE_TEMPLATE/blank_issue.md

@@ -1,4 +0,0 @@
----
-name: Blank Issue
-about: Create a blank issue.
----

.github/dependabot.yml

@@ -8,9 +8,6 @@ updates:
     schedule:
       interval: "weekly"
     groups:
-      tree-sitter:
-        patterns:
-          - "tree-sitter*"
       rust-dependencies:
         update-types:
           - "minor"

.github/workflows/build.yml

@@ -6,7 +6,12 @@ on:
       - master
   merge_group:
   schedule:
-    - cron: '00 01 * * *'
+    - cron: "00 01 * * *"
+
+env:
+  MSRV: "1.82"
+  # This key can be changed to bust the cache of tree-sitter grammars.
+  GRAMMAR_CACHE_VERSION: ""
 
 jobs:
   check:
@@ -15,17 +20,24 @@ jobs:
     if: github.repository == 'helix-editor/helix' || github.event_name != 'schedule'
     steps:
       - name: Checkout sources
-        uses: actions/checkout@v4
-      - name: Install stable toolchain
-        uses: helix-editor/rust-toolchain@v1
+        uses: actions/checkout@v5
+
+      - name: Install MSRV toolchain
+        uses: dtolnay/rust-toolchain@master
         with:
-          profile: minimal
-          override: true
+          toolchain: ${{ env.MSRV }}
 
       - uses: Swatinem/rust-cache@v2
         with:
           shared-key: "build"
 
+      - name: Cache tree-sitter grammars
+        uses: actions/cache@v4
+        with:
+          path: runtime/grammars
+          key: ${{ runner.os }}-${{ runner.arch }}-stable-v${{ env.GRAMMAR_CACHE_VERSION }}-tree-sitter-grammars-${{ hashFiles('languages.toml') }}
+          restore-keys: ${{ runner.os }}-${{ runner.arch }}-stable-v${{ env.GRAMMAR_CACHE_VERSION }}-tree-sitter-grammars-
+
       - name: Run cargo check
         run: cargo check
 
@@ -33,26 +45,29 @@ jobs:
     name: Test Suite
     runs-on: ${{ matrix.os }}
     if: github.repository == 'helix-editor/helix' || github.event_name != 'schedule'
+    timeout-minutes: 30
     env:
       RUST_BACKTRACE: 1
       HELIX_LOG_LEVEL: info
     steps:
       - name: Checkout sources
-        uses: actions/checkout@v4
+        uses: actions/checkout@v5
 
-      - name: Install stable toolchain
-        uses: dtolnay/rust-toolchain@1.70
+      - name: Install MSRV toolchain
+        uses: dtolnay/rust-toolchain@master
+        with:
+          toolchain: ${{ env.MSRV }}
 
       - uses: Swatinem/rust-cache@v2
         with:
           shared-key: "build"
 
-      - name: Cache test tree-sitter grammar
+      - name: Cache tree-sitter grammars
         uses: actions/cache@v4
         with:
           path: runtime/grammars
-          key: ${{ runner.os }}-stable-v${{ env.CACHE_VERSION }}-tree-sitter-grammars-${{ hashFiles('languages.toml') }}
-          restore-keys: ${{ runner.os }}-stable-v${{ env.CACHE_VERSION }}-tree-sitter-grammars-
+          key: ${{ runner.os }}-${{ runner.arch }}-stable-v${{ env.GRAMMAR_CACHE_VERSION }}-tree-sitter-grammars-${{ hashFiles('languages.toml') }}
+          restore-keys: ${{ runner.os }}-${{ runner.arch }}-stable-v${{ env.GRAMMAR_CACHE_VERSION }}-tree-sitter-grammars-
 
       - name: Run cargo test
         run: cargo test --workspace
@@ -62,7 +77,7 @@ jobs:
 
     strategy:
       matrix:
-        os: [ubuntu-latest, macos-latest, windows-latest]
+        os: [ubuntu-latest, macos-latest, windows-latest, ubuntu-24.04-arm]
 
   lints:
     name: Lints
@@ -70,17 +85,25 @@ jobs:
     if: github.repository == 'helix-editor/helix' || github.event_name != 'schedule'
     steps:
       - name: Checkout sources
-        uses: actions/checkout@v4
+        uses: actions/checkout@v5
 
-      - name: Install stable toolchain
-        uses: dtolnay/rust-toolchain@1.70
+      - name: Install MSRV toolchain
+        uses: dtolnay/rust-toolchain@master
         with:
+          toolchain: ${{ env.MSRV }}
           components: rustfmt, clippy
 
       - uses: Swatinem/rust-cache@v2
         with:
           shared-key: "build"
 
+      - name: Cache tree-sitter grammars
+        uses: actions/cache@v4
+        with:
+          path: runtime/grammars
+          key: ${{ runner.os }}-${{ runner.arch }}-stable-v${{ env.GRAMMAR_CACHE_VERSION }}-tree-sitter-grammars-${{ hashFiles('languages.toml') }}
+          restore-keys: ${{ runner.os }}-${{ runner.arch }}-stable-v${{ env.GRAMMAR_CACHE_VERSION }}-tree-sitter-grammars-
+
       - name: Run cargo fmt
         run: cargo fmt --all --check
 
@@ -98,18 +121,30 @@ jobs:
     if: github.repository == 'helix-editor/helix' || github.event_name != 'schedule'
     steps:
       - name: Checkout sources
-        uses: actions/checkout@v4
+        uses: actions/checkout@v5
 
-      - name: Install stable toolchain
-        uses: dtolnay/rust-toolchain@1.70
+      - name: Install MSRV toolchain
+        uses: dtolnay/rust-toolchain@master
+        with:
+          toolchain: ${{ env.MSRV }}
 
       - uses: Swatinem/rust-cache@v2
         with:
           shared-key: "build"
 
+      - name: Cache tree-sitter grammars
+        uses: actions/cache@v4
+        with:
+          path: runtime/grammars
+          key: ${{ runner.os }}-${{ runner.arch }}-stable-v${{ env.GRAMMAR_CACHE_VERSION }}-tree-sitter-grammars-${{ hashFiles('languages.toml') }}
+          restore-keys: ${{ runner.os }}-${{ runner.arch }}-stable-v${{ env.GRAMMAR_CACHE_VERSION }}-tree-sitter-grammars-
+
       - name: Validate queries
         run: cargo xtask query-check
 
+      - name: Validate themes
+        run: cargo xtask theme-check
+
       - name: Generate docs
         run: cargo xtask docgen
 
@@ -119,4 +154,3 @@ jobs:
           git diff-files --quiet \
             || (echo "Run 'cargo xtask docgen', commit the changes and push again" \
             && exit 1)
-

.github/workflows/cachix.yml

@@ -11,13 +11,13 @@ jobs:
     runs-on: ubuntu-latest
     steps:
     - name: Checkout sources
-      uses: actions/checkout@v4
+      uses: actions/checkout@v5
 
     - name: Install nix
-      uses: cachix/install-nix-action@V27
+      uses: cachix/install-nix-action@v31
 
     - name: Authenticate with Cachix
-      uses: cachix/cachix-action@v15
+      uses: cachix/cachix-action@v16
       with:
         name: helix
         authToken: ${{ secrets.CACHIX_AUTH_TOKEN }}

.github/workflows/gh-pages.yml

@@ -1,4 +1,4 @@
-name: Github Pages
+name: GitHub Pages
 
 on:
   push:
@@ -11,13 +11,13 @@ jobs:
   deploy:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v4
+      - uses: actions/checkout@v5
 
       - name: Setup mdBook
         uses: peaceiris/actions-mdbook@v2
         with:
-          mdbook-version: 'latest'
-          # mdbook-version: '0.4.8'
+          # mdbook-version: 'latest'
+          mdbook-version: '0.4.43'
 
       - run: mdbook build book
       

.github/workflows/release.yml

@@ -23,7 +23,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout sources
-        uses: actions/checkout@v4
+        uses: actions/checkout@v5
 
       - name: Install stable toolchain
         uses: dtolnay/rust-toolchain@stable
@@ -58,23 +58,21 @@ jobs:
     strategy:
       fail-fast: false # don't fail other jobs if one fails
       matrix:
-        build: [x86_64-linux, x86_64-macos, x86_64-windows] #, x86_64-win-gnu, win32-msvc
+        build: [x86_64-linux, aarch64-linux, x86_64-macos, x86_64-windows, aarch64-macos] #, x86_64-win-gnu, win32-msvc
         include:
         - build: x86_64-linux
-          os: ubuntu-latest
+          # WARN: When changing this to a newer version, make sure that the GLIBC isnt too new, as this can cause issues
+          # with portablity on older systems that dont follow ubuntus more rapid release cadence.
+          os: ubuntu-22.04
           rust: stable
           target: x86_64-unknown-linux-gnu
           cross: false
         - build: aarch64-linux
-          os: ubuntu-latest
+          # Version should be kept in lockstep with the x86_64 version
+          os: ubuntu-22.04-arm
           rust: stable
           target: aarch64-unknown-linux-gnu
-          cross: true
-        # - build: riscv64-linux
-        #   os: ubuntu-latest
-        #   rust: stable
-        #   target: riscv64gc-unknown-linux-gnu
-        #   cross: true
+          cross: false
         - build: x86_64-macos
           os: macos-latest
           rust: stable
@@ -85,13 +83,16 @@ jobs:
           rust: stable
           target: x86_64-pc-windows-msvc
           cross: false
-        # 23.03: build issues
         - build: aarch64-macos
           os: macos-latest
           rust: stable
           target: aarch64-apple-darwin
           cross: false
-          skip_tests: true  # x86_64 host can't run aarch64 code
+        # - build: riscv64-linux
+        #   os: ubuntu-22.04
+        #   rust: stable
+        #   target: riscv64gc-unknown-linux-gnu
+        #   cross: true
         # - build: x86_64-win-gnu
         #   os: windows-2019
         #   rust: stable-x86_64-gnu
@@ -103,16 +104,16 @@ jobs:
 
     steps:
       - name: Checkout sources
-        uses: actions/checkout@v4
+        uses: actions/checkout@v5
 
       - name: Download grammars
-        uses: actions/download-artifact@v4
+        uses: actions/download-artifact@v5
 
       - name: Move grammars under runtime
         if: "!startsWith(matrix.os, 'windows')"
         run: |
           mkdir -p runtime/grammars/sources
-          tar xJf grammars/grammars.tar.xz -C runtime/grammars/sources
+          tar xJf grammars.tar.xz -C runtime/grammars/sources
 
       # The rust-toolchain action ignores rust-toolchain.toml files.
       # Removing this before building with cargo ensures that the rust-toolchain
@@ -147,16 +148,8 @@ jobs:
         if: "!matrix.skip_tests"
         run: ${{ env.CARGO }} test --release --locked --target ${{ matrix.target }} --workspace
 
-      - name: Set profile.release.strip = true
-        shell: bash
-        run: |
-          cat >> .cargo/config.toml <<EOF
-          [profile.release]
-          strip = true
-          EOF
-
       - name: Build release binary
-        run: ${{ env.CARGO }} build --release --locked --target ${{ matrix.target }}
+        run: ${{ env.CARGO }} build --profile opt --locked --target ${{ matrix.target }}
 
       - name: Build AppImage
         shell: bash
@@ -183,7 +176,7 @@ jobs:
 
           mkdir -p "$APP.AppDir"/usr/{bin,lib/helix}
 
-          cp "target/${{ matrix.target }}/release/hx" "$APP.AppDir/usr/bin/hx"
+          cp "target/${{ matrix.target }}/opt/hx" "$APP.AppDir/usr/bin/hx"
           rm -rf runtime/grammars/sources
           cp -r runtime "$APP.AppDir/usr/lib/helix/runtime"
 
@@ -206,14 +199,25 @@ jobs:
           mv "$APP-$VERSION-$ARCH.AppImage" \
               "$APP-$VERSION-$ARCH.AppImage.zsync" dist
 
+      - name: Build Deb
+        shell: bash
+        if: matrix.build == 'x86_64-linux'
+        run: |
+          cargo install cargo-deb
+          mkdir -p target/release
+          cp target/${{ matrix.target }}/opt/hx target/release/
+          cargo deb --no-build
+          mkdir -p dist
+          mv target/debian/*.deb dist/
+
       - name: Build archive
         shell: bash
         run: |
           mkdir -p dist
-          if [ "${{ matrix.os }}" = "windows-2019" ]; then
-            cp "target/${{ matrix.target }}/release/hx.exe" "dist/"
+          if [ "${{ matrix.os }}" = "windows-latest" ]; then
+            cp "target/${{ matrix.target }}/opt/hx.exe" "dist/"
           else
-            cp "target/${{ matrix.target }}/release/hx" "dist/"
+            cp "target/${{ matrix.target }}/opt/hx" "dist/"
           fi
           if [ -d runtime/grammars/sources ]; then
             rm -rf runtime/grammars/sources
@@ -231,9 +235,9 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout sources
-        uses: actions/checkout@v4
+        uses: actions/checkout@v5
 
-      - uses: actions/download-artifact@v4
+      - uses: actions/download-artifact@v5
 
       - name: Build archive
         shell: bash
@@ -241,6 +245,7 @@ jobs:
           set -ex
 
           source="$(pwd)"
+          tag=${GITHUB_REF_NAME//\//}
           mkdir -p runtime/grammars/sources
           tar xJf grammars/grammars.tar.xz -C runtime/grammars/sources
           rm -rf grammars
@@ -254,7 +259,7 @@ jobs:
               if [[ $platform =~ "windows" ]]; then
                   exe=".exe"
               fi
-              pkgname=helix-$GITHUB_REF_NAME-$platform
+              pkgname=helix-$tag-$platform
               mkdir -p $pkgname
               cp $source/LICENSE $source/README.md $pkgname
               mkdir $pkgname/contrib
@@ -265,6 +270,7 @@ jobs:
 
               if [[ "$platform" = "x86_64-linux" ]]; then
                   mv bins-$platform/helix-*.AppImage* dist/
+                  mv bins-$platform/*.deb dist/
               fi
 
               if [ "$exe" = "" ]; then
@@ -274,7 +280,7 @@ jobs:
               fi
           done
 
-          tar cJf dist/helix-$GITHUB_REF_NAME-source.tar.xz -C $source .
+          tar cJf dist/helix-$tag-source.tar.xz -C $source .
           mv dist $source/
 
       - name: Upload binaries to release
@@ -286,7 +292,7 @@ jobs:
           file_glob: true
           tag: ${{ github.ref_name }}
           overwrite: true
-      
+
       - name: Upload binaries as artifact
         uses: actions/upload-artifact@v4
         if: env.preview == 'true'

.gitignore

@@ -3,3 +3,4 @@ target
 helix-term/rustfmt.toml
 result
 runtime/grammars
+.DS_Store

CHANGELOG.md

@@ -1,3 +1,602 @@
+<!--
+# YY.0M (YYYY-0M-0D)
+
+Breaking changes:
+
+Features:
+
+Commands:
+
+Usability improvements:
+
+Fixes:
+
+Themes:
+
+New languages:
+
+Updated languages and queries:
+
+Packaging:
+-->
+
+# 25.07.1 (2025-07-18)
+
+This is a patch release which lowers the GLIBC requirements of the release artifacts published to GitHub ([#13983](https://github.com/helix-editor/helix/pull/13983))
+
+# 25.07 (2025-07-15)
+
+As always, a big thank you to all of the contributors! This release saw changes from 195 contributors.
+
+Breaking changes:
+
+* The parsing of the command line has been rewritten and now supports flags and expansions ([#12527](https://github.com/helix-editor/helix/pull/12527), [#13018](https://github.com/helix-editor/helix/pull/13018), [9574e55](https://github.com/helix-editor/helix/commit/9574e55), [2d4c2a1](https://github.com/helix-editor/helix/commit/2d4c2a1), [#13192](https://github.com/helix-editor/helix/pull/13192), [67f1fe2](https://github.com/helix-editor/helix/commit/67f1fe2), [#13466](https://github.com/helix-editor/helix/pull/13466), [#13467](https://github.com/helix-editor/helix/pull/13467), [#13840](https://github.com/helix-editor/helix/pull/13840))
+    * Quoting and spaces are now handled differently. This can break existing keymaps which use typable commands, in particular `:sh`, `:set-option` or `:toggle-option`.
+    * The `:rsort` command has been removed. Use the reverse flag instead: `:sort --reverse`
+
+Features:
+
+* Add a picker which explores directories ([#11285](https://github.com/helix-editor/helix/pull/11285), [d4aed40](https://github.com/helix-editor/helix/commit/d4aed40))
+* Allow cycling through multiple LSP Hover responses with `A-n`/`A-p` ([#10122](https://github.com/helix-editor/helix/pull/10122), [2367b20](https://github.com/helix-editor/helix/commit/2367b20))
+* Add support for incomplete LSP completions ([5c1f3f8](https://github.com/helix-editor/helix/commit/5c1f3f8))
+* Add support for EditorConfig ([#13056](https://github.com/helix-editor/helix/pull/13056), [#13443](https://github.com/helix-editor/helix/pull/13443))
+* Add support for LSP document colors ([#12308](https://github.com/helix-editor/helix/pull/12308), [d43de14](https://github.com/helix-editor/helix/commit/d43de14), [47cdd23](https://github.com/helix-editor/helix/commit/47cdd23), [ba54b6a](https://github.com/helix-editor/helix/commit/ba54b6a), [#13188](https://github.com/helix-editor/helix/pull/13188))
+* Support expansions in external formatter arguments ([#13429](https://github.com/helix-editor/helix/pull/13429))
+* Switch out the highlighter for the `tree-house` crate ([#12972](https://github.com/helix-editor/helix/pull/12972), [09bc67a](https://github.com/helix-editor/helix/commit/09bc67a), [a7c3a43](https://github.com/helix-editor/helix/commit/a7c3a43), [3ceae88](https://github.com/helix-editor/helix/commit/3ceae88), [05ae617](https://github.com/helix-editor/helix/commit/05ae617), [5a1dcc2](https://github.com/helix-editor/helix/commit/5a1dcc2), [ebf96bd](https://github.com/helix-editor/helix/commit/ebf96bd), [#13644](https://github.com/helix-editor/helix/pull/13644), [b1f4717](https://github.com/helix-editor/helix/commit/b1f4717), [7410fe3](https://github.com/helix-editor/helix/commit/7410fe3), [633c5fb](https://github.com/helix-editor/helix/commit/633c5fb), [362e97e](https://github.com/helix-editor/helix/commit/362e97e), [#13828](https://github.com/helix-editor/helix/pull/13828), [6fd1efd](https://github.com/helix-editor/helix/commit/6fd1efd))
+    * This fixes a number of highlighter bugs.
+    * Locals like parameter highlights are now highlighted even when the definition is not in view.
+    * Markdown is now injected into rust doc comments (`///` and `//!`).
+* Add support for the DAP `startDebugging` reverse request ([#13403](https://github.com/helix-editor/helix/pull/13403))
+
+Commands:
+
+* Add `copy_between_registers` for interactive copying between two registers ([066e938](https://github.com/helix-editor/helix/commit/066e938))
+* Add `extend_to_file_{start,end}`, select-mode variants of `goto_file_{start,end}` ([#11767](https://github.com/helix-editor/helix/pull/11767))
+* Add `:!` alias for `:sh` and `:|` for `:pipe` ([#13263](https://github.com/helix-editor/helix/pull/13263))
+* Add `goto_column` and `extend_to_column` ([#13440](https://github.com/helix-editor/helix/pull/13440))
+* Add an `--insensitive`/`-i` flag to the `:sort` command ([#13560](https://github.com/helix-editor/helix/pull/13560))
+* Add `rotate_selections_first` and `rotate_selections_last` ([#13615](https://github.com/helix-editor/helix/pull/13615))
+* Add a `--no-format` flag for all `:write` commands ([2f56091](https://github.com/helix-editor/helix/commit/2f56091))
+* Add a `--skip-visible` flag for `:buffer-close-others` and `:buffer-close-others!` ([#5393](https://github.com/helix-editor/helix/pull/5393))
+
+Usability improvements:
+
+* Replace current file using `A-ret` in pickers rather than loading it in the background ([#12605](https://github.com/helix-editor/helix/pull/12605))
+* Set multiple selections when passing a file with multiple locations to `hx` ([#12192](https://github.com/helix-editor/helix/pull/12192))
+* Add path completion for multiple cursors ([#12550](https://github.com/helix-editor/helix/pull/12550), [c9dc940](https://github.com/helix-editor/helix/commit/c9dc940))
+* Truncate long prompt lines with "…" ([#12036](https://github.com/helix-editor/helix/pull/12036), [9d6ea77](https://github.com/helix-editor/helix/commit/9d6ea77), [0b9701e](https://github.com/helix-editor/helix/commit/0b9701e), [d3fb8fc](https://github.com/helix-editor/helix/commit/d3fb8fc))
+* Allow specifying languages in `:lsp-stop` and `:lsp-restart` ([#12578](https://github.com/helix-editor/helix/pull/12578), [3d7e273](https://github.com/helix-editor/helix/commit/3d7e273))
+* Add `m` (nearest matching pair) to infobox popups for `md` and `mr` ([#12650](https://github.com/helix-editor/helix/pull/12650))
+* Add a hint message in the statusline when using `:sort` on a single selection ([#12585](https://github.com/helix-editor/helix/pull/12585))
+* Avoid wrapping around in `goto_{next,prev}_diag` ([#12704](https://github.com/helix-editor/helix/pull/12704))
+* Support responses from multiple language servers for LSP goto-definition (and declaration, type definition and implementation) and goto-references ([f7394d5](https://github.com/helix-editor/helix/commit/f7394d5), [1a821ac](https://github.com/helix-editor/helix/commit/1a821ac), [d285a8a](https://github.com/helix-editor/helix/commit/d285a8a))
+* Show formatter errors in `:format` ([47f84d0](https://github.com/helix-editor/helix/commit/47f84d0))
+* Show typable command docs in keybinding infobox popups when the command takes no arguments ([e9c16b7](https://github.com/helix-editor/helix/commit/e9c16b7))
+* Add per-command titles to register selection infobox popups for `select_register`, `insert_register` and `copy_between_registers` ([e0da129](https://github.com/helix-editor/helix/commit/e0da129))
+* Add container name column to the LSP symbol picker ([#12930](https://github.com/helix-editor/helix/pull/12930))
+* Add a theme key for highlighting directories in completions and picker items ([#12855](https://github.com/helix-editor/helix/pull/12855), [7bebe0a](https://github.com/helix-editor/helix/commit/7bebe0a))
+* Add `editor.trim-final-newlines` and `editor.trim-trailing-whitespace` config options ([aa20eb8](https://github.com/helix-editor/helix/commit/aa20eb8))
+* Warn when the configured theme is unusable because true-color is not available ([#13058](https://github.com/helix-editor/helix/pull/13058))
+* Allow configuring `[workspace-]diagnostic` statusline element severities ([#13288](https://github.com/helix-editor/helix/pull/13288), [b0528bb](https://github.com/helix-editor/helix/commit/b0528bb))
+* Improve completion for shell commands ([#12883](https://github.com/helix-editor/helix/pull/12883), [532f241](https://github.com/helix-editor/helix/commit/532f241))
+* Show the primary selection index in the `selections` statusline element when there are multiple selections ([#12326](https://github.com/helix-editor/helix/pull/12326))
+* Use configured language server names when possible in `--health` output ([#13573](https://github.com/helix-editor/helix/pull/13573))
+* Add a statusline element for indentation style ([#13632](https://github.com/helix-editor/helix/pull/13632))
+* Set the working directory of language server commands to the workspace root ([#13691](https://github.com/helix-editor/helix/pull/13691))
+* Avoid jumpiness in the picker preview for languages with non-default tab widths ([#13761](https://github.com/helix-editor/helix/pull/13761))
+* Add a config option for limiting LSP inlay hint length ([#13742](https://github.com/helix-editor/helix/pull/13742))
+* Improve heuristics used in the diff gutter ([#13722](https://github.com/helix-editor/helix/pull/13722))
+* Allow moving a file with `:move` when its old path does not exist ([#13748](https://github.com/helix-editor/helix/pull/13748))
+* Allow moving a file into a directory with `:move` ([#13922](https://github.com/helix-editor/helix/pull/13922))
+* Show human-readable file sizes in the statusline message for file writes ([#13627](https://github.com/helix-editor/helix/pull/13627))
+* Add diagnostic source to the diagnosics pickers ([#13758](https://github.com/helix-editor/helix/pull/13758))
+* Show all active scopes under the cursor in `:tree-sitter-highlight-name` ([4a85171](https://github.com/helix-editor/helix/commit/4a85171))
+* Auto-close the LSP code-actions popup ([#13832](https://github.com/helix-editor/helix/pull/13832))
+* Add a configuration option for controlling atomic writes to disk ([#13656](https://github.com/helix-editor/helix/pull/13656))
+
+Fixes:
+
+* Fix panic from using `search_selection_detect_word_boundaries` (`*`) at the end of the file ([#12611](https://github.com/helix-editor/helix/pull/12611))
+* Discard placeholder text for zero tabstop `${0:placeholder}` ([#12647](https://github.com/helix-editor/helix/pull/12647))
+* Fix panic in `goto_file` (`gf`) on file names with non-ASCII characters ([#12673](https://github.com/helix-editor/helix/pull/12673))
+* Only accept unmodified characters in `goto_word` (`gw`) ([f5f9f49](https://github.com/helix-editor/helix/commit/f5f9f49), [0364521](https://github.com/helix-editor/helix/commit/0364521))
+* Skip recording keys pressed by macros while recording a macro ([#12733](https://github.com/helix-editor/helix/pull/12733))
+* Deny unknown fields in `editor.smart-tab` config ([28047fe](https://github.com/helix-editor/helix/commit/28047fe))
+* Fix soft-wrap word boundary detection for Unicode combining accent characters ([#12483](https://github.com/helix-editor/helix/pull/12483))
+* Fix clearing of infobox popups in `select_register` and `insert_register` commands ([e882a75](https://github.com/helix-editor/helix/commit/e882a75))
+* Fix handling of `stderr` of DAP child processes ([d0d1693](https://github.com/helix-editor/helix/commit/d0d1693))
+* Cancel all pending requests when a DAP session terminates ([26db541](https://github.com/helix-editor/helix/commit/26db541))
+* Properly discard out-of-date diagnostics ([313a647](https://github.com/helix-editor/helix/commit/313a647))
+* Fix display of multiple language servers in `hx --health` ([#12841](https://github.com/helix-editor/helix/pull/12841))
+* Respect `editor.default-yank-register` in `:yank-joined` ([#12890](https://github.com/helix-editor/helix/pull/12890))
+* Escape percent character when pasting the history register into the picker ([#12886](https://github.com/helix-editor/helix/pull/12886))
+* Render rulers before the cursor ([2d3b75a](https://github.com/helix-editor/helix/commit/2d3b75a))
+* Avoid inserting final newlines in empty files ([67879a1](https://github.com/helix-editor/helix/commit/67879a1))
+* Gracefully handle partial failure in multi-language-server requests ([#13156](https://github.com/helix-editor/helix/pull/13156), [14cab4b](https://github.com/helix-editor/helix/commit/14cab4b))
+* Improve LSP progress message display in the statusline ([#13180](https://github.com/helix-editor/helix/pull/13180))
+* Fix behavior of `<esc>` removing added indentation in documents with CRLF line endings ([702a961](https://github.com/helix-editor/helix/commit/702a961))
+* Append changes to document history before pushing jumplist jumps ([#13619](https://github.com/helix-editor/helix/pull/13619))
+* Fix overflow in the display of large chunks of text in the signature-help component ([#13566](https://github.com/helix-editor/helix/pull/13566))
+* Fix panic from clearing whitespace when changing multiple selections on one line ([#13673](https://github.com/helix-editor/helix/pull/13673))
+* Include formatting options in LSP range formatting request ([#13734](https://github.com/helix-editor/helix/pull/13734))
+* Consistently set statusline errors when LSP features are not available ([#12577](https://github.com/helix-editor/helix/pull/12577))
+* Fix `goto_file` on Windows ([#13770](https://github.com/helix-editor/helix/pull/13770))
+* Fix crash in `goto_word` (`gw`) when `editor.jump-label-alphabet` is configured to be empty ([#13863](https://github.com/helix-editor/helix/pull/13863))
+* Fix `open_above` / `open_below` (`o` / `O`) when using a count on a document with CRLF line-endings ([#13905](https://github.com/helix-editor/helix/pull/13905))
+
+Themes:
+
+* Update `modus` themes ([#12670](https://github.com/helix-editor/helix/pull/12670))
+* Update `snazzy` ([#11089](https://github.com/helix-editor/helix/pull/11089))
+* Update `gruber-darker` ([#12797](https://github.com/helix-editor/helix/pull/12797))
+* Update `cyan_light` ([#12864](https://github.com/helix-editor/helix/pull/12864), [#12891](https://github.com/helix-editor/helix/pull/12891))
+* Update `onedarker` ([#12833](https://github.com/helix-editor/helix/pull/12833))
+* Update `github_light` ([#12907](https://github.com/helix-editor/helix/pull/12907))
+* Update `kanagawa` ([#12895](https://github.com/helix-editor/helix/pull/12895))
+* Add `beans` ([#12963](https://github.com/helix-editor/helix/pull/12963))
+* Update `base16_transparent` ([#13080](https://github.com/helix-editor/helix/pull/13080))
+* Update `sunset` ([#13086](https://github.com/helix-editor/helix/pull/13086))
+* Add `carbon` ([#13067](https://github.com/helix-editor/helix/pull/13067))
+* Update `soralized` ([#13121](https://github.com/helix-editor/helix/pull/13121))
+* Add `focus_nova` ([#13144](https://github.com/helix-editor/helix/pull/13144))
+* Update `onedark` ([#13166](https://github.com/helix-editor/helix/pull/13166))
+* Update `adwaita-light` ([#13174](https://github.com/helix-editor/helix/pull/13174))
+* Add `earl_grey` ([#13203](https://github.com/helix-editor/helix/pull/13203))
+* Update `spacebones` ([#13213](https://github.com/helix-editor/helix/pull/13213))
+* Add `peachpuff` ([#13225](https://github.com/helix-editor/helix/pull/13225))
+* Update catppuccin themes ([#13262](https://github.com/helix-editor/helix/pull/13262))
+* Update gruvbox themes ([#13315](https://github.com/helix-editor/helix/pull/13315))
+* Update serika themes ([#13341](https://github.com/helix-editor/helix/pull/13341))
+* Add `gruvbox-material` ([#13311](https://github.com/helix-editor/helix/pull/13311))
+* Add `ashen` ([#13366](https://github.com/helix-editor/helix/pull/13366))
+* Update Zed themes ([#13370](https://github.com/helix-editor/helix/pull/13370))
+* Update Tokyonight themes ([#13375](https://github.com/helix-editor/helix/pull/13375))
+* Update `onelight` ([#13413](https://github.com/helix-editor/helix/pull/13413))
+* Add `ataraxia` ([#13390](https://github.com/helix-editor/helix/pull/13390))
+* Add `vesper` ([#13394](https://github.com/helix-editor/helix/pull/13394))
+* Add `kinda_nvim` and `kinda_nvim_light` ([#13406](https://github.com/helix-editor/helix/pull/13406))
+* Update `sonokai` ([#13410](https://github.com/helix-editor/helix/pull/13410))
+* Add `nyxvamp` themes ([#12185](https://github.com/helix-editor/helix/pull/12185))
+* Update nord themes ([#13574](https://github.com/helix-editor/helix/pull/13574))
+* Add `lapis_aquamarine` ([#13726](https://github.com/helix-editor/helix/pull/13726))
+* Add `sidra` ([#13575](https://github.com/helix-editor/helix/pull/13575))
+* Add `dark-synthwave` ([#13857](https://github.com/helix-editor/helix/pull/13857))
+* Update `rose_pine` ([#13908](https://github.com/helix-editor/helix/pull/13908))
+* Add `doom-one` ([#13933](https://github.com/helix-editor/helix/pull/13933))
+* Update `nightfox` ([#13957](https://github.com/helix-editor/helix/pull/13957))
+
+New languages:
+
+* Ghostty config ([#12703](https://github.com/helix-editor/helix/pull/12703))
+* Tera ([#12756](https://github.com/helix-editor/helix/pull/12756))
+* FGA ([#12763](https://github.com/helix-editor/helix/pull/12763))
+* CSV ([#11973](https://github.com/helix-editor/helix/pull/11973))
+* Yara ([#12753](https://github.com/helix-editor/helix/pull/12753))
+* Djot ([#12562](https://github.com/helix-editor/helix/pull/12562))
+* Ink ([#12773](https://github.com/helix-editor/helix/pull/12773))
+* Mail ([#12945](https://github.com/helix-editor/helix/pull/12945))
+* SourcePawn ([#13028](https://github.com/helix-editor/helix/pull/13028))
+* TLA+ ([#13081](https://github.com/helix-editor/helix/pull/13081))
+* Werk ([#13136](https://github.com/helix-editor/helix/pull/13136))
+* Debian control file ([#13245](https://github.com/helix-editor/helix/pull/13245))
+* WESL ([#13267](https://github.com/helix-editor/helix/pull/13267))
+* Fennel ([#13260](https://github.com/helix-editor/helix/pull/13260), [6081a5d](https://github.com/helix-editor/helix/commit/6081a5d))
+* Quarto ([#13339](https://github.com/helix-editor/helix/pull/13339))
+* Pug ([#13435](https://github.com/helix-editor/helix/pull/13435))
+* Slang ([#13449](https://github.com/helix-editor/helix/pull/13449))
+* Dunst config ([#13458](https://github.com/helix-editor/helix/pull/13458))
+* Luau ([#13702](https://github.com/helix-editor/helix/pull/13702))
+* Caddyfile ([#13859](https://github.com/helix-editor/helix/pull/13859))
+* Java properties ([#13874](https://github.com/helix-editor/helix/pull/13874))
+* Git notes ([#13885](https://github.com/helix-editor/helix/pull/13885))
+* systemd (split from INI) ([#13907](https://github.com/helix-editor/helix/pull/13907))
+* JSON-LD (split from JSON) ([#13925](https://github.com/helix-editor/helix/pull/13925))
+* Django HTML ([#13935](https://github.com/helix-editor/helix/pull/13935))
+
+Updated languages and queries:
+
+* Add `ruby-lsp` for Ruby ([#12511](https://github.com/helix-editor/helix/pull/12511))
+* Add `wat_server` for Wat ([#12581](https://github.com/helix-editor/helix/pull/12581))
+* Recognize `bun.lock` as JSONC ([fcf981b](https://github.com/helix-editor/helix/commit/fcf981b))
+* Update tree-sitter-rust ([#12607](https://github.com/helix-editor/helix/pull/12607), [1afa63d](https://github.com/helix-editor/helix/commit/1afa63d))
+* Fix configuration of `cs-lsp` ([#12615](https://github.com/helix-editor/helix/pull/12615))
+* Add `beancount-language-server` for Beancount ([#12610](https://github.com/helix-editor/helix/pull/12610))
+* Update tree-sitter-fish ([#12456](https://github.com/helix-editor/helix/pull/12456))
+* Add `fish-lsp` for Fish ([#12456](https://github.com/helix-editor/helix/pull/12456))
+* Update tree-sitter-ini ([#12456](https://github.com/helix-editor/helix/pull/12456), [#13088](https://github.com/helix-editor/helix/pull/13088))
+* Recognize `hgrc` as INI ([#12456](https://github.com/helix-editor/helix/pull/12456))
+* Restrict tagged template injection languages for ECMA languages ([#12217](https://github.com/helix-editor/helix/pull/12217))
+* Update tree-sitter-zig ([#11980](https://github.com/helix-editor/helix/pull/11980), [#12708](https://github.com/helix-editor/helix/pull/12708))
+* Update tree-sitter-elixir ([8bf9adf](https://github.com/helix-editor/helix/commit/8bf9adf))
+* Add `asm-lsp` for Assembly dialects ([#12684](https://github.com/helix-editor/helix/pull/12684))
+* Update tree-sitter-just ([#12692](https://github.com/helix-editor/helix/pull/12692), #)
+* Update tree-sitter-cairo ([#12712](https://github.com/helix-editor/helix/pull/12712))
+* Configure a comment token for Svelte ([#12743](https://github.com/helix-editor/helix/pull/12743))
+* Recognize `.sublime-*` files ([#12750](https://github.com/helix-editor/helix/pull/12750))
+* Highlight `$` tagged templates as shell commands in ECMA languages ([#12751](https://github.com/helix-editor/helix/pull/12751))
+* Add `#'` comment token for R ([#12748](https://github.com/helix-editor/helix/pull/12748))
+* Fix module/namespace highlight in Unison ([93fa990](https://github.com/helix-editor/helix/commit/93fa990))
+* Add missing `#not-eq?` and `#not-match?` highlights in TSQ ([3824010](https://github.com/helix-editor/helix/commit/3824010))
+* Reverse the precedence order of highlight queries ([#9458](https://github.com/helix-editor/helix/pull/9458), [#12777](https://github.com/helix-editor/helix/pull/12777), [#12795](https://github.com/helix-editor/helix/pull/12795), [144a4f4](https://github.com/helix-editor/helix/commit/144a4f4), [e1c26eb](https://github.com/helix-editor/helix/commit/e1c26eb), [e1060a2](https://github.com/helix-editor/helix/commit/e1060a2), [7f41670](https://github.com/helix-editor/helix/commit/7f41670), [#13293](https://github.com/helix-editor/helix/pull/13293))
+* Update Rust highlights ([b8bfc44](https://github.com/helix-editor/helix/commit/b8bfc44), [#12871](https://github.com/helix-editor/helix/pull/12871), [#13664](https://github.com/helix-editor/helix/pull/13664))
+* Add block comment configuration for PHP ([0ab403d](https://github.com/helix-editor/helix/commit/0ab403d))
+* Update Gren highlights ([#12769](https://github.com/helix-editor/helix/pull/12769))
+* Remove `ERROR` node highlighting from all highlight queries ([16ff063](https://github.com/helix-editor/helix/commit/16ff063))
+* Update tree-sitter-erlang and highlights ([18b9eb9](https://github.com/helix-editor/helix/commit/18b9eb9), [9f3b193](https://github.com/helix-editor/helix/commit/9f3b193), [12139a4](https://github.com/helix-editor/helix/commit/12139a4))
+* Update Nix injections ([#12776](https://github.com/helix-editor/helix/pull/12776), [#12774](https://github.com/helix-editor/helix/pull/12774), [#13851](https://github.com/helix-editor/helix/pull/13851))
+* Add indent queries for Nix ([#12829](https://github.com/helix-editor/helix/pull/12829))
+* Update Markdown highlights ([#12696](https://github.com/helix-editor/helix/pull/12696))
+* Recognize `xsl` as XML ([#12834](https://github.com/helix-editor/helix/pull/12834))
+* Remove deprecated `typst-lsp` config ([5a66270](https://github.com/helix-editor/helix/commit/5a66270))
+* Replace `pkgbuild-language-server` with `termux-language-server` ([c3c9a0d](https://github.com/helix-editor/helix/commit/c3c9a0d))
+* Update SQL highlights ([#12837](https://github.com/helix-editor/helix/pull/12837))
+* Recognize `mpd` and `smil` as XML ([#12916](https://github.com/helix-editor/helix/pull/12916))
+* Add indents and textojbects for Kotlin ([#12925](https://github.com/helix-editor/helix/pull/12925))
+* Fix module highlights in Koto ([7e87a36](https://github.com/helix-editor/helix/commit/7e87a36))
+* Update language servers for Protobuf ([#12936](https://github.com/helix-editor/helix/pull/12936))
+* Add `astro-ls` for Astro ([#12939](https://github.com/helix-editor/helix/pull/12939))
+* Fix recognition of "scons*" files as Python ([#12943](https://github.com/helix-editor/helix/pull/12943))
+* Update C# queries ([#12948](https://github.com/helix-editor/helix/pull/12948))
+* Add comment textojbect to TOML ([#12952](https://github.com/helix-editor/helix/pull/12952))
+* Add `starpls` as Starlark language server ([#12958](https://github.com/helix-editor/helix/pull/12958))
+* Add `pkl-lsp` for PKL ([#12962](https://github.com/helix-editor/helix/pull/12962))
+* Add `kdlfmt` formatter for KDL ([#12967](https://github.com/helix-editor/helix/pull/12967))
+* Update CSS highlights ([#12497](https://github.com/helix-editor/helix/pull/12497), [fed3edc](https://github.com/helix-editor/helix/commit/fed3edc))
+* Add `harper-ls` ([#13029](https://github.com/helix-editor/helix/pull/13029))
+* Change `wgsl_analyzer` to `wgsl-analyzer` ([#13063](https://github.com/helix-editor/helix/pull/13063))
+* Update tree-sitter-vhdl ([#13091](https://github.com/helix-editor/helix/pull/13091))
+* Update tree-sitter-openscad ([#13033](https://github.com/helix-editor/helix/pull/13033))
+* Update Rust injections ([694b615](https://github.com/helix-editor/helix/commit/694b615), [1bd7a39](https://github.com/helix-editor/helix/commit/1bd7a39))
+* Update Ruby highlights ([#13055](https://github.com/helix-editor/helix/pull/13055))
+* Recognize `gitconfig` as an extension ([#13115](https://github.com/helix-editor/helix/pull/13115))
+* Add `///` comment token for Amber ([#13122](https://github.com/helix-editor/helix/pull/13122))
+* Add indent queries for Starlark ([#13126](https://github.com/helix-editor/helix/pull/13126))
+* Recognize more systemd file types as INI ([#13139](https://github.com/helix-editor/helix/pull/13139))
+* Update scheme queries ([#13143](https://github.com/helix-editor/helix/pull/13143))
+* Recognize `tmTheme` as XML ([#13202](https://github.com/helix-editor/helix/pull/13202))
+* Update `golangci-lint` command for v2 ([#13204](https://github.com/helix-editor/helix/pull/13204))
+* Add `just-lsp` for Just ([#13276](https://github.com/helix-editor/helix/pull/13276))
+* Add a tree-sitter-prolog grammar ([#11611](https://github.com/helix-editor/helix/pull/11611))
+* Fix typos in Ada queries ([#13251](https://github.com/helix-editor/helix/pull/13251))
+* Update mint language server args ([#13248](https://github.com/helix-editor/helix/pull/13248))
+* Update typescript highlights ([#13250](https://github.com/helix-editor/helix/pull/13250))
+* Update tree-sitter-jjdescription ([#13329](https://github.com/helix-editor/helix/pull/13329))
+* Add injection queries for Quint ([#13322](https://github.com/helix-editor/helix/pull/13322))
+* Update tree-sitter-scss and highlights ([#13414](https://github.com/helix-editor/helix/pull/13414))
+* Update tree-sitter-go-mod ([#13395](https://github.com/helix-editor/helix/pull/13395))
+* Update tree-sitter-svelte ([#13423](https://github.com/helix-editor/helix/pull/13423))
+* Update Lua highlights ([#13401](https://github.com/helix-editor/helix/pull/13401))
+* Update Go highlights ([#13425](https://github.com/helix-editor/helix/pull/13425), [25b299a](https://github.com/helix-editor/helix/commit/25b299a), [#13825](https://github.com/helix-editor/helix/pull/13825))
+* Recognize `.git-blame-ignore-revs` as gitignore ([#13460](https://github.com/helix-editor/helix/pull/13460))
+* Update Verilog highlights ([#13473](https://github.com/helix-editor/helix/pull/13473), [#13493](https://github.com/helix-editor/helix/pull/13493))
+* Update tree-sitter-v ([#13469](https://github.com/helix-editor/helix/pull/13469))
+* Update WGSL highlights ([#13479](https://github.com/helix-editor/helix/pull/13479))
+* Update Bash highlights ([#13477](https://github.com/helix-editor/helix/pull/13477))
+* Update tree-sitter-cpp ([#13504](https://github.com/helix-editor/helix/pull/13504))
+* Update rust-analyzer config to use server-side file watching ([#13432](https://github.com/helix-editor/helix/pull/13432))
+* Update Vue injections ([#13511](https://github.com/helix-editor/helix/pull/13511))
+* Recognize `sld` as Scheme ([#13528](https://github.com/helix-editor/helix/pull/13528))
+* Recognize more files as git-attributes ([#13540](https://github.com/helix-editor/helix/pull/13540))
+* Update tree-sitter-haskell and queries ([#13475](https://github.com/helix-editor/helix/pull/13475))
+* Align INI highlights with TOML ([#13589](https://github.com/helix-editor/helix/pull/13589))
+* Add tree-sitter-rust-format-args for `format_args!` injections in Rust ([#13533](https://github.com/helix-editor/helix/pull/13533), [#13657](https://github.com/helix-editor/helix/pull/13657), [4dd4ba7](https://github.com/helix-editor/helix/commit/4dd4ba7), [86f10ae](https://github.com/helix-editor/helix/commit/86f10ae))
+* Update Ungrammar highlights ([8d58f6c](https://github.com/helix-editor/helix/commit/8d58f6c))
+* Add `ty` language server for Python ([#13643](https://github.com/helix-editor/helix/pull/13643))
+* Add `clarinet` language server for Clarity ([#13647](https://github.com/helix-editor/helix/pull/13647))
+* Update prisma config to avoid a crash in the language server ([f6878f6](https://github.com/helix-editor/helix/commit/f6878f6))
+* Add `pyrefly` for Python ([#13713](https://github.com/helix-editor/helix/pull/13713))
+* Update Python highlights ([#13715](https://github.com/helix-editor/helix/pull/13715))
+* Update Mojo language server and formatter to `pixi` ([#13648](https://github.com/helix-editor/helix/pull/13648))
+* Add `tombi` for TOML ([#13723](https://github.com/helix-editor/helix/pull/13723))
+* Add `neocmakelsp` for CMake ([#13740](https://github.com/helix-editor/helix/pull/13740))
+* Update C and C++ highlights ([#13747](https://github.com/helix-editor/helix/pull/13747), [#13772](https://github.com/helix-editor/helix/pull/13772))
+* Highlight escape sequences in ECMA languages ([#13762](https://github.com/helix-editor/helix/pull/13762))
+* Add an external formatter config for Crystal ([#13759](https://github.com/helix-editor/helix/pull/13759))
+* Add `amber-lsp` for Amber ([#13763](https://github.com/helix-editor/helix/pull/13763))
+* Update HTML highlights ([#13753](https://github.com/helix-editor/helix/pull/13753))
+* Update tree-sitter-purescript and highlights ([#13782](https://github.com/helix-editor/helix/pull/13782))
+* Update tree-sitter-gleam and highlights ([#13793](https://github.com/helix-editor/helix/pull/13793), [#13807](https://github.com/helix-editor/helix/pull/13807), [#13813](https://github.com/helix-editor/helix/pull/13813))
+* Recognize Buck files as Starlark ([#13810](https://github.com/helix-editor/helix/pull/13810))
+* Use tree-sitter-crystal instead of tree-sitter-ruby for Crystal and add custom queries ([#13805](https://github.com/helix-editor/helix/pull/13805))
+* Update tree-sitter-twig ([#13689](https://github.com/helix-editor/helix/pull/13689))
+* Recognize `jsconfig.json` as JSONC, use as JavaScript and JSX roots ([#13822](https://github.com/helix-editor/helix/pull/13822))
+* Recognize `.gem/credentials` as YAML ([#13843](https://github.com/helix-editor/helix/pull/13843))
+* Update Dockerfile injections ([#13845](https://github.com/helix-editor/helix/pull/13845), 13852)
+* Change tree-sitter parser for Git commit message files ([44293df](https://github.com/helix-editor/helix/commit/44293df))
+* Recognize `mimeapps.list` as INI ([#13850](https://github.com/helix-editor/helix/pull/13850))
+* Update tree-sitter-odin, highlights and indents ([#13877](https://github.com/helix-editor/helix/pull/13877), [#13917](https://github.com/helix-editor/helix/pull/13917))
+* Add locals queries for C, improve parameter highlighting ([#13876](https://github.com/helix-editor/helix/pull/13876))
+* Add textobjects for QML ([#13855](https://github.com/helix-editor/helix/pull/13855))
+* Add comment tokens for DTD ([#13904](https://github.com/helix-editor/helix/pull/13904))
+* Add `dts-lsp` for DeviceTree ([#13907](https://github.com/helix-editor/helix/pull/13907))
+* Update gomod highlights ([#13913](https://github.com/helix-editor/helix/pull/13913))
+* Recognize `compose.yaml` and `compose.yml` as Docker Compose ([#13930](https://github.com/helix-editor/helix/pull/13930))
+
+Packaging:
+
+* Fix handling of spaces in Bash completion ([#12828](https://github.com/helix-editor/helix/pull/12828))
+* Refactor Nix flake ([#12831](https://github.com/helix-editor/helix/pull/12831), [#13024](https://github.com/helix-editor/helix/pull/13024), [cb1ecc9](https://github.com/helix-editor/helix/commit/cb1ecc9), [#13305](https://github.com/helix-editor/helix/pull/13305))
+* Add `ConsoleOnly` to `Helix.desktop` categories ([#13236](https://github.com/helix-editor/helix/pull/13236))
+* Drop Nix flake dependency on flake-utils ([60a03a3](https://github.com/helix-editor/helix/commit/60a03a3))
+* Increase the MSRV to 1.82 ([#13275](https://github.com/helix-editor/helix/pull/13275))
+
+# 25.01.1 (2025-01-19)
+
+25.01.1 is a patch release focusing on fixing bugs and panics from changes in 25.01.
+
+Usability improvements:
+
+* Run external formatters from the document's directory ([#12315](https://github.com/helix-editor/helix/pull/12315))
+
+Fixes:
+
+* Fix blank buffer picker preview on doc with no views ([917174e](https://github.com/helix-editor/helix/commit/917174e))
+* Fix `join_selections` behavior on tabs ([#12452](https://github.com/helix-editor/helix/pull/12452))
+* Fix recognition for color LSP completion hex codes for some language servers ([#12501](https://github.com/helix-editor/helix/pull/12501))
+* Fix offsets to selections updated by `open_below`/`open_above` (`o`/`O`) in multi-cursor scenarios ([#12465](https://github.com/helix-editor/helix/pull/12465))
+* Fix offsets to selections updated by `insert_newline` when trimming whitespace in multi-cursor scenarios ([4bd17e5](https://github.com/helix-editor/helix/commit/4bd17e5))
+* Fix panic in path completion from resolving variables like `${HOME:-$HOME}` ([#12556](https://github.com/helix-editor/helix/pull/12556))
+* Prevent line comment continuation when using `change_selection` (`c`) on a line above a comment ([#12575](https://github.com/helix-editor/helix/pull/12575))
+
+Themes:
+
+* Update `onelight` ([#12399](https://github.com/helix-editor/helix/pull/12399))
+* Add cursorline color to iceberg themes ([#12404](https://github.com/helix-editor/helix/pull/12404))
+* Update `special`, `ui.text.directory` and `ui.virtual.wrap` in `dark_plus` ([#12530](https://github.com/helix-editor/helix/pull/12530))
+
+New languages:
+
+* CodeQL ([#12470](https://github.com/helix-editor/helix/pull/12470))
+* Gren ([#12525](https://github.com/helix-editor/helix/pull/12525))
+
+Updated languages and queries:
+
+* Fix Teal LSP name ([#12395](https://github.com/helix-editor/helix/pull/12395))
+* Highlight `:` in Rust as a delimiter ([#12408](https://github.com/helix-editor/helix/pull/12408))
+* Update Swift highlights ([#12409](https://github.com/helix-editor/helix/pull/12409))
+* Highlight JSX attributes as `@attribute` ([#12416](https://github.com/helix-editor/helix/pull/12416))
+* Improve markdown heading highlights ([#12417](https://github.com/helix-editor/helix/pull/12417))
+* Add comment tokens configuration for JSONC ([b26903c](https://github.com/helix-editor/helix/commit/b26903c))
+* Highlight the never type `!` as a type in Rust ([#12485](https://github.com/helix-editor/helix/pull/12485))
+* Expand builtin function highlights for ECMA languages, Rust and Haskell ([#12488](https://github.com/helix-editor/helix/pull/12488))
+* Recognize `.clang-tidy` as YAML ([#12498](https://github.com/helix-editor/helix/pull/12498))
+* Update MATLAB grammar and indent queries ([#12518](https://github.com/helix-editor/helix/pull/12518))
+* Recognize `rockspec` as Lua ([#12516](https://github.com/helix-editor/helix/pull/12516))
+* Add `///` to Dart comment tokens configuration ([99d33c7](https://github.com/helix-editor/helix/commit/99d33c7))
+* Update Solidity grammar and queries ([#12457](https://github.com/helix-editor/helix/pull/12457))
+* Update Spade grammar and queries ([#12583](https://github.com/helix-editor/helix/pull/12583))
+* Re-enable Hare fetching and building by default ([#11507](https://github.com/helix-editor/helix/pull/11507))
+
+Packaging:
+
+* `--version` now prints a leading zero for single-digit months, for example `25.01` (03f35af)
+* Pin the Ubuntu GitHub Actions runners used for releases to `ubuntu-22.04` ([#12464](https://github.com/helix-editor/helix/pull/12464))
+* Produce a Debian package (`.deb` file) in the release GitHub Actions workflow ([#12453](https://github.com/helix-editor/helix/pull/12453))
+
+# 25.01 (2025-01-03)
+
+As always, a big thank you to all of the contributors! This release saw changes from 171 contributors.
+
+Breaking changes:
+
+* The `editor.lsp.display-messages` key now controls messages sent with the LSP `window/showMessage` notification rather than progress messages. If you want to enable progress messages you should now enable the `editor.lsp.display-progress-messages` key instead. ([#5535](https://github.com/helix-editor/helix/pull/5535))
+
+Features:
+
+* Big refactor for `Picker`s ([#9647](https://github.com/helix-editor/helix/pull/9647), [#11209](https://github.com/helix-editor/helix/pull/11209), [#11216](https://github.com/helix-editor/helix/pull/11216), [#11211](https://github.com/helix-editor/helix/pull/11211), [#11343](https://github.com/helix-editor/helix/pull/11343), [#11406](https://github.com/helix-editor/helix/pull/11406))
+    * Use a table layout and allow filtering by column
+    * Reimplement `global_search` to allow changing the query dynamically
+* Add an alternative "inline" display for LSP diagnostics ([#6417](https://github.com/helix-editor/helix/pull/6417), [#11815](https://github.com/helix-editor/helix/pull/11815))
+* Support defining keybindings as macros ([#4709](https://github.com/helix-editor/helix/pull/4709))
+* Continue line comments in `o`/`O` and on `<ret>` in insert mode ([#10996](https://github.com/helix-editor/helix/pull/10996), [#12213](https://github.com/helix-editor/helix/pull/12213), [#12215](https://github.com/helix-editor/helix/pull/12215))
+* Allow configuring and switching clipboard providers at runtime ([#10839](https://github.com/helix-editor/helix/pull/10839), [b855cd0](https://github.com/helix-editor/helix/commit/b855cd0), [467fad5](https://github.com/helix-editor/helix/commit/467fad5), [191b0f0](https://github.com/helix-editor/helix/commit/191b0f0))
+* Add support for path completion ([#2608](https://github.com/helix-editor/helix/pull/2608))
+* Support bindings with the Super (Cmd/Win/Meta) modifier ([#6592](https://github.com/helix-editor/helix/pull/6592))
+* Support rendering and jumping between tabstops in snippet completions ([#9801](https://github.com/helix-editor/helix/pull/9801))
+* Allow theming directory completions ([#12205](https://github.com/helix-editor/helix/pull/12205), [#12295](https://github.com/helix-editor/helix/pull/12295))
+
+Commands:
+
+* Add commands to move within snake_case or camelCase words ([#8147](https://github.com/helix-editor/helix/pull/8147))
+* Add `search_selection_detect_word_boundaries` ([#12126](https://github.com/helix-editor/helix/pull/12126))
+    * This command takes the `*` key in normal and select mode, replacing `search_selection` which was moved to `A-*`.
+
+Usability improvements:
+
+* Add `:edit` and `:e` aliases for `:open` ([#11186](https://github.com/helix-editor/helix/pull/11186), [#11196](https://github.com/helix-editor/helix/pull/11196))
+* Trim trailing newline from pipe command outputs when the input doesn't have a trailing newline ([#11183](https://github.com/helix-editor/helix/pull/11183), [4f63a46](https://github.com/helix-editor/helix/commit/4f63a46))
+* Add `:mv` alias for `:move` ([#11256](https://github.com/helix-editor/helix/pull/11256))
+* Return document display name instead of absolute path from the `%` special register ([#11275](https://github.com/helix-editor/helix/pull/11275))
+* Track view position on a per-view instead of per-document basis ([#10559](https://github.com/helix-editor/helix/pull/10559))
+* Improve scrolloff calculation to leave a gap in the middle ([#11323](https://github.com/helix-editor/helix/pull/11323))
+* Show a popup for stderr printed by failed `:sh` commands ([#11239](https://github.com/helix-editor/helix/pull/11239))
+* Add statusline errors when nothing is selected with `s`, `K`, `A-K` ([#11370](https://github.com/helix-editor/helix/pull/11370))
+* Add `.svn` as a workspace root marker ([#11429](https://github.com/helix-editor/helix/pull/11429))
+* Trim the end of `:sh` outputs ([#11161](https://github.com/helix-editor/helix/pull/11161))
+* Show LSP `window/showMessage` messages in the statusline ([#5535](https://github.com/helix-editor/helix/pull/5535))
+* Support finding workspace directories via `.jj` directories ([#11685](https://github.com/helix-editor/helix/pull/11685))
+* Join single-line comments with `join_selections` (`J`) ([#11742](https://github.com/helix-editor/helix/pull/11742))
+* Show anonymous syntax tree nodes in `:tree-sitter-subtree` ([#11663](https://github.com/helix-editor/helix/pull/11663), [38e8382](https://github.com/helix-editor/helix/commit/38e8382))
+* Save an undo checkpoint before paste in insert mode ([#8121](https://github.com/helix-editor/helix/pull/8121))
+* Only break on ASCII spaces in `:reflow` ([#12048](https://github.com/helix-editor/helix/pull/12048))
+* Add a `default-yank-register` config option ([#11430](https://github.com/helix-editor/helix/pull/11430))
+* Show a statusline error for `:format` when a formatter is not available ([#12183](https://github.com/helix-editor/helix/pull/12183))
+* Change to the home directory with `:cd` with no arguments ([#12042](https://github.com/helix-editor/helix/pull/12042))
+* Change default comment token to `#` for unrecognized files ([#12080](https://github.com/helix-editor/helix/pull/12080), [#12266](https://github.com/helix-editor/helix/pull/12266), [bae6a58](https://github.com/helix-editor/helix/commit/bae6a58))
+* Trim all trailing whitespace on `insert_newline` ([#12177](https://github.com/helix-editor/helix/pull/12177))
+* Change to the prior directory with `:cd -` ([#12194](https://github.com/helix-editor/helix/pull/12194))
+* Allow parsing `-` (with no modifiers) as a keybinding ([#12191](https://github.com/helix-editor/helix/pull/12191))
+* Improve opening statusline and error messages when opening duplicate files or directories ([#12199](https://github.com/helix-editor/helix/pull/12199))
+* Trim trailing colons in paths passed on the argv ([#9963](https://github.com/helix-editor/helix/pull/9963))
+* Show tree-sitter parser availability in `hx --health <lang>` ([#12228](https://github.com/helix-editor/helix/pull/12228))
+* Show a preview block for colors in the LSP completion menu ([#12299](https://github.com/helix-editor/helix/pull/12299))
+* Add infobox help for `surround_add`, `surround_replace` and `surround_delete` ([#12262](https://github.com/helix-editor/helix/pull/12262))
+
+Fixes:
+
+* Respect document indentation settings in `format_selections` (`=`) ([#11169](https://github.com/helix-editor/helix/pull/11169))
+* Avoid switching the current document to normal mode during an LSP `workspace/applyEdit` operation ([#11176](https://github.com/helix-editor/helix/pull/11176))
+* Fix off-by-one in LSP `find_completion_range` ([#11266](https://github.com/helix-editor/helix/pull/11266))
+* Prefer file-system mtime to local system time for detecting external modifications ([#11142](https://github.com/helix-editor/helix/pull/11142), [#11352](https://github.com/helix-editor/helix/pull/11352), [#11358](https://github.com/helix-editor/helix/pull/11358), [#11361](https://github.com/helix-editor/helix/pull/11361))
+* Fix writing of hardlinks ([#11340](https://github.com/helix-editor/helix/pull/11340))
+* Prevent language servers from being automatically restarted when stopped with `:lsp-stop` ([#11321](https://github.com/helix-editor/helix/pull/11321))
+* Stable-sort LSP text edits ([#11357](https://github.com/helix-editor/helix/pull/11357))
+* Fix determination of current language layer in documents with nested language injections ([#11365](https://github.com/helix-editor/helix/pull/11365))
+* Fix a panic from `:move`ing a file to a new extension which starts a language server ([#11387](https://github.com/helix-editor/helix/pull/11387))
+* Fix a panic from duplicating the diff gutter ([#11092](https://github.com/helix-editor/helix/pull/11092))
+* Keep cursor position when exactly replacing text ([#5930](https://github.com/helix-editor/helix/pull/5930))
+* Fix a panic from `jump_backward` on a newly opened split ([#11508](https://github.com/helix-editor/helix/pull/11508))
+* Fix a panic from language servers sending an unknown diagnostic severity ([#11569](https://github.com/helix-editor/helix/pull/11569))
+* Fix a panic when drawing at the edge of the screen ([#11737](https://github.com/helix-editor/helix/pull/11737))
+* Fix git repo detection on symlinks ([#11732](https://github.com/helix-editor/helix/pull/11732))
+* Fix a panic from a language server sending an out-of-range active signature index in `textDocument/signatureHelp` ([#11825](https://github.com/helix-editor/helix/pull/11825))
+* Fix a panic from using `C-k` in a prompt ending in a multi-byte character ([#12237](https://github.com/helix-editor/helix/pull/12237))
+* Expand tildes in paths passed to `:read` ([#12271](https://github.com/helix-editor/helix/pull/12271))
+* Respect per-language `workspace-lsp-roots` configuration when opening new documents ([#12223](https://github.com/helix-editor/helix/pull/12223))
+* Consistently replace line-endings in paste/replace commands ([c262fe4](https://github.com/helix-editor/helix/commit/c262fe4))
+* Fix formatting in error statusline messages when inspecting variables in DAP ([#12354](https://github.com/helix-editor/helix/pull/12354))
+* Fix invisible printing of headers in `--health` output on light terminals ([#12355](https://github.com/helix-editor/helix/pull/12355))
+* Accept integers serialized as floats in the JSONRPC `id` field ([#12376](https://github.com/helix-editor/helix/pull/12376))
+
+Themes:
+
+* Bring `kanagawa` colors better in line with neovim version ([#11187](https://github.com/helix-editor/helix/pull/11187), [#11270](https://github.com/helix-editor/helix/pull/11270))
+* Add `ao` ([#11063](https://github.com/helix-editor/helix/pull/11063))
+* Update `dark_plus` ([#11415](https://github.com/helix-editor/helix/pull/11415))
+* Add `iceberg-light` and `iceberg-dark` ([#10674](https://github.com/helix-editor/helix/pull/10674))
+* Update everforest themes ([#11459](https://github.com/helix-editor/helix/pull/11459))
+* Update gruvbox themes ([#11477](https://github.com/helix-editor/helix/pull/11477))
+* Change primary selection cursor color for `naysayer` ([#11617](https://github.com/helix-editor/helix/pull/11617))
+* Style picker column names in `horizon-dark` ([#11649](https://github.com/helix-editor/helix/pull/11649))
+* Style picker column names in Darcula themes ([#11649](https://github.com/helix-editor/helix/pull/11649))
+* Update diagnostics colors in `snazzy` ([#11731](https://github.com/helix-editor/helix/pull/11731))
+* Update bogster themes ([#11353](https://github.com/helix-editor/helix/pull/11353))
+* Highlight `keyword.storage` in `onedark` ([#11802](https://github.com/helix-editor/helix/pull/11802))
+* Add `ui.virtual.jump-label` to `serika-dark` ([#11911](https://github.com/helix-editor/helix/pull/11911))
+* Add `adwaita-light` ([#10869](https://github.com/helix-editor/helix/pull/10869))
+* Add seoul256 themes ([#11466](https://github.com/helix-editor/helix/pull/11466))
+* Add yo themes ([#11703](https://github.com/helix-editor/helix/pull/11703))
+* Add `eiffel` ([#11679](https://github.com/helix-editor/helix/pull/11679))
+* Add `carbonfox` ([#11558](https://github.com/helix-editor/helix/pull/11558))
+* Set tags color in monokai themes ([#11917](https://github.com/helix-editor/helix/pull/11917))
+* Improve readability of spacebones picker selection ([#12064](https://github.com/helix-editor/helix/pull/12064))
+* Update modus themes ([#11949](https://github.com/helix-editor/helix/pull/11949))
+* Use bold for statusline mode indicator in `onedarker` ([#11958](https://github.com/helix-editor/helix/pull/11958))
+* Update hex themes, add a new hex theme ([#10849](https://github.com/helix-editor/helix/pull/10849))
+* Add `sunset` ([#12093](https://github.com/helix-editor/helix/pull/12093))
+* Add bufferline highlighting for flexoki themes ([#12146](https://github.com/helix-editor/helix/pull/12146))
+* Add colors for (un)checked list items to catppuccin themes ([#12167](https://github.com/helix-editor/helix/pull/12167))
+* Update `voxed` ([#9328](https://github.com/helix-editor/helix/pull/9328))
+* Add `vintage` ([#9361](https://github.com/helix-editor/helix/pull/9361))
+* Add directory style to everforest themes ([#12287](https://github.com/helix-editor/helix/pull/12287))
+* Add inactive text and update jump label highlights in `dark_plus` ([#12289](https://github.com/helix-editor/helix/pull/12289))
+* Sync changes with catppuccin themes ([#12304](https://github.com/helix-editor/helix/pull/12304))
+* Add `ui.text.directory` to `nightfox` ([#12328](https://github.com/helix-editor/helix/pull/12328))
+* Add `ui.text.directory` to `sunset` ([#12328](https://github.com/helix-editor/helix/pull/12328))
+* Add `diagnostic.unnecessary` to Catppuccin themes ([#12391](https://github.com/helix-editor/helix/pull/12391))
+
+New languages:
+
+* `jjdescription` ([#11271](https://github.com/helix-editor/helix/pull/11271), [#11857](https://github.com/helix-editor/helix/pull/11857), [#12305](https://github.com/helix-editor/helix/pull/12305))
+* i3wm and Sway configs ([#11424](https://github.com/helix-editor/helix/pull/11424))
+* TypeSpec ([#11412](https://github.com/helix-editor/helix/pull/11412))
+* jq ([#11393](https://github.com/helix-editor/helix/pull/11393))
+* Thrift ([#11367](https://github.com/helix-editor/helix/pull/11367))
+* Gherkin ([#11083](https://github.com/helix-editor/helix/pull/11083))
+* Circom ([#11676](https://github.com/helix-editor/helix/pull/11676))
+* Dune ([#11829](https://github.com/helix-editor/helix/pull/11829))
+* Snakemake ([#11858](https://github.com/helix-editor/helix/pull/11858), [#11936](https://github.com/helix-editor/helix/pull/11936))
+* Cylc ([#11830](https://github.com/helix-editor/helix/pull/11830))
+* textproto ([#11874](https://github.com/helix-editor/helix/pull/11874))
+* Spade ([#11448](https://github.com/helix-editor/helix/pull/11448), [#12276](https://github.com/helix-editor/helix/pull/12276))
+* NestedText ([#11987](https://github.com/helix-editor/helix/pull/11987))
+* Quint ([#11898](https://github.com/helix-editor/helix/pull/11898))
+* Amber-lang ([#12021](https://github.com/helix-editor/helix/pull/12021))
+* Vento ([#12147](https://github.com/helix-editor/helix/pull/12147))
+* Teal ([#12081](https://github.com/helix-editor/helix/pull/12081))
+* Koto ([#12307](https://github.com/helix-editor/helix/pull/12307))
+* NGINX ([#12309](https://github.com/helix-editor/helix/pull/12309))
+
+Updated languages and queries:
+
+* Add comment injections for Hare ([#11173](https://github.com/helix-editor/helix/pull/11173))
+* Improve highlights for `blade.php` files ([#11138](https://github.com/helix-editor/helix/pull/11138))
+* Update tree-sitter-slint ([#11224](https://github.com/helix-editor/helix/pull/11224), [#11757](https://github.com/helix-editor/helix/pull/11757), [#12297](https://github.com/helix-editor/helix/pull/12297))
+* Recognize `just` files as Just ([#11286](https://github.com/helix-editor/helix/pull/11286))
+* Recognize `mdx` as Markdown ([#11122](https://github.com/helix-editor/helix/pull/11122))
+* Update Just grammar and queries ([#11306](https://github.com/helix-editor/helix/pull/11306))
+* Recognize `tclsh` as TCL ([#11236](https://github.com/helix-editor/helix/pull/11236))
+* Update Godot grammar and queries ([#11235](https://github.com/helix-editor/helix/pull/11235))
+* Update Gleam grammar and queries ([#11427](https://github.com/helix-editor/helix/pull/11427))
+* Add `mesonlsp` for Meson ([#11416](https://github.com/helix-editor/helix/pull/11416))
+* Update HTML highlights ([#11400](https://github.com/helix-editor/helix/pull/11400))
+* Add comment textobjects for Verilog ([#11388](https://github.com/helix-editor/helix/pull/11388))
+* Switch tree-sitter-just grammar ([#11380](https://github.com/helix-editor/helix/pull/11380), [#11606](https://github.com/helix-editor/helix/pull/11606), [#12141](https://github.com/helix-editor/helix/pull/12141))
+* Update tree-sitter-fsharp ([#11061](https://github.com/helix-editor/helix/pull/11061))
+* Add `nixd` for Nix ([#10767](https://github.com/helix-editor/helix/pull/10767))
+* Highlight types and enum members from the Rust prelude ([#8535](https://github.com/helix-editor/helix/pull/8535))
+* Improve textobjects for HCL, Nix ([#11513](https://github.com/helix-editor/helix/pull/11513))
+* Add textobjects queries for docker-compose, dockerfile, env, git-config, hcl, hocon, prisma, SQL and YAML ([#11513](https://github.com/helix-editor/helix/pull/11513))
+* Recognize cshtml files as HTML ([#11540](https://github.com/helix-editor/helix/pull/11540))
+* Set a memory limit for the Lean language server ([#11683](https://github.com/helix-editor/helix/pull/11683))
+* Add configurations for jedi and ruff language servers ([#11630](https://github.com/helix-editor/helix/pull/11630))
+* Update Vue highlights ([#11706](https://github.com/helix-editor/helix/pull/11706))
+* Switch tree-sitter-hcl grammar ([#11749](https://github.com/helix-editor/helix/pull/11749))
+* Fix `odinfmt` formatter configuration ([#11759](https://github.com/helix-editor/helix/pull/11759))
+* Recognize `rbs` files as Ruby ([#11786](https://github.com/helix-editor/helix/pull/11786))
+* Update tree-sitter-nickel ([#11771](https://github.com/helix-editor/helix/pull/11771))
+* Recognize `ldtk` and `ldtkl` files as JSON ([#11793](https://github.com/helix-editor/helix/pull/11793))
+* Fix highlights for builtin functions in Fish ([#11792](https://github.com/helix-editor/helix/pull/11792))
+* Add `superhtml` for HTML ([#11609](https://github.com/helix-editor/helix/pull/11609))
+* Add a configuration for the Vale language server ([#11636](https://github.com/helix-editor/helix/pull/11636))
+* Add Erlang Language Platform (`elp`) for Erlang ([#11499](https://github.com/helix-editor/helix/pull/11499))
+* Update Odin highlights ([#11804](https://github.com/helix-editor/helix/pull/11804))
+* Remove auto-pairs for single quotes in SML ([#11838](https://github.com/helix-editor/helix/pull/11838))
+* Add `glsl_analyzer` for GLSL ([#11891](https://github.com/helix-editor/helix/pull/11891))
+* Recognize `.prettierrc` as YAML ([#11997](https://github.com/helix-editor/helix/pull/11997))
+* Fix `swift-format` formatter configuration ([#12052](https://github.com/helix-editor/helix/pull/12052))
+* Add `package.json` and `tsconfig.json` as JS/TS workspace roots ([#10652](https://github.com/helix-editor/helix/pull/10652))
+* Add "INVARIANT" to comment error highlights ([#12094](https://github.com/helix-editor/helix/pull/12094))
+* Update Rescript grammar and queries ([#11165](https://github.com/helix-editor/helix/pull/11165))
+* Update tree-sitter-nasm ([#11795](https://github.com/helix-editor/helix/pull/11795))
+* Update LLVM grammars ([#11851](https://github.com/helix-editor/helix/pull/11851))
+* Update Perl and Pod grammars ([#11848](https://github.com/helix-editor/helix/pull/11848))
+* Add Nim injections in Nix ([#11837](https://github.com/helix-editor/helix/pull/11837))
+* Recognize `livemd` as Markdown ([#12034](https://github.com/helix-editor/helix/pull/12034))
+* Update Unison grammar and queries ([#12039](https://github.com/helix-editor/helix/pull/12039))
+* Turn off Swift auto-format by default ([#12071](https://github.com/helix-editor/helix/pull/12071))
+* Recognize `.swift-format` as JSON ([#12071](https://github.com/helix-editor/helix/pull/12071))
+* Recognize `.clangd` and `.clang-format` as YAML ([#12032](https://github.com/helix-editor/helix/pull/12032))
+* Recognize `ssh_config.d/*.conf` as sshclientconfig ([#11947](https://github.com/helix-editor/helix/pull/11947))
+* Update comment token configs for Zig ([#12049](https://github.com/helix-editor/helix/pull/12049))
+* Update tree-sitter-bicep ([#11525](https://github.com/helix-editor/helix/pull/11525))
+* Add `hyperls` for Hyperlang ([#11056](https://github.com/helix-editor/helix/pull/11056))
+* Add highlight queries for Solidity ([#12102](https://github.com/helix-editor/helix/pull/12102))
+* Recognize `WORKSPACE.bzlmod` as Starlark ([#12103](https://github.com/helix-editor/helix/pull/12103))
+* Update Ada grammar and queries ([#12131](https://github.com/helix-editor/helix/pull/12131))
+* Restrict Hocon file-types glob patterns ([#12156](https://github.com/helix-editor/helix/pull/12156))
+* Update Mojo language server to Magic ([#12195](https://github.com/helix-editor/helix/pull/12195))
+* Switch tree-sitter-v grammar ([#12236](https://github.com/helix-editor/helix/pull/12236))
+* Add "COMPLIANCE" to comment error highlights ([#12094](https://github.com/helix-editor/helix/pull/12094))
+* Add a language server configuration for `ltex-ls-plus` ([#12251](https://github.com/helix-editor/helix/pull/12251))
+* Update tree-sitter-dockerfile ([#12230](https://github.com/helix-editor/helix/pull/12230))
+* Add `]` to PHP outdents ([#12286](https://github.com/helix-editor/helix/pull/12286))
+* Add textobjects for Odin ([#12302](https://github.com/helix-editor/helix/pull/12302))
+* Update tree-sitter-heex and queries ([#12334](https://github.com/helix-editor/helix/pull/12334))
+* Update protobuf highlights ([#12339](https://github.com/helix-editor/helix/pull/12339))
+* Switch tree-sitter-query (TSQ) grammar ([#12148](https://github.com/helix-editor/helix/pull/12148), [e0bccd2](https://github.com/helix-editor/helix/commit/e0bccd2))
+* Add block comment configurations for jinja and nunjucks ([#12348](https://github.com/helix-editor/helix/pull/12348))
+* Add `uv` shebang for python ([#12360](https://github.com/helix-editor/helix/pull/12360))
+* Update tree-sitter-vento ([#12368](https://github.com/helix-editor/helix/pull/12368))
+* Switch Protobuf tree-sitter grammar ([#12225](https://github.com/helix-editor/helix/pull/12225))
+* Recognize `hypr/*.conf` as Hyprland ([#12384](https://github.com/helix-editor/helix/pull/12384))
+
+Packaging:
+
+* Add completions for Nushell ([#11262](https://github.com/helix-editor/helix/pull/11262), [#11346](https://github.com/helix-editor/helix/pull/11346))
+* Fix completion of flags in Bash completions ([#11246](https://github.com/helix-editor/helix/pull/11246))
+* Include shell completions in Nix outputs ([#11518](https://github.com/helix-editor/helix/pull/11518))
+
 # 24.07 (2024-07-14)
 
 Thanks to all of the contributors! This release has changes from 160 contributors.
@@ -16,7 +615,7 @@ Features:
 Commands:
 
 - `select_all_siblings` (`A-a`) - select all siblings of each selection ([87c4161](https://github.com/helix-editor/helix/commit/87c4161))
-- `select_all_children` (`A-i`) - select all children of each selection ([fa67c5c](https://github.com/helix-editor/helix/commit/fa67c5c))
+- `select_all_children` (`A-I`) - select all children of each selection ([fa67c5c](https://github.com/helix-editor/helix/commit/fa67c5c))
 - `:read` - insert the contents of the given file at each selection ([#10447](https://github.com/helix-editor/helix/pull/10447))
 
 Usability improvements:
@@ -436,7 +1035,7 @@ Updated languages and queries:
 - Recognize common Dockerfile file types ([#9772](https://github.com/helix-editor/helix/pull/9772))
 - Recognize NUON files as Nu ([#9839](https://github.com/helix-editor/helix/pull/9839))
 - Add textobjects for Java native functions and constructors ([#9806](https://github.com/helix-editor/helix/pull/9806))
-- Fix "braket" typeo in JSX highlights ([#9910](https://github.com/helix-editor/helix/pull/9910))
+- Fix "braket" typo in JSX highlights ([#9910](https://github.com/helix-editor/helix/pull/9910))
 - Update tree-sitter-hurl ([#9775](https://github.com/helix-editor/helix/pull/9775))
 - Add textobjects queries for Vala ([#8541](https://github.com/helix-editor/helix/pull/8541))
 - Update tree-sitter-git-config ([9610254](https://github.com/helix-editor/helix/commit/9610254))
@@ -643,7 +1242,7 @@ Updated languages and queries:
 - Add Fortran comment injections ([#7305](https://github.com/helix-editor/helix/pull/7305))
 - Switch Vue language server to `vue-language-server` ([#7312](https://github.com/helix-editor/helix/pull/7312))
 - Update tree-sitter-sql ([#7387](https://github.com/helix-editor/helix/pull/7387), [#8464](https://github.com/helix-editor/helix/pull/8464))
-- Replace the MATLAB tre-sitter grammar ([#7388](https://github.com/helix-editor/helix/pull/7388), [#7442](https://github.com/helix-editor/helix/pull/7442), [#7491](https://github.com/helix-editor/helix/pull/7491), [#7493](https://github.com/helix-editor/helix/pull/7493), [#7511](https://github.com/helix-editor/helix/pull/7511), [#7532](https://github.com/helix-editor/helix/pull/7532), [#8040](https://github.com/helix-editor/helix/pull/8040))
+- Replace the MATLAB tree-sitter grammar ([#7388](https://github.com/helix-editor/helix/pull/7388), [#7442](https://github.com/helix-editor/helix/pull/7442), [#7491](https://github.com/helix-editor/helix/pull/7491), [#7493](https://github.com/helix-editor/helix/pull/7493), [#7511](https://github.com/helix-editor/helix/pull/7511), [#7532](https://github.com/helix-editor/helix/pull/7532), [#8040](https://github.com/helix-editor/helix/pull/8040))
 - Highlight TOML table headers ([#7441](https://github.com/helix-editor/helix/pull/7441))
 - Recognize `cppm` file-type as C++ ([#7492](https://github.com/helix-editor/helix/pull/7492))
 - Refactor ecma language queries into private and public queries ([#7207](https://github.com/helix-editor/helix/pull/7207))
@@ -1130,7 +1729,7 @@ Features:
 - Support underline styles and colors ([#4061](https://github.com/helix-editor/helix/pull/4061), [98c121c](https://github.com/helix-editor/helix/commit/98c121c))
 - Inheritance for themes ([#3067](https://github.com/helix-editor/helix/pull/3067), [#4096](https://github.com/helix-editor/helix/pull/4096))
 - Cursorcolumn ([#4084](https://github.com/helix-editor/helix/pull/4084))
-- Overhauled system for writing files and quiting ([#2267](https://github.com/helix-editor/helix/pull/2267), [#4397](https://github.com/helix-editor/helix/pull/4397))
+- Overhauled system for writing files and quitting ([#2267](https://github.com/helix-editor/helix/pull/2267), [#4397](https://github.com/helix-editor/helix/pull/4397))
 - Autosave when terminal loses focus ([#3178](https://github.com/helix-editor/helix/pull/3178))
 - Use OSC52 as a fallback for the system clipboard ([#3220](https://github.com/helix-editor/helix/pull/3220))
 - Show git diffs in the gutter ([#3890](https://github.com/helix-editor/helix/pull/3890), [#5012](https://github.com/helix-editor/helix/pull/5012), [#4995](https://github.com/helix-editor/helix/pull/4995))
@@ -1285,7 +1884,7 @@ Themes:
 - Update `pop-dark` ([#4323](https://github.com/helix-editor/helix/pull/4323))
 - Update `rose_pine` ([#4221](https://github.com/helix-editor/helix/pull/4221))
 - Add `kanagawa` ([#4300](https://github.com/helix-editor/helix/pull/4300))
-- Add `hex_steel`, `hex_toxic` and `hex_lavendar` ([#4367](https://github.com/helix-editor/helix/pull/4367), [#4990](https://github.com/helix-editor/helix/pull/4990))
+- Add `hex_steel`, `hex_toxic` and `hex_lavender` ([#4367](https://github.com/helix-editor/helix/pull/4367), [#4990](https://github.com/helix-editor/helix/pull/4990))
 - Update `tokyonight` and `tokyonight_storm` ([#4415](https://github.com/helix-editor/helix/pull/4415))
 - Update `gruvbox` ([#4626](https://github.com/helix-editor/helix/pull/4626))
 - Update `dark_plus` ([#4661](https://github.com/helix-editor/helix/pull/4661), [#4678](https://github.com/helix-editor/helix/pull/4678))
@@ -1452,7 +2051,7 @@ Usability improvements and fixes:
 - Introduce `keyword.storage` highlight scope ([#2731](https://github.com/helix-editor/helix/pull/2731))
 - Handle symlinks more consistently ([#2718](https://github.com/helix-editor/helix/pull/2718))
 - Improve markdown list rendering ([#2687](https://github.com/helix-editor/helix/pull/2687))
-- Update auto-pairs and idle-timout settings when the config is reloaded ([#2736](https://github.com/helix-editor/helix/pull/2736))
+- Update auto-pairs and idle-timeout settings when the config is reloaded ([#2736](https://github.com/helix-editor/helix/pull/2736))
 - Fix panic on closing last buffer ([#2658](https://github.com/helix-editor/helix/pull/2658))
 - Prevent modifying jumplist until jumping to a reference ([#2670](https://github.com/helix-editor/helix/pull/2670))
 - Ensure `:quit` and `:quit!` take no arguments ([#2654](https://github.com/helix-editor/helix/pull/2654))

Cargo.toml

@@ -5,6 +5,7 @@ members = [
   "helix-view",
   "helix-term",
   "helix-tui",
+  "helix-lsp-types",
   "helix-lsp",
   "helix-event",
   "helix-dap",
@@ -21,13 +22,12 @@ default-members = [
 
 [profile.release]
 lto = "thin"
-# debug = true
 
 [profile.opt]
 inherits = "release"
 lto = "fat"
 codegen-units = 1
-# strip = "debuginfo" # TODO: or strip = true
+strip = true
 opt-level = 3
 
 [profile.integration]
@@ -37,17 +37,28 @@ package.helix-tui.opt-level = 2
 package.helix-term.opt-level = 2
 
 [workspace.dependencies]
-tree-sitter = { version = "0.22" }
+tree-house = { version = "0.3.0", default-features = false }
 nucleo = "0.5.0"
 slotmap = "1.0.7"
-thiserror = "1.0"
+thiserror = "2.0"
+tempfile = "3.21.0"
+bitflags = "2.9"
+unicode-segmentation = "1.2"
+ropey = { version = "1.6.1", default-features = false, features = ["simd"] }
+foldhash = "0.1"
+parking_lot = "0.12"
+futures-executor = "0.3"
+futures-util = { version = "0.3", features = ["std", "async-await"], default-features = false }
+tokio-stream = "0.1.17"
+toml = "0.9"
+termina = "0.1.0"
 
 [workspace.package]
-version = "24.7.0"
+version = "25.7.1"
 edition = "2021"
 authors = ["Blaž Hrastnik <blaz@mxxn.io>"]
 categories = ["editor"]
 repository = "https://github.com/helix-editor/helix"
 homepage = "https://helix-editor.com"
 license = "MPL-2.0"
-rust-version = "1.70"
+rust-version = "1.82"

README.md

@@ -37,8 +37,8 @@ All shortcuts/keymaps can be found [in the documentation on the website](https:/
 - Built-in language server support
 - Smart, incremental syntax highlighting and code editing via tree-sitter
 
-It's a terminal-based editor first, but I'd like to explore a custom renderer
-(similar to Emacs) in wgpu or skulpin.
+Although it's primarily a terminal-based editor, I am interested in exploring
+a custom renderer (similar to Emacs) using wgpu or skulpin.
 
 Note: Only certain languages have indentation definitions at the moment. Check
 `runtime/queries/<lang>/` for `indents.scm`.
@@ -47,7 +47,7 @@ Note: Only certain languages have indentation definitions at the moment. Check
 
 [Installation documentation](https://docs.helix-editor.com/install.html).
 
-[![Packaging status](https://repology.org/badge/vertical-allrepos/helix.svg?exclude_unsupported=1)](https://repology.org/project/helix/versions)
+[![Packaging status](https://repology.org/badge/vertical-allrepos/helix-editor.svg?exclude_unsupported=1)](https://repology.org/project/helix-editor/versions)
 
 # Contributing
 

book/book.toml

@@ -1,7 +1,6 @@
 [book]
 authors = ["Blaž Hrastnik"]
 language = "en"
-multilingual = false
 src = "src"
 
 [output.html]

book/src/SUMMARY.md

@@ -10,10 +10,15 @@
   - [Surround](./surround.md)
   - [Textobjects](./textobjects.md)
   - [Syntax aware motions](./syntax-aware-motions.md)
+  - [Pickers](./pickers.md)
+  - [Jumplist](./jumplist.md)
   - [Keymap](./keymap.md)
+  - [Command line](./command-line.md)
   - [Commands](./commands.md)
   - [Language support](./lang-support.md)
-- [Migrating from Vim](./from-vim.md)
+- [Ecosystem](./ecosystem.md)
+  - [Migrating from Vim](./from-vim.md)
+  - [Helix mode in other software](./other-software.md)
 - [Configuration](./configuration.md)
   - [Editor](./editor.md)
   - [Themes](./themes.md)
@@ -24,3 +29,5 @@
   - [Adding textobject queries](./guides/textobject.md)
   - [Adding indent queries](./guides/indent.md)
   - [Adding injection queries](./guides/injection.md)
+  - [Adding tags queries](./guides/tags.md)
+  - [Adding rainbow bracket queries](./guides/rainbow_bracket_queries.md)

book/src/building-from-source.md

@@ -7,6 +7,7 @@
   - [Note to packagers](#note-to-packagers)
 - [Validating the installation](#validating-the-installation)
 - [Configure the desktop shortcut](#configure-the-desktop-shortcut)
+- [Building the Debian package](#building-the-debian-package)
 
 Requirements:
 
@@ -34,10 +35,19 @@ RUSTFLAGS="-C target-feature=-crt-static"
 2. Compile from source:
 
    ```sh
+   # Reproducible
    cargo install --path helix-term --locked
    ```
+   ```sh
+   # Optimized
+   cargo install \
+      --profile opt \
+      --config 'build.rustflags=["-C", "target-cpu=native"]' \
+      --path helix-term \
+      --locked
+   ```
 
-   This command will create the `hx` executable and construct the tree-sitter
+   Either command will create the `hx` executable and construct the tree-sitter
    grammars in the local `runtime` folder.
 
 > 💡 If you do not want to fetch or build grammars, set an environment variable `HELIX_DISABLE_AUTO_GRAMMAR_BUILD`
@@ -63,11 +73,9 @@ export HELIX_RUNTIME=~/src/helix/runtime
 Or, create a symbolic link:
 
 ```sh
-ln -Ts $PWD/runtime ~/.config/helix/runtime
+ln -Tsf $PWD/runtime ~/.config/helix/runtime
 ```
 
-If the above command fails to create a symbolic link because the file exists either move `~/.config/helix/runtime` to a new location or delete it, then run the symlink command above again.
-
 #### Windows
 
 Either set the `HELIX_RUNTIME` environment variable to point to the runtime files using the Windows setting (search for
@@ -75,7 +83,7 @@ Either set the `HELIX_RUNTIME` environment variable to point to the runtime file
 Cmd:
 
 ```sh
-setx HELIX_RUNTIME "%userprofile%\source\repos\helix\runtime"
+setx HELIX_RUNTIME "%userprofile%\src\helix\runtime"
 ```
 
 > 💡 `%userprofile%` resolves to your user directory like
@@ -117,7 +125,7 @@ to package the runtime into `/usr/lib/helix/runtime`. The rough steps a build
 script could follow are:
 
 1. `export HELIX_DEFAULT_RUNTIME=/usr/lib/helix/runtime`
-1. `cargo build --profile opt --locked --path helix-term`
+1. `cargo build --profile opt --locked`
 1. `cp -r runtime $BUILD_DIR/usr/lib/helix/`
 1. `cp target/opt/hx $BUILD_DIR/usr/bin/hx`
 
@@ -162,3 +170,39 @@ file. For example, to use `kitty`:
 sed -i "s|Exec=hx %F|Exec=kitty hx %F|g" ~/.local/share/applications/Helix.desktop
 sed -i "s|Terminal=true|Terminal=false|g" ~/.local/share/applications/Helix.desktop
 ```
+
+### Building the Debian package
+
+If the `.deb` file provided on the release page uses a `libc` version higher
+than that used by your Debian, Ubuntu, or Mint system, you can build the package
+from source to match your system's dependencies.
+
+Install `cargo-deb`, the tool used for building the `.deb` file:
+
+```sh
+cargo install cargo-deb
+```
+
+After cloning and entering the Helix repository as previously described,
+use the following command to build the release binary and package it into a `.deb` file in a single step.
+
+```sh
+cargo deb -- --locked
+```
+
+> 💡 This locks you into the `--release` profile. But you can also build helix in any way you like.
+> As long as you leave a `target/release/hx` file, it will get packaged with `cargo deb --no-build`
+
+> 💡 Don't worry about the following:
+> ```
+> warning: Failed to find dependency specification
+> ```
+> Cargo deb just reports which packaged files it didn't derive dependencies for. But
+> so far the dependency deriving seams very good, even if some of the grammar files are skipped.
+
+You can find the resulted `.deb` in `target/debian/`. It should contain everything it needs, including the
+
+- completions for bash, fish, zsh
+- .desktop file
+- icon (though desktop environments might use their own since the name of the package is correctly `helix`)
+- launcher to the binary with the runtime

book/src/command-line.md

@@ -0,0 +1,88 @@
+# Command line
+
+- [Quoting](#quoting)
+- [Flags](#flags)
+- [Expansions](#expansions)
+- [Exceptions](#exceptions)
+
+The command line is used for executing [typable commands](./commands.md#typable-commands) like `:write` or `:quit`. Press `:` to activate the command line.
+
+Typable commands optionally accept arguments. `:write` for example accepts an optional path to write the file contents. The command line also supports a quoting syntax for arguments, flags to modify command behaviors, and _expansions_ - a way to insert values from the editor. Most commands support these features but some have custom parsing rules (see the [exceptions](#exceptions) below).
+
+## Quoting
+
+By default, command arguments are split on tabs and space characters. `:open README.md CHANGELOG.md` for example should open two files, `README.md` and `CHANGELOG.md`. Arguments that contain spaces can be surrounded in single quotes (`'`) or backticks (`` ` ``) to prevent the space from separating the argument, like `:open 'a b.txt'`.
+
+Double quotes may be used the same way, but double quotes _expand_ their inner content. `:echo "%{cursor_line}"` for example may print `1` because of the expansion for the `cursor_line` variable. `:echo '%{cursor_line}'` though prints `%{cursor_line}` literally: content within single quotes or backticks is interpreted as-is.
+
+On Unix systems the backslash character may be used to escape certain characters depending on where it is used. Within an argument which isn't surround in quotes, the backslash can be used to escape the space or tab characters: `:open a\ b.txt` is equivalent to `:open 'a b.txt'`. The backslash may also be used to escape quote characters (`'`, `` ` ``, `"`) or the percent token (`%`) when used at the beginning of an argument. `:echo \%%sh{foo}` for example prints `%sh{foo}` instead of invoking a `foo` shell command and `:echo \"quote` prints `"quote`. The backslash character is treated literally in any other situation on Unix systems and always on Windows: `:echo \n` always prints `\n`.
+
+## Flags
+
+Command flags are optional switches that can be used to alter the behavior of a command. For example the `:sort` command accepts an optional `--reverse` (or `-r` for short) flag which causes the sort command to reverse the sorting direction. Typing the `-` character shows completions for the current command's flags, if any.
+
+The `--` flag specifies the end of flags. All arguments after `--` are treated as positional arguments: `:open -- -a.txt` opens a file called `-a.txt`.
+
+## Expansions
+
+Expansions are patterns that Helix recognizes and replaces within the command line. Helix recognizes anything starting with a percent token (`%`) as an expansion, for example `%sh{echo hi!}`. Expansions are particularly useful when used in commands like `:echo` or `:noop` for executing simple scripts. For example:
+
+```toml
+[keys.normal]
+# Print the current line's git blame information to the statusline.
+space.B = ":echo %sh{git blame -L %{cursor_line},+1 %{buffer_name}}"
+```
+
+Expansions take the form `%[<kind>]<open><contents><close>`. In `%sh{echo hi!}`, for example, the kind is `sh` - the shell expansion - and the contents are "echo hi!", with `{` and `}` acting as opening and closing delimiters. The following open/close characters are recognized as expansion delimiter pairs: `(`/`)`, `[`/`]`, `{`/`}` and `<`/`>`. Plus the single characters `'`, `"` or `|` may be used instead: `%{cursor_line}` is equivalent to `%<cursor_line>`, `%[cursor_line]` or `%|cursor_line|`.
+
+To escape a percent character instead of treating it as an expansion, use two percent characters consecutively. To execute a shell command like `date -u +'%Y-%m-%d'`, double the percent characters: `:echo %sh{date -u +'%%Y-%%m-%%d'}`.
+
+When no `<kind>` is provided, Helix will expand a **variable**. For example `%{cursor_line}` can be used as in argument to insert the line number. `:echo %{cursor_line}` for instance may print `1` to the statusline.
+
+The following variables are supported:
+
+| Name | Description |
+|---   |---          |
+| `cursor_line` | The line number of the primary cursor in the currently focused document, starting at 1. |
+| `cursor_column` | The column number of the primary cursor in the currently focused document, starting at 1. This is counted as the number of grapheme clusters from the start of the line rather than bytes or codepoints. |
+| `buffer_name` | The relative path of the currently focused document. `[scratch]` is expanded instead for scratch buffers. |
+| `line_ending` | A string containing the line ending of the currently focused document. For example on Unix systems this is usually a line-feed character (`\n`) but on Windows systems this may be a carriage-return plus a line-feed (`\r\n`). The line ending kind of the currently focused document can be inspected with the `:line-ending` command. |
+| `current_working_directory` | Current working directory |
+| `workspace_directory` | Nearest ancestor directory of the current working directory that contains `.git`, `.svn`, `jj` or `.helix` |
+| `language` | A string containing the language name of the currently focused document.|
+| `selection` | A string containing the contents of the primary selection of the currently focused document. |
+| `selection_line_start` | The line number of the start of the primary selection in the currently focused document, starting at 1. |
+| `selection_line_end` | The line number of the end of the primary selection in the currently focused document, starting at 1. |
+
+Aside from editor variables, the following expansions may be used:
+
+* Unicode `%u{..}`. The contents may contain up to six hexadecimal numbers corresponding to a Unicode codepoint value. For example `:echo %u{25CF}` prints `●` to the statusline.
+* Shell `%sh{..}`. The contents are passed to the configured shell command. For example `:echo %sh{echo "20 * 5" | bc}` may print `100` on the statusline on when using a shell with `echo` and the `bc` calculator installed. Shell expansions are evaluated recursively. `%sh{echo '%{buffer_name}:%{cursor_line}'}` for example executes a command like `echo 'README.md:1'`: the variables within the `%sh{..}` expansion are evaluated before executing the shell command.
+
+As mentioned above, double quotes can be used to surround arguments containing spaces but also support expansions within the quoted content unlike singe quotes or backticks. For example `:echo "circle: %u{25CF}"` prints `circle: ●` to the statusline while `:echo 'circle: %u{25CF}'` prints `circle: %u{25CF}`.
+
+Note that expansions are only evaluated once the Enter key is pressed in command mode.
+
+## Exceptions
+
+The following commands support expansions but otherwise pass the given argument directly to the shell program without interpreting quotes:
+
+* `:insert-output`
+* `:append-output`
+* `:pipe`
+* `:pipe-to`
+* `:run-shell-command`
+
+For example executing `:sh echo "%{buffer_name}:%{cursor_column}"` would pass text like `echo "README.md:1"` as an argument to the shell program: the expansions are evaluated but not the quotes. As mentioned above, percent characters can be used in shell commands by doubling the percent character. To insert the output of a command like `date -u +'%Y-%m-%d'` use `:insert-output date -u +'%%Y-%%m-%%d'`.
+
+The `:set-option` and `:toggle-option` commands use regular parsing for the first argument - the config option name - and parse the rest depending on the config option's type. `:set-option` interprets the second argument as a string for string config options and parses everything else as JSON.
+
+`:toggle-option`'s behavior depends on the JSON type of the config option supplied as the first argument:
+
+* Booleans: only the config option name should be provided. For example `:toggle-option auto-format` will flip the `auto-format` option.
+* Strings: the rest of the command line is parsed with regular quoting rules. For example `:toggle-option indent-heuristic hybrid tree-sitter simple` cycles through "hybrid", "tree-sitter" and "simple" values on each invocation of the command.
+* Numbers, arrays and objects: the rest of the command line is parsed as a stream of JSON values. For example `:toggle-option rulers [81] [51, 73]` cycles through `[81]` and `[51, 73]`.
+
+When providing multiple values to `:toggle-option` there should be no duplicates. `:toggle-option indent-heuristic hybrid simple tree-sitter simple` for example would only toggle between "hybrid" and "tree-sitter" values.
+
+`:lsp-workspace-command` works similarly to `:toggle-option`. The first argument (if present) is parsed according to normal rules. The rest of the line is parsed as JSON values. Unlike `:toggle-option`, string arguments for a command must be quoted. For example `:lsp-workspace-command lsp.Command "foo" "bar"`.

book/src/commands.md

@@ -1,5 +1,16 @@
 # Commands
 
-Command mode can be activated by pressing `:`. The built-in commands are:
+- [Typable commands](#typable-commands)
+- [Static commands](#static-commands)
+
+## Typable commands
+
+Typable commands are used from command mode and may take arguments. Command mode can be activated by pressing `:`. The built-in typable commands are:
 
 {{#include ./generated/typable-cmd.md}}
+
+## Static Commands
+
+Static commands take no arguments and can be bound to keys. Static commands can also be executed from the command picker (`<space>?`). The built-in static commands are:
+
+{{#include ./generated/static-cmd.md}}

book/src/configuration.md

@@ -27,8 +27,8 @@ hidden = false
 
 You can use a custom configuration file by specifying it with the `-c` or
 `--config` command line argument, for example `hx -c path/to/custom-config.toml`.
-Additionally, you can reload the configuration file by sending the USR1
-signal to the Helix process on Unix operating systems, such as by using the command `pkill -USR1 hx`.
+You can reload the config file by issuing the `:config-reload` command. Alternatively, on Unix operating systems, you can reload it by sending the USR1
+signal to the Helix process, such as by using the command `pkill -USR1 hx`.
 
 Finally, you can have a `config.toml` local to a project by putting it under a `.helix` directory in your repository.
 Its settings will be merged with the configuration directory `config.toml` and the built-in configuration.

book/src/ecosystem.md

@@ -0,0 +1,3 @@
+# Ecosystem
+
+This section has information related to the wider Helix ecosystem. 

book/src/editor.md

@@ -1,11 +1,13 @@
 ## Editor
 
 - [`[editor]` Section](#editor-section)
+- [`[editor.clipboard-provider]` Section](#editorclipboard-provider-section)
 - [`[editor.statusline]` Section](#editorstatusline-section)
 - [`[editor.lsp]` Section](#editorlsp-section)
 - [`[editor.cursor-shape]` Section](#editorcursor-shape-section)
 - [`[editor.file-picker]` Section](#editorfile-picker-section)
 - [`[editor.auto-pairs]` Section](#editorauto-pairs-section)
+- [`[editor.auto-save]` Section](#editorauto-save-section)
 - [`[editor.search]` Section](#editorsearch-section)
 - [`[editor.whitespace]` Section](#editorwhitespace-section)
 - [`[editor.indent-guides]` Section](#editorindent-guides-section)
@@ -16,6 +18,8 @@
   - [`[editor.gutters.spacer]` Section](#editorguttersspacer-section)
 - [`[editor.soft-wrap]` Section](#editorsoft-wrap-section)
 - [`[editor.smart-tab]` Section](#editorsmart-tab-section)
+- [`[editor.inline-diagnostics]` Section](#editorinline-diagnostics-section)
+- [`[editor.word-completion]` Section](#editorword-completion-section)
 
 ### `[editor]` Section
 
@@ -23,14 +27,17 @@
 |--|--|---------|
 | `scrolloff` | Number of lines of padding around the edge of the screen when scrolling | `5` |
 | `mouse` | Enable mouse mode | `true` |
+| `default-yank-register` | Default register used for yank/paste | `"` |
 | `middle-click-paste` | Middle click paste support | `true` |
 | `scroll-lines` | Number of lines to scroll per scroll wheel step | `3` |
 | `shell` | Shell to use when running external commands | Unix: `["sh", "-c"]`<br/>Windows: `["cmd", "/C"]` |
 | `line-number` | Line number display: `absolute` simply shows each line's number, while `relative` shows the distance from the current line. When unfocused or in insert mode, `relative` will still show absolute line numbers | `absolute` |
 | `cursorline` | Highlight all lines with a cursor | `false` |
 | `cursorcolumn` | Highlight all columns with a cursor | `false` |
+| `continue-comments` | if helix should automatically add a line comment token if you create a new line inside a comment. | `true` |
 | `gutters` | Gutters to display: Available are `diagnostics` and `diff` and `line-numbers` and `spacer`, note that `diagnostics` also includes other features like breakpoints, 1-width padding will be inserted if gutters is non-empty | `["diagnostics", "spacer", "line-numbers", "spacer", "diff"]` |
 | `auto-completion` | Enable automatic pop up of auto-completion | `true` |
+| `path-completion` | Enable filepath completion. Show files and directories if an existing path at the cursor was recognized, either absolute or relative to the current opened document or current working directory (if the buffer is not yet saved). Defaults to true. | `true` |
 | `auto-format` | Enable automatic formatting on save | `true` |
 | `idle-timeout` | Time in milliseconds since last keypress before idle timers trigger. | `250` |
 | `completion-timeout` | Time in milliseconds after typing a word character before completions are shown, set to 5 for instant.  | `250` |
@@ -47,9 +54,39 @@
 | `workspace-lsp-roots` | Directories relative to the workspace root that are treated as LSP roots. Should only be set in `.helix/config.toml` | `[]` |
 | `default-line-ending` | The line ending to use for new documents. Can be `native`, `lf`, `crlf`, `ff`, `cr` or `nel`. `native` uses the platform's native line ending (`crlf` on Windows, otherwise `lf`). | `native` |
 | `insert-final-newline` | Whether to automatically insert a trailing line-ending on write if missing | `true` |
+| `atomic-save` | Whether to use atomic operations to write documents to disk. This prevents data loss if the editor is interrupted while writing the file, but may confuse some file watching/hot reloading programs. | `true` |
+| `trim-final-newlines` | Whether to automatically remove line-endings after the final one on write | `false` |
+| `trim-trailing-whitespace` | Whether to automatically remove whitespace preceding line endings on write | `false` |
 | `popup-border` | Draw border around `popup`, `menu`, `all`, or `none` | `none` |
 | `indent-heuristic` | How the indentation for a newly inserted line is computed: `simple` just copies the indentation level from the previous line, `tree-sitter` computes the indentation based on the syntax tree and `hybrid` combines both approaches. If the chosen heuristic is not available, a different one will be used as a fallback (the fallback order being `hybrid` -> `tree-sitter` -> `simple`). | `hybrid`
 | `jump-label-alphabet` | The characters that are used to generate two character jump labels. Characters at the start of the alphabet are used first. | `"abcdefghijklmnopqrstuvwxyz"`
+| `end-of-line-diagnostics` | Minimum severity of diagnostics to render at the end of the line. Set to `disable` to disable entirely. Refer to the setting about `inline-diagnostics` for more details | `"hint"`
+| `clipboard-provider` | Which API to use for clipboard interaction. One of `pasteboard` (MacOS), `wayland`, `x-clip`, `x-sel`, `win-32-yank`, `termux`, `tmux`, `windows`, `termcode`, `none`, or a custom command set. | Platform and environment specific. |
+| `editor-config` | Whether to read settings from [EditorConfig](https://editorconfig.org) files | `true` |
+| `rainbow-brackets` | Whether to render rainbow colors for matching brackets. Requires tree-sitter `rainbows.scm` queries for the language. | `false` |
+
+### `[editor.clipboard-provider]` Section
+
+Helix can be configured either to use a builtin clipboard configuration or to use
+a provided command.
+
+For instance, setting it to use OSC 52 termcodes, the configuration would be:
+```toml
+[editor]
+clipboard-provider = "termcode"
+```
+
+Alternatively, Helix can be configured to use arbitrary commands for clipboard integration:
+
+```toml
+[editor.clipboard-provider.custom]
+yank = { command = "cat",  args = ["test.txt"] }
+paste = { command = "tee",  args = ["test.txt"] }
+primary-yank = { command = "cat",  args = ["test-primary.txt"] } # optional
+primary-paste = { command = "tee",  args = ["test-primary.txt"] } # optional
+```
+
+For custom commands the contents of the yank/paste is communicated over stdin/stdout.
 
 ### `[editor.statusline]` Section
 
@@ -70,6 +107,8 @@ separator = "│"
 mode.normal = "NORMAL"
 mode.insert = "INSERT"
 mode.select = "SELECT"
+diagnostics = ["warning", "error"]
+workspace-diagnostics = ["warning", "error"]
 ```
 The `[editor.statusline]` key takes the following sub-keys:
 
@@ -82,6 +121,8 @@ The `[editor.statusline]` key takes the following sub-keys:
 | `mode.normal` | The text shown in the `mode` element for normal mode | `"NOR"` |
 | `mode.insert` | The text shown in the `mode` element for insert mode | `"INS"` |
 | `mode.select` | The text shown in the `mode` element for select mode | `"SEL"` |
+| `diagnostics` | A list of severities which are displayed for the current buffer | `["warning", "error"]` |
+| `workspace-diagnostics` | A list of severities which are displayed for the workspace | `["warning", "error"]` |
 
 The following statusline elements can be configured:
 
@@ -92,15 +133,17 @@ The following statusline elements can be configured:
 | `file-name` | The path/name of the opened file |
 | `file-absolute-path` | The absolute path/name of the opened file |
 | `file-base-name` | The basename of the opened file |
+| `current-working-directory` | The current working directory  |
 | `file-modification-indicator` | The indicator to show whether the file is modified (a `[+]` appears when there are unsaved changes) |
 | `file-encoding` | The encoding of the opened file if it differs from UTF-8 |
 | `file-line-ending` | The file line endings (CRLF or LF) |
+| `file-indent-style` | The file indentation style |
 | `read-only-indicator` | An indicator that shows `[readonly]` when a file cannot be written |
 | `total-line-numbers` | The total line numbers of the opened file |
 | `file-type` | The type of the opened file |
 | `diagnostics` | The number of warnings and/or errors |
 | `workspace-diagnostics` | The number of warnings and/or errors on workspace |
-| `selections` | The number of active selections |
+| `selections` | The primary selection index out of the number of active selections |
 | `primary-selection-length` | The number of characters currently in primary selection |
 | `position` | The cursor position |
 | `position-percentage` | The cursor position as a percentage of the total number of lines |
@@ -114,16 +157,19 @@ The following statusline elements can be configured:
 | Key                   | Description                                                 | Default |
 | ---                   | -----------                                                 | ------- |
 | `enable`              | Enables LSP integration. Setting to false will completely disable language servers regardless of language settings.| `true` |
-| `display-messages`    | Display LSP progress messages below statusline[^1]          | `false` |
+| `display-messages`    | Display LSP `window/showMessage` messages below statusline[^1] | `true` |
+| `display-progress-messages` | Display LSP progress messages below statusline[^1]    | `false` |
 | `auto-signature-help` | Enable automatic popup of signature help (parameter hints)  | `true`  |
 | `display-inlay-hints` | Display inlay hints[^2]                                     | `false` |
+| `inlay-hints-length-limit` | Maximum displayed length (non-zero number) of inlay hints | Unset by default  |
+| `display-color-swatches` | Show color swatches next to colors | `true` |
 | `display-signature-help-docs` | Display docs under signature help popup             | `true`  |
 | `snippets`      | Enables snippet completions. Requires a server restart (`:lsp-restart`) to take effect after `:config-reload`/`:set`. | `true`  |
 | `goto-reference-include-declaration` | Include declaration in the goto references popup. | `true`  |
 
 [^1]: By default, a progress spinner is shown in the statusline beside the file path.
 
-[^2]: You may also have to activate them in the LSP config for them to appear, not just in Helix. Inlay hints in Helix are still being improved on and may be a little bit laggy/janky under some circumstances. Please report any bugs you see so we can fix them!
+[^2]: You may also have to activate them in the language server config for them to appear, not just in Helix. Inlay hints in Helix are still being improved on and may be a little bit laggy/janky under some circumstances. Please report any bugs you see so we can fix them!
 
 ### `[editor.cursor-shape]` Section
 
@@ -393,3 +439,52 @@ S-tab = "move_parent_node_start"
 tab = "extend_parent_node_end"
 S-tab = "extend_parent_node_start"
 ```
+
+### `[editor.inline-diagnostics]` Section
+
+Options for rendering diagnostics inside the text like shown below
+
+```
+fn main() {
+  let foo = bar;
+            └─ no such value in this scope
+}
+````
+
+| Key        | Description | Default |
+|------------|-------------|---------|
+| `cursor-line` | The minimum severity that a diagnostic must have to be shown inline on the line that contains the primary cursor. Set to `disable` to not show any diagnostics inline. This option does not have any effect when in insert-mode and will only take effect 350ms after moving the cursor to a different line. | `"warning"` |
+| `other-lines` | The minimum severity that a diagnostic must have to be shown inline on a line that does not contain the cursor-line. Set to `disable` to not show any diagnostics inline. | `"disable"` |
+| `prefix-len` | How many horizontal bars `─` are rendered before the diagnostic text.  | `1` |
+| `max-wrap` | Equivalent of the `editor.soft-wrap.max-wrap` option for diagnostics.  | `20` |
+| `max-diagnostics` | Maximum number of diagnostics to render inline for a given line  | `10` |
+
+The allowed values for `cursor-line` and `other-lines` are: `error`, `warning`, `info`, `hint`.
+
+The (first) diagnostic with the highest severity that is not shown inline is rendered at the end of the line (as long as its severity is higher than the `end-of-line-diagnostics` config option):
+
+```
+fn main() {
+  let baz = 1;
+  let foo = bar; a local variable with a similar name exists: baz
+            └─ no such value in this scope
+}
+```
+
+### `[editor.word-completion]` Section
+
+Options for controlling completion of words from open buffers.
+
+| Key                  | Description                                                    | Default  |
+| ---                  | ---                                                            | ---      |
+| `enable`             | Whether word completion is enabled                             | `true`   |
+| `trigger-length`     | Number of word characters to type before triggering completion | `7`      |
+
+Example:
+
+```toml
+[editor.word-completion]
+enable = true
+# Set the trigger length lower so that words are completed more often
+trigger-length = 4
+```

book/src/generated/lang-support.md

@@ -1,228 +1,297 @@
-| Language | Syntax Highlighting | Treesitter Textobjects | Auto Indent | Default LSP |
-| --- | --- | --- | --- | --- |
-| ada | ✓ | ✓ |  | `ada_language_server`, `ada_language_server` |
-| adl | ✓ | ✓ | ✓ |  |
-| agda | ✓ |  |  |  |
-| astro | ✓ |  |  |  |
-| awk | ✓ | ✓ |  | `awk-language-server` |
-| bash | ✓ | ✓ | ✓ | `bash-language-server` |
-| bass | ✓ |  |  | `bass` |
-| beancount | ✓ |  |  |  |
-| bibtex | ✓ |  |  | `texlab` |
-| bicep | ✓ |  |  | `bicep-langserver` |
-| bitbake | ✓ |  |  | `bitbake-language-server` |
-| blade | ✓ |  |  |  |
-| blueprint | ✓ |  |  | `blueprint-compiler` |
-| c | ✓ | ✓ | ✓ | `clangd` |
-| c-sharp | ✓ | ✓ |  | `OmniSharp` |
-| cabal |  |  |  | `haskell-language-server-wrapper` |
-| cairo | ✓ | ✓ | ✓ | `cairo-language-server` |
-| capnp | ✓ |  | ✓ |  |
-| cel | ✓ |  |  |  |
-| clojure | ✓ |  |  | `clojure-lsp` |
-| cmake | ✓ | ✓ | ✓ | `cmake-language-server` |
-| comment | ✓ |  |  |  |
-| common-lisp | ✓ |  | ✓ | `cl-lsp` |
-| cpon | ✓ |  | ✓ |  |
-| cpp | ✓ | ✓ | ✓ | `clangd` |
-| crystal | ✓ | ✓ |  | `crystalline` |
-| css | ✓ |  | ✓ | `vscode-css-language-server` |
-| cue | ✓ |  |  | `cuelsp` |
-| d | ✓ | ✓ | ✓ | `serve-d` |
-| dart | ✓ | ✓ | ✓ | `dart` |
-| dbml | ✓ |  |  |  |
-| devicetree | ✓ |  |  |  |
-| dhall | ✓ | ✓ |  | `dhall-lsp-server` |
-| diff | ✓ |  |  |  |
-| docker-compose | ✓ |  | ✓ | `docker-compose-langserver`, `yaml-language-server` |
-| dockerfile | ✓ |  |  | `docker-langserver` |
-| dot | ✓ |  |  | `dot-language-server` |
-| dtd | ✓ |  |  |  |
-| earthfile | ✓ | ✓ | ✓ | `earthlyls` |
-| edoc | ✓ |  |  |  |
-| eex | ✓ |  |  |  |
-| ejs | ✓ |  |  |  |
-| elisp | ✓ |  |  |  |
-| elixir | ✓ | ✓ | ✓ | `elixir-ls` |
-| elm | ✓ | ✓ |  | `elm-language-server` |
-| elvish | ✓ |  |  | `elvish` |
-| env | ✓ |  |  |  |
-| erb | ✓ |  |  |  |
-| erlang | ✓ | ✓ |  | `erlang_ls` |
-| esdl | ✓ |  |  |  |
-| fidl | ✓ |  |  |  |
-| fish | ✓ | ✓ | ✓ |  |
-| forth | ✓ |  |  | `forth-lsp` |
-| fortran | ✓ |  | ✓ | `fortls` |
-| fsharp | ✓ |  |  | `fsautocomplete` |
-| gas | ✓ | ✓ |  |  |
-| gdscript | ✓ | ✓ | ✓ |  |
-| gemini | ✓ |  |  |  |
-| git-attributes | ✓ |  |  |  |
-| git-commit | ✓ | ✓ |  |  |
-| git-config | ✓ |  |  |  |
-| git-ignore | ✓ |  |  |  |
-| git-rebase | ✓ |  |  |  |
-| gjs | ✓ | ✓ | ✓ | `typescript-language-server`, `vscode-eslint-language-server`, `ember-language-server` |
-| gleam | ✓ | ✓ |  | `gleam` |
-| glimmer | ✓ |  |  | `ember-language-server` |
-| glsl | ✓ | ✓ | ✓ |  |
-| gn | ✓ |  |  |  |
-| go | ✓ | ✓ | ✓ | `gopls`, `golangci-lint-langserver` |
-| godot-resource | ✓ |  |  |  |
-| gomod | ✓ |  |  | `gopls` |
-| gotmpl | ✓ |  |  | `gopls` |
-| gowork | ✓ |  |  | `gopls` |
-| graphql | ✓ | ✓ |  | `graphql-lsp` |
-| groovy | ✓ |  |  |  |
-| gts | ✓ | ✓ | ✓ | `typescript-language-server`, `vscode-eslint-language-server`, `ember-language-server` |
-| hare | ✓ |  |  |  |
-| haskell | ✓ | ✓ |  | `haskell-language-server-wrapper` |
-| haskell-persistent | ✓ |  |  |  |
-| hcl | ✓ | ✓ | ✓ | `terraform-ls` |
-| heex | ✓ | ✓ |  | `elixir-ls` |
-| helm | ✓ |  |  | `helm_ls` |
-| hocon | ✓ |  | ✓ |  |
-| hoon | ✓ |  |  |  |
-| hosts | ✓ |  |  |  |
-| html | ✓ |  |  | `vscode-html-language-server` |
-| hurl | ✓ | ✓ | ✓ |  |
-| hyprlang | ✓ |  | ✓ |  |
-| idris |  |  |  | `idris2-lsp` |
-| iex | ✓ |  |  |  |
-| ini | ✓ |  |  |  |
-| inko | ✓ | ✓ | ✓ |  |
-| janet | ✓ |  |  |  |
-| java | ✓ | ✓ | ✓ | `jdtls` |
-| javascript | ✓ | ✓ | ✓ | `typescript-language-server` |
-| jinja | ✓ |  |  |  |
-| jsdoc | ✓ |  |  |  |
-| json | ✓ | ✓ | ✓ | `vscode-json-language-server` |
-| json5 | ✓ |  |  |  |
-| jsonc | ✓ |  | ✓ | `vscode-json-language-server` |
-| jsonnet | ✓ |  |  | `jsonnet-language-server` |
-| jsx | ✓ | ✓ | ✓ | `typescript-language-server` |
-| julia | ✓ | ✓ | ✓ | `julia` |
-| just | ✓ | ✓ | ✓ |  |
-| kdl | ✓ | ✓ | ✓ |  |
-| koka | ✓ |  | ✓ | `koka` |
-| kotlin | ✓ |  |  | `kotlin-language-server` |
-| latex | ✓ | ✓ |  | `texlab` |
-| ld | ✓ |  | ✓ |  |
-| ldif | ✓ |  |  |  |
-| lean | ✓ |  |  | `lean` |
-| ledger | ✓ |  |  |  |
-| llvm | ✓ | ✓ | ✓ |  |
-| llvm-mir | ✓ | ✓ | ✓ |  |
-| llvm-mir-yaml | ✓ |  | ✓ |  |
-| log | ✓ |  |  |  |
-| lpf | ✓ |  |  |  |
-| lua | ✓ | ✓ | ✓ | `lua-language-server` |
-| make | ✓ |  | ✓ |  |
-| markdoc | ✓ |  |  | `markdoc-ls` |
-| markdown | ✓ |  |  | `marksman`, `markdown-oxide` |
-| markdown.inline | ✓ |  |  |  |
-| matlab | ✓ | ✓ | ✓ |  |
-| mermaid | ✓ |  |  |  |
-| meson | ✓ |  | ✓ |  |
-| mint |  |  |  | `mint` |
-| mojo | ✓ | ✓ | ✓ | `mojo-lsp-server` |
-| move | ✓ |  |  |  |
-| msbuild | ✓ |  | ✓ |  |
-| nasm | ✓ | ✓ |  |  |
-| nickel | ✓ |  | ✓ | `nls` |
-| nim | ✓ | ✓ | ✓ | `nimlangserver` |
-| nix | ✓ | ✓ |  | `nil` |
-| nu | ✓ |  |  | `nu` |
-| nunjucks | ✓ |  |  |  |
-| ocaml | ✓ |  | ✓ | `ocamllsp` |
-| ocaml-interface | ✓ |  |  | `ocamllsp` |
-| odin | ✓ |  | ✓ | `ols` |
-| ohm | ✓ | ✓ | ✓ |  |
-| opencl | ✓ | ✓ | ✓ | `clangd` |
-| openscad | ✓ |  |  | `openscad-lsp` |
-| org | ✓ |  |  |  |
-| pascal | ✓ | ✓ |  | `pasls` |
-| passwd | ✓ |  |  |  |
-| pem | ✓ |  |  |  |
-| perl | ✓ | ✓ | ✓ | `perlnavigator` |
-| pest | ✓ | ✓ | ✓ | `pest-language-server` |
-| php | ✓ | ✓ | ✓ | `intelephense` |
-| php-only | ✓ |  |  |  |
-| pkgbuild | ✓ | ✓ | ✓ | `pkgbuild-language-server`, `bash-language-server` |
-| pkl | ✓ |  | ✓ |  |
-| po | ✓ | ✓ |  |  |
-| pod | ✓ |  |  |  |
-| ponylang | ✓ | ✓ | ✓ |  |
-| powershell | ✓ |  |  |  |
-| prisma | ✓ |  |  | `prisma-language-server` |
-| prolog |  |  |  | `swipl` |
-| protobuf | ✓ | ✓ | ✓ | `bufls`, `pb` |
-| prql | ✓ |  |  |  |
-| purescript | ✓ | ✓ |  | `purescript-language-server` |
-| python | ✓ | ✓ | ✓ | `pylsp` |
-| qml | ✓ |  | ✓ | `qmlls` |
-| r | ✓ |  |  | `R` |
-| racket | ✓ |  | ✓ | `racket` |
-| regex | ✓ |  |  |  |
-| rego | ✓ |  |  | `regols` |
-| rescript | ✓ | ✓ |  | `rescript-language-server` |
-| rmarkdown | ✓ |  | ✓ | `R` |
-| robot | ✓ |  |  | `robotframework_ls` |
-| ron | ✓ |  | ✓ |  |
-| rst | ✓ |  |  |  |
-| ruby | ✓ | ✓ | ✓ | `solargraph` |
-| rust | ✓ | ✓ | ✓ | `rust-analyzer` |
-| sage | ✓ | ✓ |  |  |
-| scala | ✓ | ✓ | ✓ | `metals` |
-| scheme | ✓ |  | ✓ |  |
-| scss | ✓ |  |  | `vscode-css-language-server` |
-| slint | ✓ | ✓ | ✓ | `slint-lsp` |
-| smali | ✓ |  | ✓ |  |
-| smithy | ✓ |  |  | `cs` |
-| sml | ✓ |  |  |  |
-| solidity | ✓ | ✓ |  | `solc` |
-| spicedb | ✓ |  |  |  |
-| sql | ✓ |  |  |  |
-| sshclientconfig | ✓ |  |  |  |
-| starlark | ✓ | ✓ |  |  |
-| strace | ✓ |  |  |  |
-| supercollider | ✓ |  |  |  |
-| svelte | ✓ |  | ✓ | `svelteserver` |
-| sway | ✓ | ✓ | ✓ | `forc` |
-| swift | ✓ | ✓ |  | `sourcekit-lsp` |
-| t32 | ✓ |  |  |  |
-| tablegen | ✓ | ✓ | ✓ |  |
-| tact | ✓ | ✓ | ✓ |  |
-| task | ✓ |  |  |  |
-| tcl | ✓ |  | ✓ |  |
-| templ | ✓ |  |  | `templ` |
-| tfvars | ✓ |  | ✓ | `terraform-ls` |
-| todotxt | ✓ |  |  |  |
-| toml | ✓ | ✓ |  | `taplo` |
-| tsq | ✓ |  |  |  |
-| tsx | ✓ | ✓ | ✓ | `typescript-language-server` |
-| twig | ✓ |  |  |  |
-| typescript | ✓ | ✓ | ✓ | `typescript-language-server` |
-| typst | ✓ |  |  | `tinymist`, `typst-lsp` |
-| ungrammar | ✓ |  |  |  |
-| unison | ✓ |  | ✓ |  |
-| uxntal | ✓ |  |  |  |
-| v | ✓ | ✓ | ✓ | `v-analyzer` |
-| vala | ✓ | ✓ |  | `vala-language-server` |
-| verilog | ✓ | ✓ |  | `svlangserver` |
-| vhdl | ✓ |  |  | `vhdl_ls` |
-| vhs | ✓ |  |  |  |
-| vue | ✓ |  |  | `vue-language-server` |
-| wast | ✓ |  |  |  |
-| wat | ✓ |  |  |  |
-| webc | ✓ |  |  |  |
-| wgsl | ✓ |  |  | `wgsl_analyzer` |
-| wit | ✓ |  | ✓ |  |
-| wren | ✓ | ✓ | ✓ |  |
-| xit | ✓ |  |  |  |
-| xml | ✓ |  | ✓ |  |
-| xtc | ✓ |  |  |  |
-| yaml | ✓ |  | ✓ | `yaml-language-server`, `ansible-language-server` |
-| yuck | ✓ |  |  |  |
-| zig | ✓ | ✓ | ✓ | `zls` |
+| Language | Syntax Highlighting | Treesitter Textobjects | Auto Indent | Code Navigation Tags | Rainbow Brackets | Default language servers |
+| --- | --- | --- | --- | --- | --- | --- |
+| ada | ✓ | ✓ |  |  |  | `ada_language_server` |
+| adl | ✓ | ✓ | ✓ |  |  |  |
+| agda | ✓ |  |  |  |  |  |
+| alloy | ✓ |  |  |  |  |  |
+| amber | ✓ |  |  |  |  | `amber-lsp` |
+| astro | ✓ |  |  |  |  | `astro-ls` |
+| awk | ✓ | ✓ |  |  |  | `awk-language-server` |
+| bash | ✓ | ✓ | ✓ | ✓ | ✓ | `bash-language-server` |
+| bass | ✓ |  |  |  |  | `bass` |
+| beancount | ✓ |  |  |  |  | `beancount-language-server` |
+| bibtex | ✓ |  |  |  |  | `texlab` |
+| bicep | ✓ |  |  |  |  | `bicep-langserver` |
+| bitbake | ✓ |  |  |  |  | `bitbake-language-server` |
+| blade | ✓ | ✓ |  |  | ✓ |  |
+| blueprint | ✓ |  |  |  |  | `blueprint-compiler` |
+| c | ✓ | ✓ | ✓ | ✓ | ✓ | `clangd` |
+| c-sharp | ✓ | ✓ |  | ✓ |  | `OmniSharp` |
+| cabal |  |  |  |  |  | `haskell-language-server-wrapper` |
+| caddyfile | ✓ | ✓ | ✓ |  |  |  |
+| cairo | ✓ | ✓ | ✓ |  |  | `cairo-language-server` |
+| capnp | ✓ |  | ✓ |  |  |  |
+| cel | ✓ |  |  |  |  |  |
+| circom | ✓ |  |  |  |  | `circom-lsp` |
+| clarity | ✓ |  |  |  |  | `clarinet` |
+| clojure | ✓ |  |  |  | ✓ | `clojure-lsp` |
+| cmake | ✓ | ✓ | ✓ |  |  | `neocmakelsp`, `cmake-language-server` |
+| codeql | ✓ | ✓ |  |  |  | `codeql` |
+| comment | ✓ |  |  |  |  |  |
+| common-lisp | ✓ |  | ✓ |  | ✓ | `cl-lsp` |
+| cpon | ✓ |  | ✓ |  |  |  |
+| cpp | ✓ | ✓ | ✓ | ✓ | ✓ | `clangd` |
+| cross-config | ✓ | ✓ |  |  | ✓ | `taplo`, `tombi` |
+| crystal | ✓ | ✓ | ✓ | ✓ |  | `crystalline`, `ameba-ls` |
+| css | ✓ |  | ✓ |  | ✓ | `vscode-css-language-server` |
+| csv | ✓ |  |  |  |  |  |
+| cue | ✓ |  |  |  |  | `cuelsp` |
+| cylc | ✓ | ✓ | ✓ |  |  |  |
+| cython | ✓ |  | ✓ | ✓ |  |  |
+| d | ✓ | ✓ | ✓ |  |  | `serve-d` |
+| dart | ✓ | ✓ | ✓ |  |  | `dart` |
+| dbml | ✓ |  |  |  |  |  |
+| debian | ✓ |  |  |  |  |  |
+| devicetree | ✓ |  |  |  |  | `dts-lsp` |
+| dhall | ✓ | ✓ |  |  |  | `dhall-lsp-server` |
+| diff | ✓ |  |  |  |  |  |
+| djot | ✓ |  |  |  |  |  |
+| docker-compose | ✓ | ✓ | ✓ |  |  | `docker-compose-langserver`, `yaml-language-server` |
+| dockerfile | ✓ | ✓ |  |  |  | `docker-langserver` |
+| dot | ✓ |  |  |  |  | `dot-language-server` |
+| doxyfile | ✓ | ✓ | ✓ | ✓ |  |  |
+| dtd | ✓ |  |  |  |  |  |
+| dune | ✓ |  |  |  |  |  |
+| dunstrc | ✓ |  |  |  |  |  |
+| earthfile | ✓ | ✓ | ✓ |  |  | `earthlyls` |
+| edoc | ✓ |  |  |  |  |  |
+| eex | ✓ |  |  |  |  |  |
+| ejs | ✓ |  |  |  |  |  |
+| elisp | ✓ |  |  | ✓ |  |  |
+| elixir | ✓ | ✓ | ✓ | ✓ | ✓ | `elixir-ls` |
+| elm | ✓ | ✓ |  | ✓ |  | `elm-language-server` |
+| elvish | ✓ |  |  |  |  | `elvish` |
+| env | ✓ | ✓ |  |  |  |  |
+| erb | ✓ |  |  |  |  |  |
+| erlang | ✓ | ✓ |  | ✓ | ✓ | `erlang_ls`, `elp` |
+| esdl | ✓ |  |  |  |  |  |
+| fennel | ✓ |  |  |  |  | `fennel-ls` |
+| fga | ✓ | ✓ | ✓ |  |  |  |
+| fidl | ✓ |  |  |  |  |  |
+| fish | ✓ | ✓ | ✓ |  |  | `fish-lsp` |
+| flatbuffers | ✓ |  |  |  |  |  |
+| forth | ✓ |  |  |  |  | `forth-lsp` |
+| fortran | ✓ |  | ✓ |  |  | `fortls` |
+| fsharp | ✓ |  |  |  |  | `fsautocomplete` |
+| gas | ✓ | ✓ |  |  |  | `asm-lsp` |
+| gdscript | ✓ | ✓ | ✓ | ✓ |  |  |
+| gemini | ✓ |  |  |  |  |  |
+| gherkin | ✓ |  |  |  |  |  |
+| ghostty | ✓ |  |  |  |  |  |
+| git-attributes | ✓ |  |  |  |  |  |
+| git-cliff-config | ✓ | ✓ |  |  | ✓ | `taplo`, `tombi` |
+| git-commit | ✓ | ✓ |  |  |  |  |
+| git-config | ✓ | ✓ |  |  |  |  |
+| git-ignore | ✓ |  |  |  |  |  |
+| git-notes | ✓ |  |  |  |  |  |
+| git-rebase | ✓ |  |  |  |  |  |
+| gjs | ✓ | ✓ | ✓ | ✓ |  | `typescript-language-server`, `vscode-eslint-language-server`, `ember-language-server` |
+| gleam | ✓ | ✓ |  |  | ✓ | `gleam` |
+| glimmer | ✓ |  |  |  |  | `ember-language-server` |
+| glsl | ✓ | ✓ | ✓ |  |  | `glsl_analyzer` |
+| gn | ✓ |  |  |  |  |  |
+| go | ✓ | ✓ | ✓ | ✓ | ✓ | `gopls`, `golangci-lint-langserver` |
+| go-format-string | ✓ |  |  |  | ✓ |  |
+| godot-resource | ✓ | ✓ |  |  |  |  |
+| gomod | ✓ |  |  |  |  | `gopls` |
+| gotmpl | ✓ |  |  |  |  | `gopls` |
+| gowork | ✓ |  |  |  |  | `gopls` |
+| gpr | ✓ |  |  |  |  | `ada_language_server` |
+| graphql | ✓ | ✓ |  |  |  | `graphql-lsp` |
+| gren | ✓ | ✓ |  |  |  |  |
+| groovy | ✓ |  |  |  |  |  |
+| gts | ✓ | ✓ | ✓ | ✓ |  | `typescript-language-server`, `vscode-eslint-language-server`, `ember-language-server` |
+| hare | ✓ |  |  |  |  |  |
+| haskell | ✓ | ✓ |  |  |  | `haskell-language-server-wrapper` |
+| haskell-persistent | ✓ |  |  |  |  |  |
+| hcl | ✓ | ✓ | ✓ |  |  | `terraform-ls` |
+| hdl | ✓ |  |  |  |  |  |
+| heex | ✓ | ✓ |  |  |  | `elixir-ls` |
+| helm | ✓ |  |  |  |  | `helm_ls` |
+| hocon | ✓ | ✓ | ✓ |  |  |  |
+| hoon | ✓ |  |  |  |  |  |
+| hosts | ✓ |  |  |  |  |  |
+| html | ✓ | ✓ |  |  | ✓ | `vscode-html-language-server`, `superhtml` |
+| htmldjango | ✓ |  |  |  |  | `djlsp`, `vscode-html-language-server`, `superhtml` |
+| hurl | ✓ | ✓ | ✓ |  |  |  |
+| hyprlang | ✓ |  | ✓ |  |  | `hyprls` |
+| idris |  |  |  |  |  | `idris2-lsp` |
+| iex | ✓ |  |  |  |  |  |
+| ini | ✓ |  |  |  |  |  |
+| ink | ✓ |  |  |  |  |  |
+| inko | ✓ | ✓ | ✓ | ✓ |  |  |
+| janet | ✓ |  | ✓ |  | ✓ |  |
+| java | ✓ | ✓ | ✓ | ✓ | ✓ | `jdtls` |
+| javascript | ✓ | ✓ | ✓ | ✓ | ✓ | `typescript-language-server` |
+| jinja | ✓ |  |  |  |  |  |
+| jjconfig | ✓ | ✓ | ✓ |  |  | `taplo`, `tombi` |
+| jjdescription | ✓ |  |  |  |  |  |
+| jjrevset | ✓ |  |  |  |  |  |
+| jjtemplate | ✓ |  |  |  |  |  |
+| jq | ✓ | ✓ |  |  |  | `jq-lsp` |
+| jsdoc | ✓ |  |  |  |  |  |
+| json | ✓ | ✓ | ✓ |  | ✓ | `vscode-json-language-server` |
+| json-ld | ✓ | ✓ | ✓ |  |  | `vscode-json-language-server` |
+| json5 | ✓ |  |  |  |  |  |
+| jsonc | ✓ |  | ✓ |  |  | `vscode-json-language-server` |
+| jsonnet | ✓ |  |  |  |  | `jsonnet-language-server` |
+| jsx | ✓ | ✓ | ✓ | ✓ | ✓ | `typescript-language-server` |
+| julia | ✓ | ✓ | ✓ |  |  | `julia` |
+| just | ✓ | ✓ | ✓ | ✓ |  | `just-lsp` |
+| kconfig | ✓ |  | ✓ |  |  |  |
+| kdl | ✓ | ✓ | ✓ |  |  |  |
+| koka | ✓ |  | ✓ |  |  | `koka` |
+| kotlin | ✓ | ✓ | ✓ | ✓ |  | `kotlin-language-server` |
+| koto | ✓ | ✓ | ✓ |  |  | `koto-ls` |
+| latex | ✓ | ✓ |  |  |  | `texlab` |
+| ld | ✓ |  | ✓ |  |  |  |
+| ldif | ✓ |  |  |  |  |  |
+| lean | ✓ |  |  |  |  | `lean` |
+| ledger | ✓ |  |  |  |  |  |
+| llvm | ✓ | ✓ | ✓ |  |  |  |
+| llvm-mir | ✓ | ✓ | ✓ |  |  |  |
+| llvm-mir-yaml | ✓ |  | ✓ |  |  |  |
+| log | ✓ |  |  |  |  |  |
+| lpf | ✓ |  |  |  |  |  |
+| lua | ✓ | ✓ | ✓ |  | ✓ | `lua-language-server` |
+| luap | ✓ |  |  |  |  |  |
+| luau | ✓ | ✓ | ✓ |  |  | `luau-lsp` |
+| mail | ✓ | ✓ |  |  |  |  |
+| make | ✓ |  | ✓ |  |  |  |
+| markdoc | ✓ |  |  |  |  | `markdoc-ls` |
+| markdown | ✓ |  |  | ✓ |  | `marksman`, `markdown-oxide` |
+| markdown-rustdoc | ✓ |  |  |  |  |  |
+| markdown.inline | ✓ |  |  |  |  |  |
+| matlab | ✓ | ✓ | ✓ |  |  |  |
+| mermaid | ✓ |  |  |  |  |  |
+| meson | ✓ |  | ✓ |  |  | `mesonlsp` |
+| mint |  |  |  |  |  | `mint` |
+| mojo | ✓ | ✓ | ✓ |  |  | `pixi` |
+| move | ✓ |  |  |  |  |  |
+| msbuild | ✓ |  | ✓ |  |  |  |
+| nasm | ✓ | ✓ |  |  |  | `asm-lsp` |
+| nestedtext | ✓ | ✓ | ✓ |  |  |  |
+| nginx | ✓ |  |  |  |  |  |
+| nickel | ✓ |  | ✓ |  |  | `nls` |
+| nim | ✓ | ✓ | ✓ |  |  | `nimlangserver` |
+| nix | ✓ | ✓ | ✓ |  | ✓ | `nil`, `nixd` |
+| nu | ✓ |  |  |  |  | `nu` |
+| nunjucks | ✓ |  |  |  |  |  |
+| ocaml | ✓ |  | ✓ |  |  | `ocamllsp` |
+| ocaml-interface | ✓ |  |  |  |  | `ocamllsp` |
+| odin | ✓ | ✓ | ✓ |  |  | `ols` |
+| ohm | ✓ | ✓ | ✓ |  |  |  |
+| opencl | ✓ | ✓ | ✓ |  |  | `clangd` |
+| openscad | ✓ |  |  |  |  | `openscad-lsp` |
+| org | ✓ |  |  |  |  |  |
+| pascal | ✓ | ✓ |  |  |  | `pasls` |
+| passwd | ✓ |  |  |  |  |  |
+| pem | ✓ |  |  |  |  |  |
+| perl | ✓ | ✓ | ✓ |  |  | `perlnavigator` |
+| pest | ✓ | ✓ | ✓ |  |  | `pest-language-server` |
+| php | ✓ | ✓ | ✓ | ✓ | ✓ | `intelephense` |
+| php-only | ✓ |  |  | ✓ |  |  |
+| pip-requirements | ✓ |  |  |  |  |  |
+| pkgbuild | ✓ | ✓ | ✓ |  |  | `termux-language-server`, `bash-language-server` |
+| pkl | ✓ |  | ✓ |  |  | `pkl-lsp` |
+| po | ✓ | ✓ |  |  |  |  |
+| pod | ✓ |  |  |  |  |  |
+| ponylang | ✓ | ✓ | ✓ |  |  |  |
+| powershell | ✓ |  |  |  |  |  |
+| prisma | ✓ | ✓ |  |  |  | `prisma-language-server` |
+| prolog | ✓ |  | ✓ |  |  | `swipl` |
+| properties | ✓ | ✓ |  |  |  |  |
+| protobuf | ✓ | ✓ | ✓ | ✓ |  | `buf`, `pb`, `protols` |
+| prql | ✓ |  |  |  |  |  |
+| pug | ✓ |  |  |  |  |  |
+| purescript | ✓ | ✓ |  |  |  | `purescript-language-server` |
+| python | ✓ | ✓ | ✓ | ✓ | ✓ | `ty`, `ruff`, `jedi-language-server`, `pylsp` |
+| qml | ✓ | ✓ | ✓ |  |  | `qmlls` |
+| quarto | ✓ |  | ✓ |  |  |  |
+| quint | ✓ |  |  |  |  | `quint-language-server` |
+| r | ✓ |  |  |  |  | `R` |
+| racket | ✓ |  | ✓ |  | ✓ | `racket` |
+| regex | ✓ |  |  |  | ✓ |  |
+| rego | ✓ |  |  |  |  | `regols` |
+| rescript | ✓ | ✓ |  |  |  | `rescript-language-server` |
+| rmarkdown | ✓ |  | ✓ |  |  | `R` |
+| robot | ✓ |  |  |  |  | `robotframework_ls` |
+| robots.txt | ✓ | ✓ |  | ✓ |  |  |
+| ron | ✓ |  | ✓ | ✓ | ✓ |  |
+| rst | ✓ |  |  |  |  |  |
+| ruby | ✓ | ✓ | ✓ | ✓ | ✓ | `ruby-lsp`, `solargraph` |
+| rust | ✓ | ✓ | ✓ | ✓ | ✓ | `rust-analyzer` |
+| rust-format-args | ✓ |  |  |  |  |  |
+| rust-format-args-macro | ✓ | ✓ | ✓ |  | ✓ |  |
+| sage | ✓ | ✓ |  |  |  |  |
+| scala | ✓ | ✓ | ✓ |  |  | `metals` |
+| scheme | ✓ |  | ✓ |  | ✓ |  |
+| scss | ✓ |  |  |  | ✓ | `vscode-css-language-server` |
+| shellcheckrc | ✓ | ✓ |  |  |  |  |
+| slang | ✓ | ✓ | ✓ |  |  | `slangd` |
+| slint | ✓ | ✓ | ✓ |  |  | `slint-lsp` |
+| smali | ✓ |  | ✓ |  |  |  |
+| smithy | ✓ |  |  |  |  | `cs` |
+| sml | ✓ |  |  |  |  |  |
+| snakemake | ✓ |  | ✓ |  |  | `pylsp` |
+| solidity | ✓ | ✓ |  |  |  | `solc` |
+| sourcepawn | ✓ | ✓ |  |  |  | `sourcepawn-studio` |
+| spade | ✓ |  | ✓ |  |  | `spade-language-server` |
+| spicedb | ✓ |  |  | ✓ |  |  |
+| sql | ✓ | ✓ |  |  |  |  |
+| sshclientconfig | ✓ |  |  |  |  |  |
+| starlark | ✓ | ✓ | ✓ |  | ✓ | `starpls` |
+| strace | ✓ |  |  |  |  |  |
+| strictdoc | ✓ |  |  | ✓ |  |  |
+| supercollider | ✓ |  |  |  |  |  |
+| svelte | ✓ |  | ✓ |  |  | `svelteserver` |
+| sway | ✓ | ✓ | ✓ |  |  | `forc` |
+| swift | ✓ | ✓ |  |  | ✓ | `sourcekit-lsp` |
+| systemd | ✓ |  |  |  |  | `systemd-lsp` |
+| systemverilog | ✓ |  |  |  |  |  |
+| t32 | ✓ |  |  |  |  |  |
+| tablegen | ✓ | ✓ | ✓ |  |  |  |
+| tact | ✓ | ✓ | ✓ |  |  |  |
+| task | ✓ |  |  |  |  |  |
+| tcl | ✓ |  | ✓ |  |  |  |
+| teal | ✓ |  |  |  |  | `teal-language-server` |
+| templ | ✓ |  |  |  |  | `templ` |
+| tera | ✓ |  |  |  |  |  |
+| textproto | ✓ | ✓ | ✓ |  |  |  |
+| tfvars | ✓ |  | ✓ |  |  | `terraform-ls` |
+| thrift | ✓ |  |  |  |  |  |
+| tlaplus | ✓ |  |  |  |  |  |
+| todotxt | ✓ |  |  |  |  |  |
+| toml | ✓ | ✓ |  |  | ✓ | `taplo`, `tombi` |
+| tsq | ✓ |  |  |  | ✓ | `ts_query_ls` |
+| tsx | ✓ | ✓ | ✓ | ✓ | ✓ | `typescript-language-server` |
+| twig | ✓ |  |  |  |  |  |
+| typescript | ✓ | ✓ | ✓ | ✓ | ✓ | `typescript-language-server` |
+| typespec | ✓ | ✓ | ✓ |  |  | `tsp-server` |
+| typst | ✓ |  |  | ✓ |  | `tinymist` |
+| ungrammar | ✓ |  |  |  |  |  |
+| unison | ✓ | ✓ | ✓ |  |  |  |
+| uxntal | ✓ |  |  |  |  |  |
+| v | ✓ | ✓ | ✓ |  |  | `v-analyzer` |
+| vala | ✓ | ✓ |  |  |  | `vala-language-server` |
+| vento | ✓ |  |  |  |  |  |
+| verilog | ✓ | ✓ |  |  |  | `svlangserver` |
+| vhdl | ✓ |  |  |  |  | `vhdl_ls` |
+| vhs | ✓ |  |  |  |  |  |
+| vim | ✓ |  |  |  |  |  |
+| vue | ✓ |  |  |  |  | `vue-language-server` |
+| wast | ✓ |  |  |  |  |  |
+| wat | ✓ |  |  |  |  | `wat_server` |
+| webc | ✓ |  |  |  |  |  |
+| werk | ✓ |  |  |  |  |  |
+| wesl | ✓ | ✓ |  |  |  |  |
+| wgsl | ✓ | ✓ | ✓ | ✓ | ✓ | `wgsl-analyzer` |
+| wit | ✓ |  | ✓ |  |  |  |
+| wren | ✓ | ✓ | ✓ |  |  |  |
+| xit | ✓ |  |  |  |  |  |
+| xml | ✓ | ✓ | ✓ |  | ✓ |  |
+| xtc | ✓ |  |  |  |  |  |
+| yaml | ✓ | ✓ | ✓ |  | ✓ | `yaml-language-server`, `ansible-language-server` |
+| yara | ✓ |  |  |  |  | `yls` |
+| yuck | ✓ |  |  |  |  |  |
+| zig | ✓ | ✓ | ✓ |  |  | `zls` |

book/src/generated/static-cmd.md

@@ -0,0 +1,315 @@
+| Name | Description | Default keybinds |
+| --- | --- | --- |
+| `no_op` | Do nothing |  |
+| `move_char_left` | Move left | normal: `` h ``, `` <left> ``, insert: `` <left> `` |
+| `move_char_right` | Move right | normal: `` l ``, `` <right> ``, insert: `` <right> `` |
+| `move_line_up` | Move up | normal: `` gk `` |
+| `move_line_down` | Move down | normal: `` gj `` |
+| `move_visual_line_up` | Move up | normal: `` k ``, `` <up> ``, insert: `` <up> `` |
+| `move_visual_line_down` | Move down | normal: `` j ``, `` <down> ``, insert: `` <down> `` |
+| `extend_char_left` | Extend left | select: `` h ``, `` <left> `` |
+| `extend_char_right` | Extend right | select: `` l ``, `` <right> `` |
+| `extend_line_up` | Extend up | select: `` gk `` |
+| `extend_line_down` | Extend down | select: `` gj `` |
+| `extend_visual_line_up` | Extend up | select: `` k ``, `` <up> `` |
+| `extend_visual_line_down` | Extend down | select: `` j ``, `` <down> `` |
+| `copy_selection_on_next_line` | Copy selection on next line | normal: `` C ``, select: `` C `` |
+| `copy_selection_on_prev_line` | Copy selection on previous line | normal: `` <A-C> ``, select: `` <A-C> `` |
+| `move_next_word_start` | Move to start of next word | normal: `` w `` |
+| `move_prev_word_start` | Move to start of previous word | normal: `` b `` |
+| `move_next_word_end` | Move to end of next word | normal: `` e `` |
+| `move_prev_word_end` | Move to end of previous word |  |
+| `move_next_long_word_start` | Move to start of next long word | normal: `` W `` |
+| `move_prev_long_word_start` | Move to start of previous long word | normal: `` B `` |
+| `move_next_long_word_end` | Move to end of next long word | normal: `` E `` |
+| `move_prev_long_word_end` | Move to end of previous long word |  |
+| `move_next_sub_word_start` | Move to start of next sub word |  |
+| `move_prev_sub_word_start` | Move to start of previous sub word |  |
+| `move_next_sub_word_end` | Move to end of next sub word |  |
+| `move_prev_sub_word_end` | Move to end of previous sub word |  |
+| `move_parent_node_end` | Move to end of the parent node | normal: `` <A-e> `` |
+| `move_parent_node_start` | Move to beginning of the parent node | normal: `` <A-b> `` |
+| `extend_next_word_start` | Extend to start of next word | select: `` w `` |
+| `extend_prev_word_start` | Extend to start of previous word | select: `` b `` |
+| `extend_next_word_end` | Extend to end of next word | select: `` e `` |
+| `extend_prev_word_end` | Extend to end of previous word |  |
+| `extend_next_long_word_start` | Extend to start of next long word | select: `` W `` |
+| `extend_prev_long_word_start` | Extend to start of previous long word | select: `` B `` |
+| `extend_next_long_word_end` | Extend to end of next long word | select: `` E `` |
+| `extend_prev_long_word_end` | Extend to end of prev long word |  |
+| `extend_next_sub_word_start` | Extend to start of next sub word |  |
+| `extend_prev_sub_word_start` | Extend to start of previous sub word |  |
+| `extend_next_sub_word_end` | Extend to end of next sub word |  |
+| `extend_prev_sub_word_end` | Extend to end of prev sub word |  |
+| `extend_parent_node_end` | Extend to end of the parent node | select: `` <A-e> `` |
+| `extend_parent_node_start` | Extend to beginning of the parent node | select: `` <A-b> `` |
+| `find_till_char` | Move till next occurrence of char | normal: `` t `` |
+| `find_next_char` | Move to next occurrence of char | normal: `` f `` |
+| `extend_till_char` | Extend till next occurrence of char | select: `` t `` |
+| `extend_next_char` | Extend to next occurrence of char | select: `` f `` |
+| `till_prev_char` | Move till previous occurrence of char | normal: `` T `` |
+| `find_prev_char` | Move to previous occurrence of char | normal: `` F `` |
+| `extend_till_prev_char` | Extend till previous occurrence of char | select: `` T `` |
+| `extend_prev_char` | Extend to previous occurrence of char | select: `` F `` |
+| `repeat_last_motion` | Repeat last motion | normal: `` <A-.> ``, select: `` <A-.> `` |
+| `replace` | Replace with new char | normal: `` r ``, select: `` r `` |
+| `switch_case` | Switch (toggle) case | normal: `` ~ ``, select: `` ~ `` |
+| `switch_to_uppercase` | Switch to uppercase | normal: `` <A-`> ``, select: `` <A-`> `` |
+| `switch_to_lowercase` | Switch to lowercase | normal: `` ` ``, select: `` ` `` |
+| `page_up` | Move page up | normal: `` <C-b> ``, `` Z<C-b> ``, `` z<C-b> ``, `` <pageup> ``, `` Z<pageup> ``, `` z<pageup> ``, select: `` <C-b> ``, `` Z<C-b> ``, `` z<C-b> ``, `` <pageup> ``, `` Z<pageup> ``, `` z<pageup> ``, insert: `` <pageup> `` |
+| `page_down` | Move page down | normal: `` <C-f> ``, `` Z<C-f> ``, `` z<C-f> ``, `` <pagedown> ``, `` Z<pagedown> ``, `` z<pagedown> ``, select: `` <C-f> ``, `` Z<C-f> ``, `` z<C-f> ``, `` <pagedown> ``, `` Z<pagedown> ``, `` z<pagedown> ``, insert: `` <pagedown> `` |
+| `half_page_up` | Move half page up |  |
+| `half_page_down` | Move half page down |  |
+| `page_cursor_up` | Move page and cursor up |  |
+| `page_cursor_down` | Move page and cursor down |  |
+| `page_cursor_half_up` | Move page and cursor half up | normal: `` <C-u> ``, `` Z<C-u> ``, `` z<C-u> ``, `` Z<backspace> ``, `` z<backspace> ``, select: `` <C-u> ``, `` Z<C-u> ``, `` z<C-u> ``, `` Z<backspace> ``, `` z<backspace> `` |
+| `page_cursor_half_down` | Move page and cursor half down | normal: `` <C-d> ``, `` Z<C-d> ``, `` z<C-d> ``, `` Z<space> ``, `` z<space> ``, select: `` <C-d> ``, `` Z<C-d> ``, `` z<C-d> ``, `` Z<space> ``, `` z<space> `` |
+| `select_all` | Select whole document | normal: `` % ``, select: `` % `` |
+| `select_regex` | Select all regex matches inside selections | normal: `` s ``, select: `` s `` |
+| `split_selection` | Split selections on regex matches | normal: `` S ``, select: `` S `` |
+| `split_selection_on_newline` | Split selection on newlines | normal: `` <A-s> ``, select: `` <A-s> `` |
+| `merge_selections` | Merge selections | normal: `` <A-minus> ``, select: `` <A-minus> `` |
+| `merge_consecutive_selections` | Merge consecutive selections | normal: `` <A-_> ``, select: `` <A-_> `` |
+| `search` | Search for regex pattern | normal: `` / ``, `` Z/ ``, `` z/ ``, select: `` / ``, `` Z/ ``, `` z/ `` |
+| `rsearch` | Reverse search for regex pattern | normal: `` ? ``, `` Z? ``, `` z? ``, select: `` ? ``, `` Z? ``, `` z? `` |
+| `search_next` | Select next search match | normal: `` n ``, `` Zn ``, `` zn ``, select: `` Zn ``, `` zn `` |
+| `search_prev` | Select previous search match | normal: `` N ``, `` ZN ``, `` zN ``, select: `` ZN ``, `` zN `` |
+| `extend_search_next` | Add next search match to selection | select: `` n `` |
+| `extend_search_prev` | Add previous search match to selection | select: `` N `` |
+| `search_selection` | Use current selection as search pattern | normal: `` <A-*> ``, select: `` <A-*> `` |
+| `search_selection_detect_word_boundaries` | Use current selection as the search pattern, automatically wrapping with `\b` on word boundaries | normal: `` * ``, select: `` * `` |
+| `make_search_word_bounded` | Modify current search to make it word bounded |  |
+| `global_search` | Global search in workspace folder | normal: `` <space>/ ``, select: `` <space>/ `` |
+| `extend_line` | Select current line, if already selected, extend to another line based on the anchor |  |
+| `extend_line_below` | Select current line, if already selected, extend to next line | normal: `` x ``, select: `` x `` |
+| `extend_line_above` | Select current line, if already selected, extend to previous line |  |
+| `select_line_above` | Select current line, if already selected, extend or shrink line above based on the anchor |  |
+| `select_line_below` | Select current line, if already selected, extend or shrink line below based on the anchor |  |
+| `extend_to_line_bounds` | Extend selection to line bounds | normal: `` X ``, select: `` X `` |
+| `shrink_to_line_bounds` | Shrink selection to line bounds | normal: `` <A-x> ``, select: `` <A-x> `` |
+| `delete_selection` | Delete selection | normal: `` d ``, select: `` d `` |
+| `delete_selection_noyank` | Delete selection without yanking | normal: `` <A-d> ``, select: `` <A-d> `` |
+| `change_selection` | Change selection | normal: `` c ``, select: `` c `` |
+| `change_selection_noyank` | Change selection without yanking | normal: `` <A-c> ``, select: `` <A-c> `` |
+| `collapse_selection` | Collapse selection into single cursor | normal: `` ; ``, select: `` ; `` |
+| `flip_selections` | Flip selection cursor and anchor | normal: `` <A-;> ``, select: `` <A-;> `` |
+| `ensure_selections_forward` | Ensure all selections face forward | normal: `` <A-:> ``, select: `` <A-:> `` |
+| `insert_mode` | Insert before selection | normal: `` i ``, select: `` i `` |
+| `append_mode` | Append after selection | normal: `` a ``, select: `` a `` |
+| `command_mode` | Enter command mode | normal: `` : ``, select: `` : `` |
+| `file_picker` | Open file picker | normal: `` <space>f ``, select: `` <space>f `` |
+| `file_picker_in_current_buffer_directory` | Open file picker at current buffer's directory |  |
+| `file_picker_in_current_directory` | Open file picker at current working directory | normal: `` <space>F ``, select: `` <space>F `` |
+| `file_explorer` | Open file explorer in workspace root | normal: `` <space>e ``, select: `` <space>e `` |
+| `file_explorer_in_current_buffer_directory` | Open file explorer at current buffer's directory | normal: `` <space>E ``, select: `` <space>E `` |
+| `file_explorer_in_current_directory` | Open file explorer at current working directory |  |
+| `code_action` | Perform code action | normal: `` <space>a ``, select: `` <space>a `` |
+| `buffer_picker` | Open buffer picker | normal: `` <space>b ``, select: `` <space>b `` |
+| `jumplist_picker` | Open jumplist picker | normal: `` <space>j ``, select: `` <space>j `` |
+| `symbol_picker` | Open symbol picker |  |
+| `syntax_symbol_picker` | Open symbol picker from syntax information |  |
+| `lsp_or_syntax_symbol_picker` | Open symbol picker from LSP or syntax information | normal: `` <space>s ``, select: `` <space>s `` |
+| `changed_file_picker` | Open changed file picker | normal: `` <space>g ``, select: `` <space>g `` |
+| `select_references_to_symbol_under_cursor` | Select symbol references | normal: `` <space>h ``, select: `` <space>h `` |
+| `workspace_symbol_picker` | Open workspace symbol picker |  |
+| `syntax_workspace_symbol_picker` | Open workspace symbol picker from syntax information |  |
+| `lsp_or_syntax_workspace_symbol_picker` | Open workspace symbol picker from LSP or syntax information | normal: `` <space>S ``, select: `` <space>S `` |
+| `diagnostics_picker` | Open diagnostic picker | normal: `` <space>d ``, select: `` <space>d `` |
+| `workspace_diagnostics_picker` | Open workspace diagnostic picker | normal: `` <space>D ``, select: `` <space>D `` |
+| `last_picker` | Open last picker | normal: `` <space>' ``, select: `` <space>' `` |
+| `insert_at_line_start` | Insert at start of line | normal: `` I ``, select: `` I `` |
+| `insert_at_line_end` | Insert at end of line | normal: `` A ``, select: `` A `` |
+| `open_below` | Open new line below selection | normal: `` o ``, select: `` o `` |
+| `open_above` | Open new line above selection | normal: `` O ``, select: `` O `` |
+| `normal_mode` | Enter normal mode | normal: `` <esc> ``, select: `` v ``, insert: `` <esc> `` |
+| `select_mode` | Enter selection extend mode | normal: `` v `` |
+| `exit_select_mode` | Exit selection mode | select: `` <esc> `` |
+| `goto_definition` | Goto definition | normal: `` gd ``, select: `` gd `` |
+| `goto_declaration` | Goto declaration | normal: `` gD ``, select: `` gD `` |
+| `add_newline_above` | Add newline above | normal: `` [<space> ``, select: `` [<space> `` |
+| `add_newline_below` | Add newline below | normal: `` ]<space> ``, select: `` ]<space> `` |
+| `goto_type_definition` | Goto type definition | normal: `` gy ``, select: `` gy `` |
+| `goto_implementation` | Goto implementation | normal: `` gi ``, select: `` gi `` |
+| `goto_file_start` | Goto line number <n> else file start | normal: `` gg `` |
+| `goto_file_end` | Goto file end |  |
+| `extend_to_file_start` | Extend to line number<n> else file start | select: `` gg `` |
+| `extend_to_file_end` | Extend to file end |  |
+| `goto_file` | Goto files/URLs in selections | normal: `` gf ``, select: `` gf `` |
+| `goto_file_hsplit` | Goto files in selections (hsplit) | normal: `` <C-w>f ``, `` <space>wf ``, select: `` <C-w>f ``, `` <space>wf `` |
+| `goto_file_vsplit` | Goto files in selections (vsplit) | normal: `` <C-w>F ``, `` <space>wF ``, select: `` <C-w>F ``, `` <space>wF `` |
+| `goto_reference` | Goto references | normal: `` gr ``, select: `` gr `` |
+| `goto_window_top` | Goto window top | normal: `` gt ``, select: `` gt `` |
+| `goto_window_center` | Goto window center | normal: `` gc ``, select: `` gc `` |
+| `goto_window_bottom` | Goto window bottom | normal: `` gb ``, select: `` gb `` |
+| `goto_last_accessed_file` | Goto last accessed file | normal: `` ga ``, select: `` ga `` |
+| `goto_last_modified_file` | Goto last modified file | normal: `` gm ``, select: `` gm `` |
+| `goto_last_modification` | Goto last modification | normal: `` g. ``, select: `` g. `` |
+| `goto_line` | Goto line | normal: `` G ``, select: `` G `` |
+| `goto_last_line` | Goto last line | normal: `` ge `` |
+| `extend_to_last_line` | Extend to last line | select: `` ge `` |
+| `goto_first_diag` | Goto first diagnostic | normal: `` [D ``, select: `` [D `` |
+| `goto_last_diag` | Goto last diagnostic | normal: `` ]D ``, select: `` ]D `` |
+| `goto_next_diag` | Goto next diagnostic | normal: `` ]d ``, select: `` ]d `` |
+| `goto_prev_diag` | Goto previous diagnostic | normal: `` [d ``, select: `` [d `` |
+| `goto_next_change` | Goto next change | normal: `` ]g ``, select: `` ]g `` |
+| `goto_prev_change` | Goto previous change | normal: `` [g ``, select: `` [g `` |
+| `goto_first_change` | Goto first change | normal: `` [G ``, select: `` [G `` |
+| `goto_last_change` | Goto last change | normal: `` ]G ``, select: `` ]G `` |
+| `goto_line_start` | Goto line start | normal: `` gh ``, `` <home> ``, select: `` gh ``, insert: `` <home> `` |
+| `goto_line_end` | Goto line end | normal: `` gl ``, `` <end> ``, select: `` gl `` |
+| `goto_column` | Goto column | normal: `` g\| `` |
+| `extend_to_column` | Extend to column | select: `` g\| `` |
+| `goto_next_buffer` | Goto next buffer | normal: `` gn ``, select: `` gn `` |
+| `goto_previous_buffer` | Goto previous buffer | normal: `` gp ``, select: `` gp `` |
+| `goto_line_end_newline` | Goto newline at line end | insert: `` <end> `` |
+| `goto_first_nonwhitespace` | Goto first non-blank in line | normal: `` gs ``, select: `` gs `` |
+| `trim_selections` | Trim whitespace from selections | normal: `` _ ``, select: `` _ `` |
+| `extend_to_line_start` | Extend to line start | select: `` <home> `` |
+| `extend_to_first_nonwhitespace` | Extend to first non-blank in line |  |
+| `extend_to_line_end` | Extend to line end | select: `` <end> `` |
+| `extend_to_line_end_newline` | Extend to line end |  |
+| `signature_help` | Show signature help |  |
+| `smart_tab` | Insert tab if all cursors have all whitespace to their left; otherwise, run a separate command. | insert: `` <tab> `` |
+| `insert_tab` | Insert tab char | insert: `` <S-tab> `` |
+| `insert_newline` | Insert newline char | insert: `` <C-j> ``, `` <ret> `` |
+| `insert_char_interactive` | Insert an interactively-chosen char |  |
+| `append_char_interactive` | Append an interactively-chosen char |  |
+| `delete_char_backward` | Delete previous char | insert: `` <C-h> ``, `` <backspace> ``, `` <S-backspace> `` |
+| `delete_char_forward` | Delete next char | insert: `` <C-d> ``, `` <del> `` |
+| `delete_word_backward` | Delete previous word | insert: `` <C-w> ``, `` <A-backspace> `` |
+| `delete_word_forward` | Delete next word | insert: `` <A-d> ``, `` <A-del> `` |
+| `kill_to_line_start` | Delete till start of line | insert: `` <C-u> `` |
+| `kill_to_line_end` | Delete till end of line | insert: `` <C-k> `` |
+| `undo` | Undo change | normal: `` u ``, select: `` u `` |
+| `redo` | Redo change | normal: `` U ``, select: `` U `` |
+| `earlier` | Move backward in history | normal: `` <A-u> ``, select: `` <A-u> `` |
+| `later` | Move forward in history | normal: `` <A-U> ``, select: `` <A-U> `` |
+| `commit_undo_checkpoint` | Commit changes to new checkpoint | insert: `` <C-s> `` |
+| `yank` | Yank selection | normal: `` y ``, select: `` y `` |
+| `yank_to_clipboard` | Yank selections to clipboard | normal: `` <space>y ``, select: `` <space>y `` |
+| `yank_to_primary_clipboard` | Yank selections to primary clipboard |  |
+| `yank_joined` | Join and yank selections |  |
+| `yank_joined_to_clipboard` | Join and yank selections to clipboard |  |
+| `yank_main_selection_to_clipboard` | Yank main selection to clipboard | normal: `` <space>Y ``, select: `` <space>Y `` |
+| `yank_joined_to_primary_clipboard` | Join and yank selections to primary clipboard |  |
+| `yank_main_selection_to_primary_clipboard` | Yank main selection to primary clipboard |  |
+| `replace_with_yanked` | Replace with yanked text | normal: `` R ``, select: `` R `` |
+| `replace_selections_with_clipboard` | Replace selections by clipboard content | normal: `` <space>R ``, select: `` <space>R `` |
+| `replace_selections_with_primary_clipboard` | Replace selections by primary clipboard |  |
+| `paste_after` | Paste after selection | normal: `` p ``, select: `` p `` |
+| `paste_before` | Paste before selection | normal: `` P ``, select: `` P `` |
+| `paste_clipboard_after` | Paste clipboard after selections | normal: `` <space>p ``, select: `` <space>p `` |
+| `paste_clipboard_before` | Paste clipboard before selections | normal: `` <space>P ``, select: `` <space>P `` |
+| `paste_primary_clipboard_after` | Paste primary clipboard after selections |  |
+| `paste_primary_clipboard_before` | Paste primary clipboard before selections |  |
+| `indent` | Indent selection | normal: `` <gt> ``, select: `` <gt> `` |
+| `unindent` | Unindent selection | normal: `` <lt> ``, select: `` <lt> `` |
+| `format_selections` | Format selection | normal: `` = ``, select: `` = `` |
+| `join_selections` | Join lines inside selection | normal: `` J ``, select: `` J `` |
+| `join_selections_space` | Join lines inside selection and select spaces | normal: `` <A-J> ``, select: `` <A-J> `` |
+| `keep_selections` | Keep selections matching regex | normal: `` K ``, select: `` K `` |
+| `remove_selections` | Remove selections matching regex | normal: `` <A-K> ``, select: `` <A-K> `` |
+| `align_selections` | Align selections in column | normal: `` & ``, select: `` & `` |
+| `keep_primary_selection` | Keep primary selection | normal: `` , ``, select: `` , `` |
+| `remove_primary_selection` | Remove primary selection | normal: `` <A-,> ``, select: `` <A-,> `` |
+| `completion` | Invoke completion popup | insert: `` <C-x> `` |
+| `hover` | Show docs for item under cursor | normal: `` <space>k ``, select: `` <space>k `` |
+| `toggle_comments` | Comment/uncomment selections | normal: `` <C-c> ``, `` <space>c ``, select: `` <C-c> ``, `` <space>c `` |
+| `toggle_line_comments` | Line comment/uncomment selections | normal: `` <space><A-c> ``, select: `` <space><A-c> `` |
+| `toggle_block_comments` | Block comment/uncomment selections | normal: `` <space>C ``, select: `` <space>C `` |
+| `rotate_selections_forward` | Rotate selections forward | normal: `` ) ``, select: `` ) `` |
+| `rotate_selections_backward` | Rotate selections backward | normal: `` ( ``, select: `` ( `` |
+| `rotate_selection_contents_forward` | Rotate selection contents forward | normal: `` <A-)> ``, select: `` <A-)> `` |
+| `rotate_selection_contents_backward` | Rotate selections contents backward | normal: `` <A-(> ``, select: `` <A-(> `` |
+| `reverse_selection_contents` | Reverse selections contents |  |
+| `expand_selection` | Expand selection to parent syntax node | normal: `` <A-o> ``, `` <A-up> ``, select: `` <A-o> ``, `` <A-up> `` |
+| `shrink_selection` | Shrink selection to previously expanded syntax node | normal: `` <A-i> ``, `` <A-down> ``, select: `` <A-i> ``, `` <A-down> `` |
+| `select_next_sibling` | Select next sibling in the syntax tree | normal: `` <A-n> ``, `` <A-right> ``, select: `` <A-n> ``, `` <A-right> `` |
+| `select_prev_sibling` | Select previous sibling the in syntax tree | normal: `` <A-p> ``, `` <A-left> ``, select: `` <A-p> ``, `` <A-left> `` |
+| `select_all_siblings` | Select all siblings of the current node | normal: `` <A-a> ``, select: `` <A-a> `` |
+| `select_all_children` | Select all children of the current node | normal: `` <A-I> ``, `` <S-A-down> ``, select: `` <A-I> ``, `` <S-A-down> `` |
+| `jump_forward` | Jump forward on jumplist | normal: `` <C-i> ``, `` <tab> ``, select: `` <C-i> ``, `` <tab> `` |
+| `jump_backward` | Jump backward on jumplist | normal: `` <C-o> ``, select: `` <C-o> `` |
+| `save_selection` | Save current selection to jumplist | normal: `` <C-s> ``, select: `` <C-s> `` |
+| `jump_view_right` | Jump to right split | normal: `` <C-w>l ``, `` <space>wl ``, `` <C-w><C-l> ``, `` <C-w><right> ``, `` <space>w<C-l> ``, `` <space>w<right> ``, select: `` <C-w>l ``, `` <space>wl ``, `` <C-w><C-l> ``, `` <C-w><right> ``, `` <space>w<C-l> ``, `` <space>w<right> `` |
+| `jump_view_left` | Jump to left split | normal: `` <C-w>h ``, `` <space>wh ``, `` <C-w><C-h> ``, `` <C-w><left> ``, `` <space>w<C-h> ``, `` <space>w<left> ``, select: `` <C-w>h ``, `` <space>wh ``, `` <C-w><C-h> ``, `` <C-w><left> ``, `` <space>w<C-h> ``, `` <space>w<left> `` |
+| `jump_view_up` | Jump to split above | normal: `` <C-w>k ``, `` <C-w><up> ``, `` <space>wk ``, `` <C-w><C-k> ``, `` <space>w<up> ``, `` <space>w<C-k> ``, select: `` <C-w>k ``, `` <C-w><up> ``, `` <space>wk ``, `` <C-w><C-k> ``, `` <space>w<up> ``, `` <space>w<C-k> `` |
+| `jump_view_down` | Jump to split below | normal: `` <C-w>j ``, `` <space>wj ``, `` <C-w><C-j> ``, `` <C-w><down> ``, `` <space>w<C-j> ``, `` <space>w<down> ``, select: `` <C-w>j ``, `` <space>wj ``, `` <C-w><C-j> ``, `` <C-w><down> ``, `` <space>w<C-j> ``, `` <space>w<down> `` |
+| `swap_view_right` | Swap with right split | normal: `` <C-w>L ``, `` <space>wL ``, select: `` <C-w>L ``, `` <space>wL `` |
+| `swap_view_left` | Swap with left split | normal: `` <C-w>H ``, `` <space>wH ``, select: `` <C-w>H ``, `` <space>wH `` |
+| `swap_view_up` | Swap with split above | normal: `` <C-w>K ``, `` <space>wK ``, select: `` <C-w>K ``, `` <space>wK `` |
+| `swap_view_down` | Swap with split below | normal: `` <C-w>J ``, `` <space>wJ ``, select: `` <C-w>J ``, `` <space>wJ `` |
+| `transpose_view` | Transpose splits | normal: `` <C-w>t ``, `` <space>wt ``, `` <C-w><C-t> ``, `` <space>w<C-t> ``, select: `` <C-w>t ``, `` <space>wt ``, `` <C-w><C-t> ``, `` <space>w<C-t> `` |
+| `rotate_view` | Goto next window | normal: `` <C-w>w ``, `` <space>ww ``, `` <C-w><C-w> ``, `` <space>w<C-w> ``, select: `` <C-w>w ``, `` <space>ww ``, `` <C-w><C-w> ``, `` <space>w<C-w> `` |
+| `rotate_view_reverse` | Goto previous window |  |
+| `hsplit` | Horizontal bottom split | normal: `` <C-w>s ``, `` <space>ws ``, `` <C-w><C-s> ``, `` <space>w<C-s> ``, select: `` <C-w>s ``, `` <space>ws ``, `` <C-w><C-s> ``, `` <space>w<C-s> `` |
+| `hsplit_new` | Horizontal bottom split scratch buffer | normal: `` <C-w>ns ``, `` <space>wns ``, `` <C-w>n<C-s> ``, `` <space>wn<C-s> ``, select: `` <C-w>ns ``, `` <space>wns ``, `` <C-w>n<C-s> ``, `` <space>wn<C-s> `` |
+| `vsplit` | Vertical right split | normal: `` <C-w>v ``, `` <space>wv ``, `` <C-w><C-v> ``, `` <space>w<C-v> ``, select: `` <C-w>v ``, `` <space>wv ``, `` <C-w><C-v> ``, `` <space>w<C-v> `` |
+| `vsplit_new` | Vertical right split scratch buffer | normal: `` <C-w>nv ``, `` <space>wnv ``, `` <C-w>n<C-v> ``, `` <space>wn<C-v> ``, select: `` <C-w>nv ``, `` <space>wnv ``, `` <C-w>n<C-v> ``, `` <space>wn<C-v> `` |
+| `wclose` | Close window | normal: `` <C-w>q ``, `` <space>wq ``, `` <C-w><C-q> ``, `` <space>w<C-q> ``, select: `` <C-w>q ``, `` <space>wq ``, `` <C-w><C-q> ``, `` <space>w<C-q> `` |
+| `wonly` | Close windows except current | normal: `` <C-w>o ``, `` <space>wo ``, `` <C-w><C-o> ``, `` <space>w<C-o> ``, select: `` <C-w>o ``, `` <space>wo ``, `` <C-w><C-o> ``, `` <space>w<C-o> `` |
+| `select_register` | Select register | normal: `` " ``, select: `` " `` |
+| `insert_register` | Insert register | insert: `` <C-r> `` |
+| `copy_between_registers` | Copy between two registers |  |
+| `align_view_middle` | Align view middle | normal: `` Zm ``, `` zm ``, select: `` Zm ``, `` zm `` |
+| `align_view_top` | Align view top | normal: `` Zt ``, `` zt ``, select: `` Zt ``, `` zt `` |
+| `align_view_center` | Align view center | normal: `` Zc ``, `` Zz ``, `` zc ``, `` zz ``, select: `` Zc ``, `` Zz ``, `` zc ``, `` zz `` |
+| `align_view_bottom` | Align view bottom | normal: `` Zb ``, `` zb ``, select: `` Zb ``, `` zb `` |
+| `scroll_up` | Scroll view up | normal: `` Zk ``, `` zk ``, `` Z<up> ``, `` z<up> ``, select: `` Zk ``, `` zk ``, `` Z<up> ``, `` z<up> `` |
+| `scroll_down` | Scroll view down | normal: `` Zj ``, `` zj ``, `` Z<down> ``, `` z<down> ``, select: `` Zj ``, `` zj ``, `` Z<down> ``, `` z<down> `` |
+| `match_brackets` | Goto matching bracket | normal: `` mm ``, select: `` mm `` |
+| `surround_add` | Surround add | normal: `` ms ``, select: `` ms `` |
+| `surround_replace` | Surround replace | normal: `` mr ``, select: `` mr `` |
+| `surround_delete` | Surround delete | normal: `` md ``, select: `` md `` |
+| `select_textobject_around` | Select around object | normal: `` ma ``, select: `` ma `` |
+| `select_textobject_inner` | Select inside object | normal: `` mi ``, select: `` mi `` |
+| `goto_next_function` | Goto next function | normal: `` ]f ``, select: `` ]f `` |
+| `goto_prev_function` | Goto previous function | normal: `` [f ``, select: `` [f `` |
+| `goto_next_class` | Goto next type definition | normal: `` ]t ``, select: `` ]t `` |
+| `goto_prev_class` | Goto previous type definition | normal: `` [t ``, select: `` [t `` |
+| `goto_next_parameter` | Goto next parameter | normal: `` ]a ``, select: `` ]a `` |
+| `goto_prev_parameter` | Goto previous parameter | normal: `` [a ``, select: `` [a `` |
+| `goto_next_comment` | Goto next comment | normal: `` ]c ``, select: `` ]c `` |
+| `goto_prev_comment` | Goto previous comment | normal: `` [c ``, select: `` [c `` |
+| `goto_next_test` | Goto next test | normal: `` ]T ``, select: `` ]T `` |
+| `goto_prev_test` | Goto previous test | normal: `` [T ``, select: `` [T `` |
+| `goto_next_xml_element` | Goto next (X)HTML element | normal: `` ]x ``, select: `` ]x `` |
+| `goto_prev_xml_element` | Goto previous (X)HTML element | normal: `` [x ``, select: `` [x `` |
+| `goto_next_entry` | Goto next pairing | normal: `` ]e ``, select: `` ]e `` |
+| `goto_prev_entry` | Goto previous pairing | normal: `` [e ``, select: `` [e `` |
+| `goto_next_paragraph` | Goto next paragraph | normal: `` ]p ``, select: `` ]p `` |
+| `goto_prev_paragraph` | Goto previous paragraph | normal: `` [p ``, select: `` [p `` |
+| `dap_launch` | Launch debug target | normal: `` <space>Gl ``, select: `` <space>Gl `` |
+| `dap_restart` | Restart debugging session | normal: `` <space>Gr ``, select: `` <space>Gr `` |
+| `dap_toggle_breakpoint` | Toggle breakpoint | normal: `` <space>Gb ``, select: `` <space>Gb `` |
+| `dap_continue` | Continue program execution | normal: `` <space>Gc ``, select: `` <space>Gc `` |
+| `dap_pause` | Pause program execution | normal: `` <space>Gh ``, select: `` <space>Gh `` |
+| `dap_step_in` | Step in | normal: `` <space>Gi ``, select: `` <space>Gi `` |
+| `dap_step_out` | Step out | normal: `` <space>Go ``, select: `` <space>Go `` |
+| `dap_next` | Step to next | normal: `` <space>Gn ``, select: `` <space>Gn `` |
+| `dap_variables` | List variables | normal: `` <space>Gv ``, select: `` <space>Gv `` |
+| `dap_terminate` | End debug session | normal: `` <space>Gt ``, select: `` <space>Gt `` |
+| `dap_edit_condition` | Edit breakpoint condition on current line | normal: `` <space>G<C-c> ``, select: `` <space>G<C-c> `` |
+| `dap_edit_log` | Edit breakpoint log message on current line | normal: `` <space>G<C-l> ``, select: `` <space>G<C-l> `` |
+| `dap_switch_thread` | Switch current thread | normal: `` <space>Gst ``, select: `` <space>Gst `` |
+| `dap_switch_stack_frame` | Switch stack frame | normal: `` <space>Gsf ``, select: `` <space>Gsf `` |
+| `dap_enable_exceptions` | Enable exception breakpoints | normal: `` <space>Ge ``, select: `` <space>Ge `` |
+| `dap_disable_exceptions` | Disable exception breakpoints | normal: `` <space>GE ``, select: `` <space>GE `` |
+| `shell_pipe` | Pipe selections through shell command | normal: `` \| ``, select: `` \| `` |
+| `shell_pipe_to` | Pipe selections into shell command ignoring output | normal: `` <A-\|> ``, select: `` <A-\|> `` |
+| `shell_insert_output` | Insert shell command output before selections | normal: `` ! ``, select: `` ! `` |
+| `shell_append_output` | Append shell command output after selections | normal: `` <A-!> ``, select: `` <A-!> `` |
+| `shell_keep_pipe` | Filter selections with shell predicate | normal: `` $ ``, select: `` $ `` |
+| `suspend` | Suspend and return to shell | normal: `` <C-z> ``, select: `` <C-z> `` |
+| `rename_symbol` | Rename symbol | normal: `` <space>r ``, select: `` <space>r `` |
+| `increment` | Increment item under cursor | normal: `` <C-a> ``, select: `` <C-a> `` |
+| `decrement` | Decrement item under cursor | normal: `` <C-x> ``, select: `` <C-x> `` |
+| `record_macro` | Record macro | normal: `` Q ``, select: `` Q `` |
+| `replay_macro` | Replay macro | normal: `` q ``, select: `` q `` |
+| `command_palette` | Open command palette | normal: `` <space>? ``, select: `` <space>? `` |
+| `goto_word` | Jump to a two-character label | normal: `` gw `` |
+| `extend_to_word` | Extend to a two-character label | select: `` gw `` |
+| `goto_next_tabstop` | Goto next snippet placeholder |  |
+| `goto_prev_tabstop` | Goto next snippet placeholder |  |
+| `rotate_selections_first` | Make the first selection your primary one |  |
+| `rotate_selections_last` | Make the last selection your primary one |  |

book/src/generated/typable-cmd.md

@@ -2,7 +2,7 @@
 | --- | --- |
 | `:quit`, `:q` | Close the current view. |
 | `:quit!`, `:q!` | Force close the current view, ignoring unsaved changes. |
-| `:open`, `:o` | Open a file from disk into the current view. |
+| `:open`, `:o`, `:edit`, `:e` | Open a file from disk into the current view. |
 | `:buffer-close`, `:bc`, `:bclose` | Close the current buffer. |
 | `:buffer-close!`, `:bc!`, `:bclose!` | Close the current buffer forcefully, ignoring unsaved changes. |
 | `:buffer-close-others`, `:bco`, `:bcloseother` | Close all buffers but the currently focused one. |
@@ -16,7 +16,7 @@
 | `:write-buffer-close`, `:wbc` | Write changes to disk and closes the buffer. Accepts an optional path (:write-buffer-close some/path.txt) |
 | `:write-buffer-close!`, `:wbc!` | Force write changes to disk creating necessary subdirectories and closes the buffer. Accepts an optional path (:write-buffer-close! some/path.txt) |
 | `:new`, `:n` | Create a new scratch buffer. |
-| `:format`, `:fmt` | Format the file using the LSP formatter. |
+| `:format`, `:fmt` | Format the file using an external formatter or language server. |
 | `:indent-style` | Set the indentation style for editing. ('t' for tabs or 1-16 for number of spaces.) |
 | `:line-ending` | Set the document's default line ending. Options: crlf, lf. |
 | `:earlier`, `:ear` | Jump back to an earlier point in edit history. Accepts a number of steps or a time span. |
@@ -52,8 +52,8 @@
 | `:reload-all`, `:rla` | Discard changes and reload all documents from the source files. |
 | `:update`, `:u` | Write changes only if the file has been modified. |
 | `:lsp-workspace-command` | Open workspace command picker |
-| `:lsp-restart` | Restarts the language servers used by the current doc |
-| `:lsp-stop` | Stops the language servers that are used by the current doc |
+| `:lsp-restart` | Restarts the given language servers, or all language servers that are used by the current file if no arguments are supplied |
+| `:lsp-stop` | Stops the given language servers, or all language servers that are used by the current file if no arguments are supplied |
 | `:tree-sitter-scopes` | Display tree sitter scopes, primarily for theming and development. |
 | `:tree-sitter-highlight-name` | Display name of tree-sitter highlight scope under the cursor. |
 | `:debug-start`, `:dbg` | Start a debug session from a given template with given parameters. |
@@ -67,24 +67,25 @@
 | `:goto`, `:g` | Goto line number. |
 | `:set-language`, `:lang` | Set the language of current buffer (show current language if no value specified). |
 | `:set-option`, `:set` | Set a config option at runtime.<br>For example to disable smart case search, use `:set search.smart-case false`. |
-| `:toggle-option`, `:toggle` | Toggle a boolean config option at runtime.<br>For example to toggle smart case search, use `:toggle search.smart-case`. |
+| `:toggle-option`, `:toggle` | Toggle a config option at runtime.<br>For example to toggle smart case search, use `:toggle search.smart-case`. |
 | `:get-option`, `:get` | Get the current value of a config option. |
 | `:sort` | Sort ranges in selection. |
-| `:rsort` | Sort ranges in selection in reverse order. |
 | `:reflow` | Hard-wrap the current selection of lines to a given width. |
-| `:tree-sitter-subtree`, `:ts-subtree` | Display tree sitter subtree under cursor, primarily for debugging queries. |
+| `:tree-sitter-subtree`, `:ts-subtree` | Display the smallest tree-sitter subtree that spans the primary selection, primarily for debugging queries. |
 | `:config-reload` | Refresh user config. |
 | `:config-open` | Open the user config.toml file. |
 | `:config-open-workspace` | Open the workspace config.toml file. |
 | `:log-open` | Open the helix log file. |
 | `:insert-output` | Run shell command, inserting output before each selection. |
 | `:append-output` | Run shell command, appending output after each selection. |
-| `:pipe` | Pipe each selection to the shell command. |
+| `:pipe`, `:\|` | Pipe each selection to the shell command. |
 | `:pipe-to` | Pipe each selection to the shell command, ignoring output. |
-| `:run-shell-command`, `:sh` | Run a shell command |
+| `:run-shell-command`, `:sh`, `:!` | Run a shell command |
 | `:reset-diff-change`, `:diffget`, `:diffg` | Reset the diff change at the cursor position. |
 | `:clear-register` | Clear given register. If no argument is provided, clear all registers. |
 | `:redraw` | Clear and re-render the whole UI |
-| `:move` | Move the current buffer and its corresponding file to a different path |
+| `:move`, `:mv` | Move the current buffer and its corresponding file to a different path |
 | `:yank-diagnostic` | Yank diagnostic(s) under primary cursor to register, or clipboard by default |
 | `:read`, `:r` | Load a file into buffer |
+| `:echo` | Prints the given arguments to the statusline. |
+| `:noop` | Does nothing. |

book/src/guides/README.md

@@ -1,4 +1,4 @@
 # Guides
 
 This section contains guides for adding new language server configurations,
-tree-sitter grammars, textobject queries, and other similar items.
+tree-sitter grammars, textobject and rainbow bracket queries, and other similar items.

book/src/guides/adding_languages.md

@@ -1,4 +1,4 @@
-# Adding new languages to Helix
+## Adding new languages to Helix
 
 In order to add a new language to Helix, you will need to follow the steps
 below.
@@ -34,16 +34,10 @@ below.
 2. Create a new directory for the language with the path
    `runtime/queries/<name>/`.
 3. Refer to the
-   [tree-sitter website](https://tree-sitter.github.io/tree-sitter/syntax-highlighting#queries)
+   [tree-sitter website](https://tree-sitter.github.io/tree-sitter/3-syntax-highlighting.html#highlights)
    for more information on writing queries.
 4. A list of highlight captures can be found [on the themes page](https://docs.helix-editor.com/themes.html#scopes).
 
-> 💡 In Helix, the first matching query takes precedence when evaluating
-> queries, which is different from other editors such as Neovim where the last
-> matching query supersedes the ones before it. See
-> [this issue](https://github.com/helix-editor/helix/pull/1170#issuecomment-997294090)
-> for an example.
-
 ## Common issues
 
 - If you encounter errors when running Helix after switching branches, you may

book/src/guides/indent.md

@@ -1,4 +1,4 @@
-# Adding indent queries
+## Adding indent queries
 
 Helix uses tree-sitter to correctly indent new lines. This requires a tree-
 sitter grammar and an `indent.scm` query file placed in `runtime/queries/

book/src/guides/injection.md

@@ -1,14 +1,19 @@
-# Adding Injection Queries
+## Adding Injection Queries
 
 Writing language injection queries allows one to highlight a specific node as a different language.
 In addition to the [standard][upstream-docs] language injection options used by tree-sitter, there
 are a few Helix specific extensions that allow for more control.
 
-And example of a simple query that would highlight all strings as bash in Nix:
+An example of a simple query that would highlight all strings as bash in Nix:
 ```scm
 ((string_expression (string_fragment) @injection.content)
   (#set! injection.language "bash"))
 ```
+Another example is this query, which highlights links in comments and keywords like "TODO", by reusing the dedicated "comment" language:
+```
+((comment) @injection.content
+  (#set! injection.language "comment"))
+```
 
 ## Capture Types
 
@@ -57,4 +62,4 @@ second argument (a string).
 - `#any-of?` (standard):
 The first argument (a capture) must be one of the other arguments (strings).
 
-[upstream-docs]: http://tree-sitter.github.io/tree-sitter/syntax-highlighting#language-injection
+[upstream-docs]: https://tree-sitter.github.io/tree-sitter/3-syntax-highlighting.html#language-injection

book/src/guides/rainbow_bracket_queries.md

@@ -0,0 +1,132 @@
+# Adding Rainbow Bracket Queries
+
+Helix uses `rainbows.scm` tree-sitter query files to provide rainbow bracket
+functionality.
+
+Tree-sitter queries are documented in the tree-sitter online documentation.
+If you're writing queries for the first time, be sure to check out the section
+on [syntax highlighting queries] and on [query syntax].
+
+Rainbow queries have two captures: `@rainbow.scope` and `@rainbow.bracket`.
+`@rainbow.scope` should capture any node that increases the nesting level
+while `@rainbow.bracket` should capture any bracket nodes. Put another way:
+`@rainbow.scope` switches to the next rainbow color for all nodes in the tree
+under it while `@rainbow.bracket` paints captured nodes with the current
+rainbow color.
+
+For an example, let's add rainbow queries for the tree-sitter query (TSQ)
+language itself. These queries will go into a
+`runtime/queries/tsq/rainbows.scm` file in the repository root.
+
+First we'll add the `@rainbow.bracket` captures. TSQ only has parentheses and
+square brackets:
+
+```tsq
+["(" ")" "[" "]"] @rainbow.bracket
+```
+
+The ordering of the nodes within the alternation (square brackets) is not
+taken into consideration.
+
+> Note: Why are these nodes quoted? Most syntax highlights capture text
+> surrounded by parentheses. These are _named nodes_ and correspond to the
+> names of rules in the grammar. Brackets are usually written in tree-sitter
+> grammars as literal strings, for example:
+>
+> ```js
+> {
+>   // ...
+>   arguments: seq("(", repeat($.argument), ")"),
+>   // ...
+> }
+> ```
+>
+> Nodes written as literal strings in tree-sitter grammars may be captured
+> in queries with those same literal strings.
+
+Then we'll add `@rainbow.scope` captures. The easiest way to do this is to
+view the `grammar.js` file in the tree-sitter grammar's repository. For TSQ,
+that file is [here][tsq grammar.js]. As we scroll down the `grammar.js`, we
+see that the `(alternation)`, (L36) `(group)` (L57), `(named_node)` (L59),
+`(predicate)` (L87) and  `(wildcard_node)` (L97) nodes all contain literal
+parentheses or square brackets in their definitions. These nodes are all
+direct parents of brackets and happen to also be the nodes we want to change
+to the next rainbow color, so we capture them as `@rainbow.scope`.
+
+```tsq
+[
+  (group)
+  (named_node)
+  (wildcard_node)
+  (predicate)
+  (alternation)
+] @rainbow.scope
+```
+
+This strategy works as a rule of thumb for most programming and configuration
+languages. Markup languages can be trickier and may take additional
+experimentation to find the correct nodes to use for scopes and brackets.
+
+The `:tree-sitter-subtree` command shows the syntax tree under the primary
+selection in S-expression format and can be a useful tool for determining how
+to write a query.
+
+### Properties
+
+The `rainbow.include-children` property may be applied to `@rainbow.scope`
+captures. By default, all `@rainbow.bracket` captures must be direct descendant
+of a node captured with `@rainbow.scope` in a syntax tree in order to be
+highlighted. The `rainbow.include-children` property disables that check and
+allows `@rainbow.bracket` captures to be highlighted if they are direct or
+indirect descendants of some node captured with `@rainbow.scope`.
+
+For example, this property is used in the HTML rainbow queries.
+
+For a document like `<a>link</a>`, the syntax tree is:
+
+```tsq
+(element                   ; <a>link</a>
+  (start_tag               ; <a>
+    (tag_name))            ; a
+  (text)                   ; link
+  (end_tag                 ; </a>
+    (tag_name)))           ; a
+```
+
+If we want to highlight the `<`, `>` and `</` nodes with rainbow colors, we
+capture them as `@rainbow.bracket`:
+
+```tsq
+["<" ">" "</"] @rainbow.bracket
+```
+
+And we capture `(element)` as `@rainbow.scope` because `(element)` nodes nest
+within each other: they increment the nesting level and switch to the next
+color in the rainbow.
+
+```tsq
+(element) @rainbow.scope
+```
+
+But this combination of `@rainbow.scope` and `@rainbow.bracket` will not
+highlight any nodes. `<`, `>` and `</` are children of the `(start_tag)` and
+`(end_tag)` nodes. We can't capture `(start_tag)` and `(end_tag)` as
+`@rainbow.scope` because they don't nest other elements. We can fix this case
+by removing the requirement that `<`, `>` and `</` are direct descendants of
+`(element)` using the `rainbow.include-children` property.
+
+```tsq
+((element) @rainbow.scope
+ (#set! rainbow.include-children))
+```
+
+With this property set, `<`, `>`, and `</` will highlight with rainbow colors
+even though they aren't direct descendents of the `(element)` node.
+
+`rainbow.include-children` is not necessary for the vast majority of programming
+languages. It is only necessary when the node that increments the nesting level
+(changes rainbow color) is not the direct parent of the bracket node.
+
+[syntax highlighting queries]: https://tree-sitter.github.io/tree-sitter/syntax-highlighting#highlights
+[query syntax]: https://tree-sitter.github.io/tree-sitter/using-parsers#pattern-matching-with-queries
+[tsq grammar.js]: https://github.com/the-mikedavis/tree-sitter-tsq/blob/48b5e9f82ae0a4727201626f33a17f69f8e0ff86/grammar.js

book/src/guides/tags.md

@@ -0,0 +1,34 @@
+## Adding tags queries
+
+See tree-sitter's documentation on [Code Navigation Systems] for more
+background on tags queries.
+
+Helix provides LSP-like features such as document and workspace symbol pickers
+out-of-the-box for languages with `tags.scm` queries based on syntax trees. To
+be analyzed a language must have a tree-sitter grammar and a `tags.scm` query
+file which pattern matches interesting nodes from syntax trees.
+
+Query files should be placed in `runtime/queries/{language}/tags.scm`
+when contributing to Helix. You may place these under your local runtime
+directory (`~/.config/helix/runtime` in Linux for example) for the sake of
+testing.
+
+The following [captures][tree-sitter-captures] are recognized:
+
+| Capture name           |
+|---                     |
+| `definition.class`     |
+| `definition.constant`  |
+| `definition.function`  |
+| `definition.interface` |
+| `definition.macro`     |
+| `definition.module`    |
+| `definition.struct`    |
+| `definition.type`      |
+
+[Example query files][example-queries] can be found in the Helix GitHub
+repository.
+
+[Code Navigation Systems]: https://tree-sitter.github.io/tree-sitter/4-code-navigation.html
+[tree-sitter-captures]: https://tree-sitter.github.io/tree-sitter/using-parsers/queries/index.html
+[example-queries]: https://github.com/search?q=repo%3Ahelix-editor%2Fhelix+path%3A%2A%2A/tags.scm&type=Code

book/src/guides/textobject.md

@@ -1,4 +1,4 @@
-# Adding textobject queries
+## Adding textobject queries
 
 Helix supports textobjects that are language specific, such as functions, classes, etc.
 These textobjects require an accompanying tree-sitter grammar and a `textobjects.scm` query file
@@ -23,10 +23,13 @@ The following [captures][tree-sitter-captures] are recognized:
 | `test.inside`      |
 | `test.around`      |
 | `parameter.inside` |
+| `parameter.around` |
 | `comment.inside`   |
 | `comment.around`   |
 | `entry.inside`     |
 | `entry.around`     |
+| `xml-element.inside` |
+| `xml-element.around` |
 
 [Example query files][textobject-examples] can be found in the helix GitHub repository.
 
@@ -44,6 +47,6 @@ in its `textobjects.scm` file, function navigation should also work automaticall
 `function.movement` should be defined only if the node captured by `function.around`
 doesn't make sense in a navigation context.
 
-[tree-sitter-queries]: https://tree-sitter.github.io/tree-sitter/using-parsers#query-syntax
-[tree-sitter-captures]: https://tree-sitter.github.io/tree-sitter/using-parsers#capturing-nodes
+[tree-sitter-queries]: https://tree-sitter.github.io/tree-sitter/using-parsers/queries/1-syntax.html
+[tree-sitter-captures]: https://tree-sitter.github.io/tree-sitter/using-parsers/queries/2-operators.html#capturing-nodes
 [textobject-examples]: https://github.com/search?q=repo%3Ahelix-editor%2Fhelix+path%3A%2A%2A/textobjects.scm&type=Code&ref=advsearch&l=&l=

book/src/install.md

@@ -1,6 +1,7 @@
 # Installing Helix
 
-To install Helix, follow the instructions specific to your operating system.
+The typical way to install Helix is via [your operating system's package manager](./package-managers.md).
+
 Note that:
 
 - To get the latest nightly version of Helix, you need to
@@ -8,12 +9,16 @@ Note that:
 
 - To take full advantage of Helix, install the language servers for your
   preferred programming languages. See the
-  [wiki](https://github.com/helix-editor/helix/wiki/How-to-install-the-default-language-servers)
+  [wiki](https://github.com/helix-editor/helix/wiki/Language-Server-Configurations)
   for instructions.
 
 ## Pre-built binaries
 
 Download pre-built binaries from the [GitHub Releases page](https://github.com/helix-editor/helix/releases).
-Add the `hx` binary to your system's `$PATH` to use it from the command line, and copy the `runtime` directory into the config directory (for example `~/.config/helix/runtime` on Linux/macOS).
-The runtime location can be overriden via the HELIX_RUNTIME environment variable.
+The tarball contents include an `hx` binary and a `runtime` directory.
+To set up Helix:
 
+1. Add the `hx` binary to your system's `$PATH` to allow it to be used from the command line.
+2. Copy the `runtime` directory to a location that `hx` searches for runtime files. A typical location on Linux/macOS is `~/.config/helix/runtime`.
+
+To see the runtime directories that `hx` searches, run `hx --health`. If necessary, you can override the default runtime location by setting the `HELIX_RUNTIME` environment variable.

book/src/jumplist.md

@@ -0,0 +1,36 @@
+## Using the jumplist
+
+To help with quick navigation, Helix maintains a list of "jumps" called the jumplist.
+Whenever you make a significant movement (see next section), Helix stores your selections from before the move as a jump.
+A jump serves as a kind of checkpoint, allowing you to jump to a separate location, make edits, and return to where you were with your previous selections.
+This way, the jumplist tracks both your previous location and your selections.
+You can manually save a jump by using `Ctrl-s`.
+To jump backward in the jumplist, use `Ctrl-o`; to go forward, use `Ctrl-i`. To view and select from the full jumplist, use `Space-j` to open the jumplist picker.
+
+### What makes a jump
+The following is a non-exhaustive list of which actions add a jump to the jumplist:
+- Switching buffers
+  - Using the buffer picker, going to the next/previous buffer
+  - Going to the last accessed/modified file
+  - Making a new file (`:new FILE`)
+  - Opening a file (`:open FILE`)
+    - Includes `:log-open`, `:config-open`, `:config-open-workspace`, `:tutor`
+  - Navigating by pickers, global search, or the file explorer
+  - `goto_file` (`gf`)
+- Big in-file movements
+  - `select_regex` (`s`)
+  - `split_regex` (`S`)
+  - `search` (`/`)
+  - `keep_selections` and `remove_selections` (`K` and `<A-K>`)
+  - `goto_file_start` (`gg`)
+  - `goto_file_end`
+  - `goto_last_line` (`ge`)
+  - `:goto 123` / `:123` / `123G`
+  - `goto_definition` (`gd`)
+  - `goto_declaration` (`gD`)
+  - `goto_type_definition` (`gy`)
+  - `goto_reference` (`gr`)
+- Other
+  - `Ctrl-s` manually creates a jump
+  - Trying to close a modified buffer can switch you to that buffer and create a jump
+  - The debugger can create jumps as you jump stack frames

book/src/keymap.md

@@ -35,6 +35,8 @@ Normal mode is the default mode when you launch helix. You can return to it from
 
 > NOTE: Unlike Vim, `f`, `F`, `t` and `T` are not confined to the current line.
 
+> Hereafter, `<n>` represents an integer by typing a sequence of digits.
+
 | Key                   | Description                                        | Command                     |
 | -----                 | -----------                                        | -------                     |
 | `h`, `Left`           | Move left                                          | `move_char_left`            |
@@ -47,11 +49,11 @@ Normal mode is the default mode when you launch helix. You can return to it from
 | `W`                   | Move next WORD start                               | `move_next_long_word_start` |
 | `B`                   | Move previous WORD start                           | `move_prev_long_word_start` |
 | `E`                   | Move next WORD end                                 | `move_next_long_word_end`   |
-| `t`                   | Find 'till next char                               | `find_till_char`            |
+| `t`                   | Find till next char                                | `find_till_char`            |
 | `f`                   | Find next char                                     | `find_next_char`            |
-| `T`                   | Find 'till previous char                           | `till_prev_char`            |
+| `T`                   | Find till previous char                            | `till_prev_char`            |
 | `F`                   | Find previous char                                 | `find_prev_char`            |
-| `G`                   | Go to line number `<n>`                            | `goto_line`                 |
+| `<n>G`, `<n>gg`       | Go to line number `<n>`                            | `goto_line`                 |
 | `Alt-.`               | Repeat last motion (`f`, `t`, `m`, `[` or `]`)     | `repeat_last_motion`        |
 | `Home`                | Move to the start of the line                      | `goto_line_start`           |
 | `End`                 | Move to the end of the line                        | `goto_line_end`             |
@@ -112,39 +114,43 @@ Normal mode is the default mode when you launch helix. You can return to it from
 
 ### Selection manipulation
 
-| Key                   | Description                                                       | Command                              |
-| -----                 | -----------                                                       | -------                              |
-| `s`                   | Select all regex matches inside selections                        | `select_regex`                       |
-| `S`                   | Split selection into sub selections on regex matches              | `split_selection`                    |
-| `Alt-s`               | Split selection on newlines                                       | `split_selection_on_newline`         |
-| `Alt-minus`           | Merge selections                                                  | `merge_selections`                   |
-| `Alt-_`               | Merge consecutive selections                                      | `merge_consecutive_selections`       |
-| `&`                   | Align selection in columns                                        | `align_selections`                   |
-| `_`                   | Trim whitespace from the selection                                | `trim_selections`                    |
-| `;`                   | Collapse selection onto a single cursor                           | `collapse_selection`                 |
-| `Alt-;`               | Flip selection cursor and anchor                                  | `flip_selections`                    |
-| `Alt-:`               | Ensures the selection is in forward direction                     | `ensure_selections_forward`          |
-| `,`                   | Keep only the primary selection                                   | `keep_primary_selection`             |
-| `Alt-,`               | Remove the primary selection                                      | `remove_primary_selection`           |
-| `C`                   | Copy selection onto the next line (Add cursor below)              | `copy_selection_on_next_line`        |
-| `Alt-C`               | Copy selection onto the previous line (Add cursor above)          | `copy_selection_on_prev_line`        |
-| `(`                   | Rotate main selection backward                                    | `rotate_selections_backward`         |
-| `)`                   | Rotate main selection forward                                     | `rotate_selections_forward`          |
-| `Alt-(`               | Rotate selection contents backward                                | `rotate_selection_contents_backward` |
-| `Alt-)`               | Rotate selection contents forward                                 | `rotate_selection_contents_forward`  |
-| `%`                   | Select entire file                                                | `select_all`                         |
-| `x`                   | Select current line, if already selected, extend to next line     | `extend_line_below`                  |
-| `X`                   | Extend selection to line bounds (line-wise selection)             | `extend_to_line_bounds`              |
-| `Alt-x`               | Shrink selection to line bounds (line-wise selection)             | `shrink_to_line_bounds`              |
-| `J`                   | Join lines inside selection                                       | `join_selections`                    |
-| `Alt-J`               | Join lines inside selection and select the inserted space         | `join_selections_space`              |
-| `K`                   | Keep selections matching the regex                                | `keep_selections`                    |
-| `Alt-K`               | Remove selections matching the regex                              | `remove_selections`                  |
-| `Ctrl-c`              | Comment/uncomment the selections                                  | `toggle_comments`                    |
-| `Alt-o`, `Alt-up`     | Expand selection to parent syntax node (**TS**)                   | `expand_selection`                   |
-| `Alt-i`, `Alt-down`   | Shrink syntax tree object selection (**TS**)                      | `shrink_selection`                   |
-| `Alt-p`, `Alt-left`   | Select previous sibling node in syntax tree (**TS**)              | `select_prev_sibling`                |
-| `Alt-n`, `Alt-right`  | Select next sibling node in syntax tree (**TS**)                  | `select_next_sibling`                |
+| Key                      | Description                                                       | Command                              |
+| -----                    | -----------                                                       | -------                              |
+| `s`                      | Select all regex matches inside selections                        | `select_regex`                       |
+| `S`                      | Split selection into sub selections on regex matches              | `split_selection`                    |
+| `Alt-s`                  | Split selection on newlines                                       | `split_selection_on_newline`         |
+| `Alt-minus`              | Merge selections                                                  | `merge_selections`                   |
+| `Alt-_`                  | Merge consecutive selections                                      | `merge_consecutive_selections`       |
+| `&`                      | Align selection in columns                                        | `align_selections`                   |
+| `_`                      | Trim whitespace from the selection                                | `trim_selections`                    |
+| `;`                      | Collapse selection onto a single cursor                           | `collapse_selection`                 |
+| `Alt-;`                  | Flip selection cursor and anchor                                  | `flip_selections`                    |
+| `Alt-:`                  | Ensures the selection is in forward direction                     | `ensure_selections_forward`          |
+| `,`                      | Keep only the primary selection                                   | `keep_primary_selection`             |
+| `Alt-,`                  | Remove the primary selection                                      | `remove_primary_selection`           |
+| `C`                      | Copy selection onto the next line (Add cursor below)              | `copy_selection_on_next_line`        |
+| `Alt-C`                  | Copy selection onto the previous line (Add cursor above)          | `copy_selection_on_prev_line`        |
+| `(`                      | Rotate main selection backward                                    | `rotate_selections_backward`         |
+| `)`                      | Rotate main selection forward                                     | `rotate_selections_forward`          |
+| `Alt-(`                  | Rotate selection contents backward                                | `rotate_selection_contents_backward` |
+| `Alt-)`                  | Rotate selection contents forward                                 | `rotate_selection_contents_forward`  |
+| `%`                      | Select entire file                                                | `select_all`                         |
+| `x`                      | Select current line, if already selected, extend to next line     | `extend_line_below`                  |
+| `X`                      | Extend selection to line bounds (line-wise selection)             | `extend_to_line_bounds`              |
+| `Alt-x`                  | Shrink selection to line bounds (line-wise selection)             | `shrink_to_line_bounds`              |
+| `J`                      | Join lines inside selection                                       | `join_selections`                    |
+| `Alt-J`                  | Join lines inside selection and select the inserted space         | `join_selections_space`              |
+| `K`                      | Keep selections matching the regex                                | `keep_selections`                    |
+| `Alt-K`                  | Remove selections matching the regex                              | `remove_selections`                  |
+| `Ctrl-c`                 | Comment/uncomment the selections                                  | `toggle_comments`                    |
+| `Alt-o`, `Alt-up`        | Expand selection to parent syntax node (**TS**)                   | `expand_selection`                   |
+| `Alt-i`, `Alt-down`      | Shrink syntax tree object selection (**TS**)                      | `shrink_selection`                   |
+| `Alt-p`, `Alt-left`      | Select previous sibling node in syntax tree (**TS**)              | `select_prev_sibling`                |
+| `Alt-n`, `Alt-right`     | Select next sibling node in syntax tree (**TS**)                  | `select_next_sibling`                |
+| `Alt-a`                  | Select all sibling nodes in syntax tree (**TS**)                  | `select_all_siblings`                |
+| `Alt-I`, `Alt-Shift-down`| Select all children nodes in syntax tree (**TS**)                 | `select_all_children`                |
+| `Alt-e`                  | Move to end of parent node in syntax tree (**TS**)                | `move_parent_node_end`               |
+| `Alt-b`                  | Move to start of parent node in syntax tree (**TS**)              | `move_parent_node_start`             |
 
 ### Search
 
@@ -156,7 +162,8 @@ Search commands all operate on the `/` register by default. To use a different r
 | `?`   | Search for previous pattern                 | `rsearch`            |
 | `n`   | Select next search match                    | `search_next`        |
 | `N`   | Select previous search match                | `search_prev`        |
-| `*`   | Use current selection as the search pattern | `search_selection`   |
+| `*`   | Use current selection as the search pattern, automatically wrapping with `\b` on word boundaries | `search_selection_detect_word_boundaries` |
+| `Alt-*` | Use current selection as the search pattern | `search_selection` |
 
 ### Minor modes
 
@@ -207,7 +214,10 @@ Jumps to various locations.
 
 | Key   | Description                                      | Command                    |
 | ----- | -----------                                      | -------                    |
-| `g`   | Go to line number `<n>` else start of file       | `goto_file_start`          |
+| `<n>g`| Go to line number `<n>`                          | `goto_file_start`          |
+| `g`   | Go to the start of the file                      | `goto_file_start`          |
+| <code>&lt;n&gt;&#124;</code>  | Go to column number `<n>`      | `goto_column`              |
+| <code>&#124;</code>     | Go to the start of line        | `goto_column`              |
 | `e`   | Go to the end of the file                        | `goto_last_line`           |
 | `f`   | Go to files in the selections                    | `goto_file`                |
 | `h`   | Go to the start of the line                      | `goto_line_start`          |
@@ -278,7 +288,7 @@ This layer is a kludge of mappings, mostly pickers.
 
 | Key     | Description                                                             | Command                                    |
 | -----   | -----------                                                             | -------                                    |
-| `f`     | Open file picker                                                        | `file_picker`                              |
+| `f`     | Open file picker at LSP workspace root                                  | `file_picker`                              |
 | `F`     | Open file picker at current working directory                           | `file_picker_in_current_directory`         |
 | `b`     | Open buffer picker                                                      | `buffer_picker`                            |
 | `j`     | Open jumplist picker                                                    | `jumplist_picker`                          |
@@ -320,10 +330,14 @@ Displays documentation for item under cursor. Remapping currently not supported.
 
 Displays documentation for the selected completion item. Remapping currently not supported.
 
-| Key                         | Description    |
-| ----                        | -----------    |
-| `Shift-Tab`, `Ctrl-p`, `Up` | Previous entry |
-| `Tab`, `Ctrl-n`, `Down`     | Next entry     |
+| Key                         | Description                      |
+| ----                        | -----------                      |
+| `Shift-Tab`, `Ctrl-p`, `Up` | Previous entry                   |
+| `Tab`, `Ctrl-n`, `Down`     | Next entry                       |
+| `Enter`                     | Close menu and accept completion |
+| `Ctrl-c`                    | Close menu and reject completion |
+
+Any other keypresses result in the completion being accepted.
 
 ##### Signature-help Popup
 
@@ -338,30 +352,32 @@ Displays the signature of the selected completion item. Remapping currently not
 
 These mappings are in the style of [vim-unimpaired](https://github.com/tpope/vim-unimpaired).
 
-| Key      | Description                                  | Command               |
-| -----    | -----------                                  | -------               |
-| `]d`     | Go to next diagnostic (**LSP**)              | `goto_next_diag`      |
-| `[d`     | Go to previous diagnostic (**LSP**)          | `goto_prev_diag`      |
-| `]D`     | Go to last diagnostic in document (**LSP**)  | `goto_last_diag`      |
-| `[D`     | Go to first diagnostic in document (**LSP**) | `goto_first_diag`     |
-| `]f`     | Go to next function (**TS**)                 | `goto_next_function`  |
-| `[f`     | Go to previous function (**TS**)             | `goto_prev_function`  |
-| `]t`     | Go to next type definition (**TS**)          | `goto_next_class`     |
-| `[t`     | Go to previous type definition (**TS**)      | `goto_prev_class`     |
-| `]a`     | Go to next argument/parameter (**TS**)       | `goto_next_parameter` |
-| `[a`     | Go to previous argument/parameter (**TS**)   | `goto_prev_parameter` |
-| `]c`     | Go to next comment (**TS**)                  | `goto_next_comment`   |
-| `[c`     | Go to previous comment (**TS**)              | `goto_prev_comment`   |
-| `]T`     | Go to next test (**TS**)                     | `goto_next_test`      |
-| `[T`     | Go to previous test (**TS**)                 | `goto_prev_test`      |
-| `]p`     | Go to next paragraph                         | `goto_next_paragraph` |
-| `[p`     | Go to previous paragraph                     | `goto_prev_paragraph` |
-| `]g`     | Go to next change                            | `goto_next_change`    |
-| `[g`     | Go to previous change                        | `goto_prev_change`    |
-| `]G`     | Go to last change                            | `goto_last_change`    |
-| `[G`     | Go to first change                           | `goto_first_change`   |
-| `]Space` | Add newline below                            | `add_newline_below`   |
-| `[Space` | Add newline above                            | `add_newline_above`   |
+| Key      | Description                                  | Command                 |
+| -----    | -----------                                  | -------                 |
+| `]d`     | Go to next diagnostic (**LSP**)              | `goto_next_diag`        |
+| `[d`     | Go to previous diagnostic (**LSP**)          | `goto_prev_diag`        |
+| `]D`     | Go to last diagnostic in document (**LSP**)  | `goto_last_diag`        |
+| `[D`     | Go to first diagnostic in document (**LSP**) | `goto_first_diag`       |
+| `]f`     | Go to next function (**TS**)                 | `goto_next_function`    |
+| `[f`     | Go to previous function (**TS**)             | `goto_prev_function`    |
+| `]t`     | Go to next type definition (**TS**)          | `goto_next_class`       |
+| `[t`     | Go to previous type definition (**TS**)      | `goto_prev_class`       |
+| `]a`     | Go to next argument/parameter (**TS**)       | `goto_next_parameter`   |
+| `[a`     | Go to previous argument/parameter (**TS**)   | `goto_prev_parameter`   |
+| `]c`     | Go to next comment (**TS**)                  | `goto_next_comment`     |
+| `[c`     | Go to previous comment (**TS**)              | `goto_prev_comment`     |
+| `]T`     | Go to next test (**TS**)                     | `goto_next_test`        |
+| `[T`     | Go to previous test (**TS**)                 | `goto_prev_test`        |
+| `]p`     | Go to next paragraph                         | `goto_next_paragraph`   |
+| `[p`     | Go to previous paragraph                     | `goto_prev_paragraph`   |
+| `]g`     | Go to next change                            | `goto_next_change`      |
+| `[g`     | Go to previous change                        | `goto_prev_change`      |
+| `]G`     | Go to last change                            | `goto_last_change`      |
+| `[G`     | Go to first change                           | `goto_first_change`     |
+| `[x`     | Go to next (X)HTML element                   | `goto_next_xml_element` |
+| `]x`     | Go to previous (X)HTML element               | `goto_prev_xml_element` |
+| `]Space` | Add newline below                            | `add_newline_below`     |
+| `[Space` | Add newline above                            | `add_newline_above`     |
 
 ## Insert mode
 
@@ -436,6 +452,8 @@ you to selectively add search terms to your selections.
 ## Picker
 
 Keys to use within picker. Remapping currently not supported.
+See the documentation page on [pickers](./pickers.md) for more info.
+[Prompt](#prompt) keybinds also work in pickers, except where they conflict with picker keybinds.
 
 | Key                          | Description                                                |
 | -----                        | -------------                                              |

book/src/lang-support.md

@@ -1,4 +1,4 @@
-# Language Support
+## Language Support
 
 The following languages and Language Servers are supported. To use
 Language Server features, you must first [configure][lsp-config-wiki] the

book/src/languages.md

@@ -1,4 +1,4 @@
-# Languages
+## Languages
 
 Language-specific settings and settings for language servers are configured
 in `languages.toml` files.
@@ -13,7 +13,7 @@ There are three possible locations for a `languages.toml` file:
 
 2. In your [configuration directory](./configuration.md). This overrides values
    from the built-in language configuration. For example, to disable
-   auto-LSP-formatting in Rust:
+   auto-formatting for Rust:
 
    ```toml
    # in <config_dir>/helix/languages.toml
@@ -60,17 +60,21 @@ These configuration keys are available:
 | `shebangs`            | The interpreters from the shebang line, for example `["sh", "bash"]` |
 | `roots`               | A set of marker files to look for when trying to find the workspace root. For example `Cargo.lock`, `yarn.lock` |
 | `auto-format`         | Whether to autoformat this language when saving               |
-| `diagnostic-severity` | Minimal severity of diagnostic for it to be displayed. (Allowed values: `Error`, `Warning`, `Info`, `Hint`) |
+| `diagnostic-severity` | Minimal severity of diagnostic for it to be displayed. (Allowed values: `error`, `warning`, `info`, `hint`) |
 | `comment-tokens`      | The tokens to use as a comment token, either a single token `"//"` or an array `["//", "///", "//!"]` (the first token will be used for commenting). Also configurable as `comment-token` for backwards compatibility|
 | `block-comment-tokens`| The start and end tokens for a multiline comment either an array or single table of `{ start = "/*", end = "*/"}`. The first set of tokens will be used for commenting, any pairs in the array can be uncommented |
 | `indent`              | The indent to use. Has sub keys `unit` (the text inserted into the document when indenting; usually set to N spaces or `"\t"` for tabs) and `tab-width` (the number of spaces rendered for a tab) |
 | `language-servers`    | The Language Servers used for this language. See below for more information in the section [Configuring Language Servers for a language](#configuring-language-servers-for-a-language)   |
 | `grammar`             | The tree-sitter grammar to use (defaults to the value of `name`) |
-| `formatter`           | The formatter for the language, it will take precedence over the lsp when defined. The formatter must be able to take the original file as input from stdin and write the formatted file to stdout |
-| `soft-wrap` | [editor.softwrap](./configuration.md#editorsoft-wrap-section)
+| `formatter`           | The formatter for the language, it will take precedence over the lsp when defined. The formatter must be able to take the original file as input from stdin and write the formatted file to stdout. The filename of the current buffer can be passed as argument by using the `%{buffer_name}` expansion variable. See below for more information in the [Configuring the formatter command](#configuring-the-formatter-command) |
+| `soft-wrap`           | [editor.softwrap](./editor.md#editorsoft-wrap-section)
 | `text-width`          |  Maximum line length. Used for the `:reflow` command and soft-wrapping if `soft-wrap.wrap-at-text-width` is set, defaults to `editor.text-width`   |
+| `rulers`              | Overrides the `editor.rulers` config key for the language. |
+| `path-completion`     | Overrides the `editor.path-completion` config key for the language. |
+| `word-completion`     | Overrides the [`editor.word-completion`](./editor.md#editorword-completion-section) configuration for the language. |
 | `workspace-lsp-roots`     | Directories relative to the workspace root that are treated as LSP roots. Should only be set in `.helix/config.toml`. Overwrites the setting of the same name in `config.toml` if set. |
 | `persistent-diagnostic-sources` | An array of LSP diagnostic sources assumed unchanged when the language server resends the same set of diagnostics. Helix can track the position for these diagnostics internally instead. Useful for diagnostics that are recomputed on save.
+| `rainbow-brackets` | Overrides the `editor.rainbow-brackets` config key for the language |
 
 ### File-type detection and the `file-types` key
 
@@ -100,6 +104,16 @@ with the following priorities:
    the file extension of a given file wins. In the example above, the `"toml"`
    config matches files like `Cargo.toml` or `languages.toml`.
 
+### Configuring the formatter command
+
+[Command line expansions](./command-line.md#expansions) are supported in the arguments
+of the formatter command. In particular, the `%{buffer_name}` variable can be passed as
+argument to the formatter:
+
+```toml
+formatter = { command = "mylang-formatter" , args = ["--stdin", "--stdin-filename", "%{buffer_name}"] }
+```
+
 ## Language Server configuration
 
 Language servers are configured separately in the table `language-server` in the same file as the languages `languages.toml`
@@ -127,7 +141,7 @@ These are the available options for a language server.
 | ----                       | -----------                                                                                                                       |
 | `command`                  | The name or path of the language server binary to execute. Binaries must be in `$PATH`                                            |
 | `args`                     | A list of arguments to pass to the language server binary                                                                         |
-| `config`                   | LSP initialization options                                                                                                        |
+| `config`                   | Language server initialization options                                                                                            |
 | `timeout`                  | The maximum time a request to the language server may take, in seconds. Defaults to `20`                                          |
 | `environment`              | Any environment variables that will be used when starting the language server `{ "KEY1" = "Value1", "KEY2" = "Value2" }`          |
 | `required-root-patterns`   | A list of `glob` patterns to look for in the working directory. The language server is started if at least one of them is found.  |
@@ -240,4 +254,4 @@ use-grammars = { except = [ "yaml", "json" ] }
 
 When omitted, all grammars are fetched and built.
 
-[treesitter-language-injection]: https://tree-sitter.github.io/tree-sitter/syntax-highlighting#language-injection
+[treesitter-language-injection]: https://tree-sitter.github.io/tree-sitter/3-syntax-highlighting.html#language-injection

book/src/other-software.md

@@ -0,0 +1,35 @@
+# Helix mode in other software
+
+Helix' keymap and interaction model ([Using Helix](#usage.md)) is easier to adopt if it can be used consistently in many editing contexts. Yet, certain use cases cannot easily be addressed directly in Helix. Similar to vim, this leads to the creation of "Helix mode" in various other software products, allowing Helix-style editing for a greater variety of use cases.
+
+"Helix mode" is frequently still in early stages or missing entirely. For such cases, we also link to relevant bugs or discussions.
+
+## Other editors
+
+| Editor | Plugin or feature providing Helix editing | Comments
+| --- | --- | --- |
+| [Vim](https://www.vim.org/) | [helix.vim](https://github.com/chtenb/helix.vim) config |
+| [IntelliJ IDEA](https://www.jetbrains.com/idea/) / [Android Studio](https://developer.android.com/studio)| [IdeaVim](https://plugins.jetbrains.com/plugin/164-ideavim) plugin + [helix.idea.vim](https://github.com/chtenb/helix.vim) config | Minimum recommended version is IdeaVim 2.19.0.
+| [Visual Studio](https://visualstudio.microsoft.com/) | [VsVim](https://marketplace.visualstudio.com/items?itemName=JaredParMSFT.VsVim) plugin + [helix.vs.vim](https://github.com/chtenb/helix.vim) config | 
+| [Visual Studio Code](https://code.visualstudio.com/) | [Dance](https://marketplace.visualstudio.com/items?itemName=gregoire.dance) extension, or its [Helix fork](https://marketplace.visualstudio.com/items?itemName=kend.dancehelixkey) | The Helix fork has diverged. You can also use the original Dance and tweak its keybindings directly (try [this config](https://github.com/71/dance/issues/299#issuecomment-1655509531)).
+| [Visual Studio Code](https://code.visualstudio.com/) | [Helix for VS Code](https://marketplace.visualstudio.com/items?itemName=jasew.vscode-helix-emulation) extension|
+| [Zed](https://zed.dev/) | native via keybindings ([Bug](https://github.com/zed-industries/zed/issues/4642)) |
+| [CodeMirror](https://codemirror.net/) | [codemirror-helix](https://gitlab.com/_rvidal/codemirror-helix) |
+| [Lite XL](https://lite-xl.com/) | [lite-modal-hx](https://codeberg.org/Mandarancio/lite-modal-hx) |
+| [Lapce](https://lap.dev/lapce/) | | Requested: https://github.com/lapce/lapce/issues/281 |
+
+
+## Shells
+
+| Shell | Plugin or feature providing Helix editing 
+| --- | --- 
+| Fish | [Feature Request](https://github.com/fish-shell/fish-shell/issues/7748) 
+| Fish | [fish-helix](https://github.com/sshilovsky/fish-helix/tree/main) 
+| Zsh | [helix-zsh](https://github.com/john-h-k/helix-zsh) or [zsh-helix-mode](https://github.com/Multirious/zsh-helix-mode)
+| Nushell | [Feature Request](https://github.com/nushell/reedline/issues/639) 
+
+## Other software
+
+| Software | Plugin or feature providing Helix editing. | Comments
+| --- | --- | --- |
+| [Obsidian](https://obsidian.md/) | [Obsidian-Helix](https://github.com/Sinono3/obsidian-helix) | Uses `codemirror-helix` listed above.

book/src/package-managers.md

@@ -1,7 +1,7 @@
 ## Package managers
 
 - [Linux](#linux)
-  - [Ubuntu](#ubuntu)
+  - [Ubuntu/Debian](#ubuntudebian)
   - [Fedora/RHEL](#fedorarhel)
   - [Arch Linux extra](#arch-linux-extra)
   - [NixOS](#nixos)
@@ -17,21 +17,18 @@
   - [Chocolatey](#chocolatey)
   - [MSYS2](#msys2)
 
-[![Packaging status](https://repology.org/badge/vertical-allrepos/helix.svg)](https://repology.org/project/helix/versions)
+[![Packaging status](https://repology.org/badge/vertical-allrepos/helix-editor.svg)](https://repology.org/project/helix-editor/versions)
 
 ## Linux
 
 The following third party repositories are available:
 
-### Ubuntu
+### Ubuntu/Debian
 
-Add the `PPA` for Helix:
+Install the Debian package [from the release page](https://github.com/helix-editor/helix/releases/latest).
 
-```sh
-sudo add-apt-repository ppa:maveonair/helix-editor
-sudo apt update
-sudo apt install helix
-```
+If you are running a system older than Ubuntu 22.04, Mint 21, or Debian 12, you can build the `.deb` file locally
+[from source](./building-from-source.md#building-the-debian-package).
 
 ### Fedora/RHEL
 
@@ -101,7 +98,15 @@ Download the official Helix AppImage from the [latest releases](https://github.c
 chmod +x helix-*.AppImage # change permission for executable mode
 ./helix-*.AppImage # run helix
 ```
- 
+
+You can optionally [add the `.desktop` file](./building-from-source.md#configure-the-desktop-shortcut). Helix must be installed in `PATH` with the name `hx`. For example:
+```sh
+mkdir -p "$HOME/.local/bin"
+mv helix-*.AppImage "$HOME/.local/bin/hx"
+```
+
+and make sure `~/.local/bin` is in your `PATH`.
+
 ## macOS
 
 ### Homebrew Core
@@ -113,7 +118,7 @@ brew install helix
 ### MacPorts
 
 ```sh
-port install helix
+sudo port install helix
 ```
 
 ## Windows

book/src/pickers.md

@@ -0,0 +1,11 @@
+## Using pickers
+
+Helix has a variety of pickers, which are interactive windows used to select various kinds of items. These include a file picker, global search picker, and more. Most pickers are accessed via keybindings in [space mode](./keymap.md#space-mode). Pickers have their own [keymap](./keymap.md#picker) for navigation.
+
+### Filtering Picker Results
+
+Most pickers perform fuzzy matching using [fzf syntax](https://github.com/junegunn/fzf?tab=readme-ov-file#search-syntax). Two exceptions are the global search picker, which uses regex, and the workspace symbol picker, which passes search terms to the language server. Note that OR operations (`|`) are not currently supported.
+
+If a picker shows multiple columns, you may apply the filter to a specific column by prefixing the column name with `%`. Column names can be shortened to any prefix, so `%p`, `%pa` or `%pat` all mean the same as `%path`. For example, a query of `helix %p .toml !lang` in the global search picker searches for the term "helix" within files with paths ending in ".toml" but not including "lang".
+
+You can insert the contents of a [register](./registers.md) using `Ctrl-r` followed by a register name. For example, one could insert the currently selected text using `Ctrl-r`-`.`, or the directory of the current file using `Ctrl-r`-`%` followed by `Ctrl-w` to remove the last path section. The global search picker will use the contents of the [search register](./registers.md#default-registers) if you press `Enter` without typing a filter. For example, pressing `*`-`Space-/`-`Enter` will start a global search for the currently selected text.

book/src/remapping.md

@@ -1,13 +1,34 @@
-# Key remapping
+## Key remapping
 
 Helix currently supports one-way key remapping through a simple TOML configuration
 file. (More powerful solutions such as rebinding via commands will be
 available in the future).
 
+There are three kinds of commands that can be used in keymaps:
+
+* Static commands: commands like `move_char_right` which are usually bound to
+  keys and used for movement and editing. A list of static commands is
+  available in the [Keymap](./keymap.html) documentation and in the source code
+  in [`helix-term/src/commands.rs`](https://github.com/helix-editor/helix/blob/master/helix-term/src/commands.rs)
+  at the invocation of `static_commands!` macro.
+* Typable commands: commands that can be executed from command mode (`:`), for
+  example `:write!`. See the [Commands](./commands.md) documentation for a
+  list of available typeable commands or the `TypableCommandList` declaration in
+  the source code at [`helix-term/src/commands/typed.rs`](https://github.com/helix-editor/helix/blob/master/helix-term/src/commands/typed.rs).
+* Macros: sequences of keys that are executed in order. These keybindings
+  start with `@` and then list any number of keys to be executed. For example
+  `@miw` can be used to select the surrounding word. For now, macro keybindings
+  are not allowed in keybinding sequences due to limitations in the way that
+  command sequences are executed. Modifier keys (e.g. Alt+o) can be used
+  like `"<A-o>"`, e.g. `"@miw<A-o>"`
+
 To remap keys, create a `config.toml` file in your `helix` configuration
 directory (default `~/.config/helix` on Linux systems) with a structure like
 this:
 
+> 💡 To set a modifier + key as a keymap, type `A-X = ...` or  `C-X = ...` for Alt + X or Ctrl + X. Combine with Shift using a dash, e.g. `C-S-esc`. 
+> Within macros, wrap them in `<>`, e.g. `<A-X>` and `<C-X>` to distinguish from the `A` or `C` keys.
+
 ```toml
 # At most one section each of 'keys.normal', 'keys.insert' and 'keys.select'
 [keys.normal]
@@ -18,6 +39,7 @@ w = "move_line_up" # Maps the 'w' key move_line_up
 "C-S-esc" = "extend_line" # Maps Ctrl-Shift-Escape to extend_line
 g = { a = "code_action" } # Maps `ga` to show possible code actions
 "ret" = ["open_below", "normal_mode"] # Maps the enter key to open_below then re-enter normal mode
+"A-x" = "@x<A-d>" # Maps Alt-x to a macro selecting the whole line and deleting it without yanking it
 
 [keys.insert]
 "A-x" = "normal_mode"     # Maps Alt-X to enter normal mode
@@ -50,30 +72,53 @@ t = ":run-shell-command cargo test"
 
 ## Special keys and modifiers
 
-Ctrl, Shift and Alt modifiers are encoded respectively with the prefixes
-`C-`, `S-` and `A-`. Special keys are encoded as follows:
+Ctrl, Shift and Alt modifiers are encoded respectively with the prefixes `C-`, `S-` and `A-`.
 
-| Key name     | Representation |
-| ---          | ---            |
-| Backspace    | `"backspace"`  |
-| Space        | `"space"`      |
-| Return/Enter | `"ret"`        |
-| \-           | `"minus"`      |
-| Left         | `"left"`       |
-| Right        | `"right"`      |
-| Up           | `"up"`         |
-| Down         | `"down"`       |
-| Home         | `"home"`       |
-| End          | `"end"`        |
-| Page Up      | `"pageup"`     |
-| Page Down    | `"pagedown"`   |
-| Tab          | `"tab"`        |
-| Delete       | `"del"`        |
-| Insert       | `"ins"`        |
-| Null         | `"null"`       |
-| Escape       | `"esc"`        |
+The [Super key](https://en.wikipedia.org/wiki/Super_key_(keyboard_button)) - the Windows/Linux
+key or the Command key on Mac keyboards - is also supported when using a terminal emulator that
+supports the [enhanced keyboard protocol](https://github.com/helix-editor/helix/wiki/Terminal-Support#enhanced-keyboard-protocol).
+The super key is encoded with prefixes `Meta-`, `Cmd-` or `Win-`. These are all synonyms for the
+super modifier - binding a key with a `Win-` modifier will mean it can be used with the
+Windows/Linux key or the Command key.
+
+```toml
+[keys.normal]
+C-s = ":write" # Ctrl and 's' to write
+Cmd-s = ":write" # Cmd or Win or Meta and 's' to write
+```
+
+Special keys are encoded as follows:
+
+| Key name         | Representation |
+| ---              | ---            |
+| Backspace        | `"backspace"`  |
+| Space            | `"space"`      |
+| Return/Enter     | `"ret"`        |
+| Left             | `"left"`       |
+| Right            | `"right"`      |
+| Up               | `"up"`         |
+| Down             | `"down"`       |
+| Home             | `"home"`       |
+| End              | `"end"`        |
+| Page Up          | `"pageup"`     |
+| Page Down        | `"pagedown"`   |
+| Tab              | `"tab"`        |
+| Delete           | `"del"`        |
+| Insert           | `"ins"`        |
+| Null             | `"null"`       |
+| Escape           | `"esc"`        |
+| Less Than (<)    | `"lt"`         |
+| Greater Than (>) | `"gt"`         |
 
 Keys can be disabled by binding them to the `no_op` command.
 
-A list of commands is available in the [Keymap](https://docs.helix-editor.com/keymap.html) documentation
- and in the source code at [`helix-term/src/commands.rs`](https://github.com/helix-editor/helix/blob/master/helix-term/src/commands.rs) at the invocation of `static_commands!` macro and the `TypableCommandList`.
+All other keys such as `?`, `!`, `-` etc. can be used literally:
+
+```toml
+[keys.normal]
+"?" = ":write"
+"!" = ":write"
+"-" = ":write"
+```
+
+Note: `-` can't be used when combined with a modifier, for example `Alt` + `-` should be written as `A-minus`. `A--` is not accepted.

book/src/textobjects.md

@@ -24,6 +24,7 @@ function or block of code.
 | `c`                    | Comment                  |
 | `T`                    | Test                     |
 | `g`                    | Change                   |
+| `x`                    | (X)HTML element          |
 
 > 💡 `f`, `t`, etc. need a tree-sitter grammar active for the current
 document and a special tree-sitter query file to work properly. [Only

book/src/themes.md

@@ -1,4 +1,4 @@
-# Themes
+## Themes
 
 To use a theme add `theme = "<name>"` to the top of your [`config.toml`](./configuration.md) file, or select it during runtime using `:theme <name>`.
 
@@ -130,13 +130,24 @@ inherits = "boo_berry"
 berry = "#2A2A4D"
 ```
 
+### Rainbow
+
+The `rainbow` key is used for rainbow highlight for matching brackets.
+The key is a list of styles.
+
+```toml
+rainbow = ["#ff0000", "#ffa500", "#fff000", { fg = "#00ff00", modifiers = ["bold"] }]
+```
+
+Colors from the palette and modifiers may be used.
+
 ### Scopes
 
 The following is a list of scopes available to use for styling:
 
 #### Syntax highlighting
 
-These keys match [tree-sitter scopes](https://tree-sitter.github.io/tree-sitter/syntax-highlighting#theme).
+These keys match [tree-sitter scopes](https://tree-sitter.github.io/tree-sitter/3-syntax-highlighting.html#highlights).
 
 When determining styling for a highlight, the longest matching theme key will be used. For example, if the highlight is `function.builtin.static`, the key `function.builtin` will be used instead of `function`.
 
@@ -171,8 +182,10 @@ We use a similar set of scopes as
 
 - `comment` - Code comments
   - `line` - Single line comments (`//`)
+    - `documentation` - Line documentation comments (e.g. `///` in Rust)
   - `block` - Block comments (e.g. (`/* */`)
-    - `documentation` - Documentation comments (e.g. `///` in Rust)
+    - `documentation` - Block documentation comments (e.g. `/** */` in Rust)
+  - `unused` - Unused variables and patterns, e.g. `_` and `_foo`
 
 - `variable` - Variables
   - `builtin` - Reserved language variables (`self`, `this`, `super`, etc.)
@@ -181,7 +194,7 @@ We use a similar set of scopes as
     - `member` - Fields of composite data types (e.g. structs, unions)
       - `private` - Private fields that use a unique syntax (currently just ECMAScript-based languages)
 
-- `label`
+- `label` - `.class`, `#id` in CSS, etc.
 
 - `punctuation`
   - `delimiter` - Commas, colons
@@ -216,7 +229,7 @@ We use a similar set of scopes as
 
 - `namespace`
 
-- `special`
+- `special` - `derive` in Rust, etc.
 
 - `markup`
   - `heading`
@@ -283,7 +296,6 @@ These scopes are used for theming the editor interface:
 | `ui.debug.active`                 | Indicator for the line at which debugging execution is paused at, found in the gutter          |
 | `ui.gutter`                       | Gutter                                                                                         |
 | `ui.gutter.selected`              | Gutter for the line the cursor is on                                                           |
-| `ui.highlight.frameline`          | Line at which debugging execution is paused at                                                 |
 | `ui.linenr`                       | Line numbers                                                                                   |
 | `ui.linenr.selected`              | Line number for the line the cursor is on                                                      |
 | `ui.statusline`                   | Statusline                                                                                     |
@@ -293,24 +305,26 @@ These scopes are used for theming the editor interface:
 | `ui.statusline.select`            | Statusline mode during select mode ([only if `editor.color-modes` is enabled][editor-section]) |
 | `ui.statusline.separator`         | Separator character in statusline                                                              |
 | `ui.bufferline`                   | Style for the buffer line                                                                      |
-| `ui.bufferline.active`            | Style for the active buffer in buffer line                                                        |
+| `ui.bufferline.active`            | Style for the active buffer in buffer line                                                     |
 | `ui.bufferline.background`        | Style for bufferline background                                                                |
 | `ui.popup`                        | Documentation popups (e.g. Space + k)                                                          |
 | `ui.popup.info`                   | Prompt for multiple key options                                                                |
-| `ui.picker.header`                | Column names in pickers with multiple columns                                                  |
-| `ui.picker.header.active`         | The column name in pickers with multiple columns where the cursor is entering into.            |
+| `ui.picker.header`                | Header row area in pickers with multiple columns                                               |
+| `ui.picker.header.column`         | Column names in pickers with multiple columns                                                  |
+| `ui.picker.header.column.active`  | The column name in pickers with multiple columns where the cursor is entering into.            |
 | `ui.window`                       | Borderlines separating splits                                                                  |
 | `ui.help`                         | Description box for commands                                                                   |
 | `ui.text`                         | Default text style, command prompts, popup text, etc.                                          |
 | `ui.text.focus`                   | The currently selected line in the picker                                                      |
 | `ui.text.inactive`                | Same as `ui.text` but when the text is inactive (e.g. suggestions)                             |
 | `ui.text.info`                    | The key: command text in `ui.popup.info` boxes                                                 |
+| `ui.text.directory`               | Directory names in prompt completion                                                           |
 | `ui.virtual.ruler`                | Ruler columns (see the [`editor.rulers` config][editor-section])                               |
 | `ui.virtual.whitespace`           | Visible whitespace characters                                                                  |
 | `ui.virtual.indent-guide`         | Vertical indent width guides                                                                   |
 | `ui.virtual.inlay-hint`           | Default style for inlay hints of all kinds                                                     |
-| `ui.virtual.inlay-hint.parameter` | Style for inlay hints of kind `parameter` (LSPs are not required to set a kind)                |
-| `ui.virtual.inlay-hint.type`      | Style for inlay hints of kind `type` (LSPs are not required to set a kind)                     |
+| `ui.virtual.inlay-hint.parameter` | Style for inlay hints of kind `parameter` (language servers are not required to set a kind)    |
+| `ui.virtual.inlay-hint.type`      | Style for inlay hints of kind `type` (language servers are not required to set a kind)         |
 | `ui.virtual.wrap`                 | Soft-wrap indicator (see the [`editor.soft-wrap` config][editor-section])                      |
 | `ui.virtual.jump-label`           | Style for virtual jump labels                                                                  |
 | `ui.menu`                         | Code and command completion menus                                                              |
@@ -319,6 +333,7 @@ These scopes are used for theming the editor interface:
 | `ui.selection`                    | For selections in the editing area                                                             |
 | `ui.selection.primary`            |                                                                                                |
 | `ui.highlight`                    | Highlighted lines in the picker preview                                                        |
+| `ui.highlight.frameline`          | Line at which debugging execution is paused at                                                 |
 | `ui.cursorline.primary`           | The line of the primary cursor ([if cursorline is enabled][editor-section])                    |
 | `ui.cursorline.secondary`         | The lines of any other cursors ([if cursorline is enabled][editor-section])                    |
 | `ui.cursorcolumn.primary`         | The column of the primary cursor ([if cursorcolumn is enabled][editor-section])                |
@@ -334,5 +349,6 @@ These scopes are used for theming the editor interface:
 | `diagnostic.error`                | Diagnostics error (editing area)                                                               |
 | `diagnostic.unnecessary`          | Diagnostics with unnecessary tag (editing area)                                                |
 | `diagnostic.deprecated`           | Diagnostics with deprecated tag (editing area)                                                 |
+| `tabstop`                         | Snippet placeholder                                                                            |
 
 [editor-section]: ./configuration.md#editor-section

book/src/usage.md

@@ -7,3 +7,27 @@ can be accessed via the command `hx --tutor` or `:tutor`.
 > 💡 Currently, not all functionality is fully documented, please refer to the
 > [key mappings](./keymap.md) list.
 
+## Modes
+
+Helix is a modal editor, meaning it has different modes for different tasks. The main modes are:
+
+* [Normal mode](./keymap.md#normal-mode): For navigation and editing commands. This is the default mode.
+* [Insert mode](./keymap.md#insert-mode): For typing text directly into the document. Access by typing `i` in normal mode.
+* [Select/extend mode](./keymap.md#select--extend-mode): For making selections and performing operations on them. Access by typing `v` in normal mode.
+
+## Buffers
+
+Buffers are in-memory representations of files. You can have multiple buffers open at once. Use [pickers](./pickers.md) or commands like `:buffer-next` and `:buffer-previous` to open buffers or switch between them.
+
+## Selection-first editing
+
+Inspired by [Kakoune](http://kakoune.org/), Helix follows the `selection → action` model. This means that whatever you are going to act on (a word, a paragraph, a line, etc.) is selected first and the action itself (delete, change, yank, etc.) comes second. A cursor is simply a single width selection.
+
+## Multiple selections
+
+Also inspired by Kakoune, multiple selections are a core mode of interaction in Helix. For example, the standard way of replacing multiple instances of a word is to first select all instances (so there is one selection per instance) and then use the change action (`c`) to edit them all at the same time.
+
+## Motions
+
+Motions are commands that move the cursor or modify selections. They're used for navigation and text manipulation. Examples include `w` to move to the next word, or `f` to find a character. See the [Movement](./keymap.md#movement) section of the keymap for more motions.
+

book/theme/index.hbs

@@ -1,5 +1,5 @@
 <!DOCTYPE HTML>
-<html lang="{{ language }}" class="{{ default_theme }}" dir="{{ text_direction }}">
+<html lang="{{ language }}" class="{{ default_theme }} sidebar-visible" dir="{{ text_direction }}">
     <head>
         <!-- Book generated using mdBook -->
         <meta charset="UTF-8">
@@ -52,15 +52,17 @@
         <!-- MathJax -->
         <script async src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.1/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
         {{/if}}
-    </head>
-    <body class="sidebar-visible no-js">
-    <div id="body-container">
+
         <!-- Provide site root to javascript -->
         <script>
             var path_to_root = "{{ path_to_root }}";
             var default_theme = window.matchMedia("(prefers-color-scheme: dark)").matches ? "{{ preferred_dark_theme }}" : "{{ default_theme }}";
         </script>
-
+        <!-- Start loading toc.js asap -->
+        <script src="{{ path_to_root }}toc.js"></script>
+    </head>
+    <body>
+    <div id="body-container">
         <!-- Work around some values being stored in localStorage wrapped in quotes -->
         <script>
             try {
@@ -82,19 +84,16 @@
             var theme;
             try { theme = localStorage.getItem('mdbook-theme'); } catch(e) { }
             if (theme === null || theme === undefined) { theme = default_theme; }
-            var html = document.querySelector('html');
+            const html = document.documentElement;
             html.classList.remove('{{ default_theme }}')
             html.classList.add(theme);
-            var body = document.querySelector('body');
-            body.classList.remove('no-js')
-            body.classList.add('js');
+            html.classList.add("js");
         </script>
 
         <input type="checkbox" id="sidebar-toggle-anchor" class="hidden">
 
         <!-- Hide / unhide sidebar before it is displayed -->
         <script>
-            var body = document.querySelector('body');
             var sidebar = null;
             var sidebar_toggle = document.getElementById("sidebar-toggle-anchor");
             if (document.body.clientWidth >= 1080) {
@@ -104,39 +103,21 @@
                 sidebar = 'hidden';
             }
             sidebar_toggle.checked = sidebar === 'visible';
-            body.classList.remove('sidebar-visible');
-            body.classList.add("sidebar-" + sidebar);
+            html.classList.remove('sidebar-visible');
+            html.classList.add("sidebar-" + sidebar);
         </script>
 
         <nav id="sidebar" class="sidebar" aria-label="Table of contents">
-            <div class="sidebar-scrollbox">
-                {{#toc}}{{/toc}}
+            <!-- populated by js -->
+            <mdbook-sidebar-scrollbox class="sidebar-scrollbox"></mdbook-sidebar-scrollbox>
+            <noscript>
+                <iframe class="sidebar-iframe-outer" src="{{ path_to_root }}toc.html"></iframe>
+            </noscript>
+            <div id="sidebar-resize-handle" class="sidebar-resize-handle">
+                <div class="sidebar-resize-indicator"></div>
             </div>
-            <div id="sidebar-resize-handle" class="sidebar-resize-handle"></div>
         </nav>
 
-        <!-- Track and set sidebar scroll position -->
-        <script>
-            var sidebarScrollbox = document.querySelector('#sidebar .sidebar-scrollbox');
-            sidebarScrollbox.addEventListener('click', function(e) {
-                if (e.target.tagName === 'A') {
-                    sessionStorage.setItem('sidebar-scroll', sidebarScrollbox.scrollTop);
-                }
-            }, { passive: true });
-            var sidebarScrollTop = sessionStorage.getItem('sidebar-scroll');
-            sessionStorage.removeItem('sidebar-scroll');
-            if (sidebarScrollTop) {
-                // preserve sidebar scroll position when navigating via links within sidebar
-                sidebarScrollbox.scrollTop = sidebarScrollTop;
-            } else {
-                // scroll sidebar to current active section when navigating via "next/previous chapter" buttons
-                var activeSection = document.querySelector('#sidebar .active');
-                if (activeSection) {
-                    activeSection.scrollIntoView({ block: 'center' });
-                }
-            }
-        </script>
-
         <div id="page-wrapper" class="page-wrapper">
 
             <div class="page">

contrib/Helix.appdata.xml

@@ -47,6 +47,18 @@
   <content_rating type="oars-1.1" />
 
   <releases>
+    <release version="25.07.1" date="2025-07-18">
+      <url>https://github.com/helix-editor/helix/releases/tag/25.07.1</url>
+    </release>
+    <release version="25.07" date="2025-07-15">
+      <url>https://helix-editor.com/news/release-25-07-highlights/</url>
+    </release>
+    <release version="25.01.1" date="2025-01-19">
+      <url>https://github.com/helix-editor/helix/releases/tag/25.01.1</url>
+    </release>
+    <release version="25.01" date="2025-01-03">
+      <url>https://helix-editor.com/news/release-25-01-highlights/</url>
+    </release>
     <release version="24.07" date="2024-07-14">
       <url>https://github.com/helix-editor/helix/releases/tag/24.07</url>
     </release>

contrib/Helix.desktop

@@ -86,6 +86,6 @@ Keywords[ru]=текст;текстовый редактор;
 Keywords[sr]=Текст;едитор;
 Keywords[tr]=Metin;düzenleyici;
 Icon=helix
-Categories=Utility;TextEditor;
+Categories=Utility;TextEditor;ConsoleOnly
 StartupNotify=false
 MimeType=text/english;text/plain;text/x-makefile;text/x-c++hdr;text/x-c++src;text/x-chdr;text/x-csrc;text/x-java;text/x-moc;text/x-pascal;text/x-tcl;text/x-tex;application/x-shellscript;text/x-c;text/x-c++;

contrib/completion/hx.bash

@@ -2,23 +2,31 @@
 # Bash completion script for Helix editor
 
 _hx() {
-	# $1 command name
-	# $2 word being completed
-	# $3 word preceding
+    local cur prev languages
+    COMPREPLY=()
+    cur="${COMP_WORDS[COMP_CWORD]}"
+    prev="${COMP_WORDS[COMP_CWORD - 1]}"
 
-	case "$3" in
-	-g | --grammar)
-		COMPREPLY="$(compgen -W 'fetch build' -- $2)"
-		;;
-	--health)
-		local languages=$(hx --health |tail -n '+7' |awk '{print $1}' |sed 's/\x1b\[[0-9;]*m//g')
-		COMPREPLY="$(compgen -W """$languages""" -- $2)"
-		;;
-	*)
-		COMPREPLY="$(compgen -fd -W "-h --help --tutor -V --version -v -vv -vvv --health -g --grammar --vsplit --hsplit -c --config --log" -- """$2""")"
-		;;
-	esac
+    case "$prev" in
+    -g | --grammar)
+        mapfile -t COMPREPLY < <(compgen -W 'fetch build' -- "$cur")
+        return 0
+        ;;
+    --health)
+        languages=$(hx --health all-languages | tail -n '+2' | awk '{print $1}' | sed 's/\x1b\[[0-9;]*m//g')
+        mapfile -t COMPREPLY < <(compgen -W """clipboard languages all-languages all $languages""" -- "$cur")
+        return 0
+        ;;
+    esac
 
-	local IFS=$'\n'
-	COMPREPLY=($COMPREPLY)
+    case "$2" in
+    -*)
+        mapfile -t COMPREPLY < <(compgen -W "-h --help --tutor -V --version -v -vv -vvv --health -g --grammar --vsplit --hsplit -c --config --log" -- """$2""")
+        return 0
+        ;;
+    *)
+        mapfile -t COMPREPLY < <(compgen -fd -- """$2""")
+        return 0
+        ;;
+    esac
 } && complete -o filenames -F _hx hx

contrib/completion/hx.elv

@@ -20,7 +20,7 @@ var config = [ "--config" "-c" ]
 set edit:completion:arg-completer[hx] = {|@args|
   var n = (count $args)
   if (>= $n 3) {
-    # Stop completions if passed arg will take presedence
+    # Stop completions if passed arg will take precedence
     # and invalidate further input
     if (has-value $skips $args[-2]) {
       return

contrib/completion/hx.fish

@@ -4,6 +4,10 @@
 complete -c hx -s h -l help -d "Prints help information"
 complete -c hx -l tutor -d "Loads the tutorial"
 complete -c hx -l health -xa "(__hx_langs_ops)" -d "Checks for errors"
+complete -c hx -l health -xka all -d "Prints all diagnostic informations"
+complete -c hx -l health -xka all-languages -d "Lists all languages"
+complete -c hx -l health -xka languages -d "Lists user configured languages"
+complete -c hx -l health -xka clipboard -d "Prints system clipboard provider"
 complete -c hx -s g -l grammar -x -a "fetch build" -d "Fetch or build tree-sitter grammars"
 complete -c hx -s v -o vv -o vvv -d "Increases logging verbosity"
 complete -c hx -s V -l version -d "Prints version information"
@@ -14,5 +18,5 @@ complete -c hx -l log -r -d "Specifies a file to use for logging"
 complete -c hx -s w -l working-dir -d "Specify initial working directory" -xa "(__fish_complete_directories)"
 
 function __hx_langs_ops
-    hx --health languages | tail -n '+2' | string replace -fr '^(\S+) .*' '$1'
+    hx --health all-languages | tail -n '+2' | string replace -fr '^(\S+) .*' '$1'
 end

contrib/completion/hx.nu

@@ -0,0 +1,29 @@
+# Completions for Helix: <https://github.com/helix-editor/helix>
+#
+# NOTE: the `+N` syntax is not supported in Nushell (https://github.com/nushell/nushell/issues/13418)
+#       so it has not been specified here and will not be proposed in the autocompletion of Nushell.
+#       The help message won't be overridden though, so it will still be present here
+
+def health_categories [] {
+    let languages = ^hx --health all-languages | detect columns | get Language | where { $in != null }
+    let completions = [ "all", "clipboard", "languages", "all-languages" ] | append $languages
+    return $completions
+}
+
+def grammar_categories [] { ["fetch", "build"] }
+
+# A post-modern text editor.
+export extern hx [
+    --help(-h),                                 # Prints help information
+    --tutor,                                    # Loads the tutorial
+    --health: string@health_categories,         # Checks for potential errors in editor setup
+    --grammar(-g): string@grammar_categories,   # Fetches or builds tree-sitter grammars listed in `languages.toml`
+    --config(-c): glob,                         # Specifies a file to use for configuration
+    -v,                                         # Increases logging verbosity each use for up to 3 times
+    --log: glob,                                # Specifies a file to use for logging
+    --version(-V),                              # Prints version information
+    --vsplit,                                   # Splits all given files vertically into different windows
+    --hsplit,                                   # Splits all given files horizontally into different windows
+    --working-dir(-w): glob,                    # Specify an initial working directory
+    ...files: glob,                             # Sets the input file to use, position can also be specified via file[:row[:col]]
+]

contrib/completion/hx.zsh

@@ -25,7 +25,7 @@ _hx() {
 
 	case "$state" in
 	health)
-		local languages=($(hx --health | tail -n '+11' | awk '{print $1}' | sed 's/\x1b\[[0-9;]*m//g;s/[✘✓]//g'))
+		local languages=($(hx --health all-languages | tail -n '+2' | awk '{print $1}' | sed 's/\x1b\[[0-9;]*m//g;s/[✘✓]//g'))
 		_values 'language' $languages
 		;;
 	grammar)

contrib/hx_launcher.sh

@@ -0,0 +1,3 @@
+#!/usr/bin/env sh
+
+HELIX_RUNTIME=/usr/lib/helix/runtime exec /usr/lib/helix/hx "$@"

default.nix

@@ -1,8 +1,87 @@
-# Flake's default package for non-flake-enabled nix instances
-let
-  compat = builtins.fetchTarball {
-    url = "https://github.com/edolstra/flake-compat/archive/b4a34015c698c7793d592d66adbab377907a2be8.tar.gz";
-    sha256 = "sha256:1qc703yg0babixi6wshn5wm2kgl5y1drcswgszh4xxzbrwkk9sv7";
-  };
+{
+  lib,
+  rustPlatform,
+  callPackage,
+  runCommand,
+  installShellFiles,
+  git,
+  gitRev ? null,
+  grammarOverlays ? [],
+  includeGrammarIf ? _: true,
+}: let
+  fs = lib.fileset;
+
+  src = fs.difference (fs.gitTracked ./.) (fs.unions [
+    ./.envrc
+    ./rustfmt.toml
+    ./screenshot.png
+    ./book
+    ./docs
+    ./runtime
+    ./flake.lock
+    (fs.fileFilter (file: lib.strings.hasInfix ".git" file.name) ./.)
+    (fs.fileFilter (file: file.hasExt "svg") ./.)
+    (fs.fileFilter (file: file.hasExt "md") ./.)
+    (fs.fileFilter (file: file.hasExt "nix") ./.)
+  ]);
+
+  # Next we actually need to build the grammars and the runtime directory
+  # that they reside in. It is built by calling the derivation in the
+  # grammars.nix file, then taking the runtime directory in the git repo
+  # and hooking symlinks up to it.
+  grammars = callPackage ./grammars.nix {inherit grammarOverlays includeGrammarIf;};
+  runtimeDir = runCommand "helix-runtime" {} ''
+    mkdir -p $out
+    ln -s ${./runtime}/* $out
+    rm -r $out/grammars
+    ln -s ${grammars} $out/grammars
+  '';
 in
-  (import compat {src = ./.;}).defaultNix
+  rustPlatform.buildRustPackage (self: {
+    cargoLock = {
+      lockFile = ./Cargo.lock;
+      # This is not allowed in nixpkgs but is very convenient here: it allows us to
+      # avoid specifying `outputHashes` here for any git dependencies we might take
+      # on temporarily.
+      allowBuiltinFetchGit = true;
+    };
+
+    propagatedBuildInputs = [ runtimeDir ];
+    
+    nativeBuildInputs = [
+      installShellFiles
+      git
+    ];
+
+    buildType = "release";
+
+    name = with builtins; (fromTOML (readFile ./helix-term/Cargo.toml)).package.name;
+    src = fs.toSource {
+      root = ./.;
+      fileset = src;
+    };
+
+    # Helix attempts to reach out to the network and get the grammars. Nix doesn't allow this.
+    HELIX_DISABLE_AUTO_GRAMMAR_BUILD = "1";
+
+    # So Helix knows what rev it is.
+    HELIX_NIX_BUILD_REV = gitRev;
+
+    doCheck = false;
+    strictDeps = true;
+
+    # Sets the Helix runtime dir to the grammars
+    env.HELIX_DEFAULT_RUNTIME = "${runtimeDir}";
+
+    # Get all the application stuff in the output directory.
+    postInstall = ''
+      mkdir -p $out/lib
+      installShellCompletion ${./contrib/completion}/hx.{bash,fish,zsh}
+      mkdir -p $out/share/{applications,icons/hicolor/{256x256,scalable}/apps}
+      cp ${./contrib/Helix.desktop} $out/share/applications/Helix.desktop
+      cp ${./logo.svg} $out/share/icons/hicolor/scalable/apps/helix.svg
+      cp ${./contrib/helix.png} $out/share/icons/hicolor/256x256/apps/helix.png
+    '';
+
+    meta.mainProgram = "hx";
+  })

docs/CONTRIBUTING.md

@@ -13,8 +13,10 @@ Some suggestions to get started:
 - Instead of running a release version of Helix, while developing you may want to run in debug mode with `cargo run` which is way faster to compile
 - Looking for even faster compile times? Give a try to [mold](https://github.com/rui314/mold)
 - If your preferred language is missing, integrating a tree-sitter grammar for
-    it and defining syntax highlight queries for it is straight forward and
+    it and defining syntax highlight queries for it is straightforward and
     doesn't require much knowledge of the internals.
+- If you don't use the Nix development shell and are getting your rust-analyzer binary from rustup, you may need to run `rustup component add rust-analyzer`.
+  This is because `rust-toolchain.toml` selects our MSRV for the development toolchain but doesn't download the matching rust-analyzer automatically.
 
 We provide an [architecture.md][architecture.md] that should give you
 a good overview of the internals.
@@ -53,12 +55,22 @@ Existing tests can be used as examples. Helpers can be found in
 [helpers.rs][helpers.rs]. The log level can be set with the `HELIX_LOG_LEVEL`
 environment variable, e.g. `HELIX_LOG_LEVEL=debug cargo integration-test`.
 
+Contributors using MacOS might encounter `Too many open files (os error 24)`
+failures while running integration tests. This can be resolved by increasing
+the default value (e.g. to `10240` from `256`) by running `ulimit -n 10240`.
+
 ## Minimum Stable Rust Version (MSRV) Policy
 
-Helix follows the MSRV of Firefox.
-The current MSRV and future changes to the MSRV are listed in the [Firefox documentation].
+Helix keeps an intentionally low MSRV for the sake of easy building and packaging
+downstream. We follow [Firefox's MSRV policy]. Once Firefox's MSRV increases we
+may bump ours as well, but be sure to check that popular distributions like Ubuntu
+package the new MSRV version. When increasing the MSRV, update these three places:
 
-[Firefox documentation]: https://firefox-source-docs.mozilla.org/writing-rust-code/update-policy.html
+* the `workspace.package.rust-version` key in `Cargo.toml` in the repository root
+* the `env.MSRV` key at the top of `.github/workflows/build.yml`
+* the `toolchain.channel` key in `rust-toolchain.toml`
+
+[Firefox's MSRV policy]: https://firefox-source-docs.mozilla.org/writing-rust-code/update-policy.html
 [good-first-issue]: https://github.com/helix-editor/helix/labels/E-easy
 [log-file]: https://github.com/helix-editor/helix/wiki/FAQ#access-the-log-file
 [architecture.md]: ./architecture.md

docs/architecture.md

@@ -1,13 +1,16 @@
 
-| Crate        | Description                                                      |
-| -----------  | -----------                                                      |
-| helix-core   | Core editing primitives, functional.                             |
-| helix-lsp    | Language server client                                           |
-| helix-dap    | Debug Adapter Protocol (DAP) client                              |
-| helix-loader | Functions for building, fetching, and loading external resources |
-| helix-view   | UI abstractions for use in backends, imperative shell.           |
-| helix-term   | Terminal UI                                                      |
-| helix-tui    | TUI primitives, forked from tui-rs, inspired by Cursive          |
+| Crate           | Description                                                      |
+| -----------     | -----------                                                      |
+| helix-stdx      | Extensions to the standard library (similar to [`rust-analyzer`'s](https://github.com/rust-lang/rust-analyzer/blob/ea413f67a8f730b4211c09e103f8207c62e7dbc3/crates/stdx/Cargo.toml#L5)) |
+| helix-core      | Core editing primitives, functional.                             |
+| helix-lsp       | Language server client                                           |
+| helix-lsp-types | Language Server Protocol type definitions                        |
+| helix-dap       | Debug Adapter Protocol (DAP) client                              |
+| helix-event     | Primitives for defining and handling events within the editor    |
+| helix-loader    | Functions for building, fetching, and loading external resources |
+| helix-view      | UI abstractions for use in backends, imperative shell.           |
+| helix-term      | Terminal UI                                                      |
+| helix-tui       | TUI primitives, forked from tui-rs, inspired by Cursive          |
 
 
 This document contains a high-level overview of Helix internals.
@@ -109,3 +112,17 @@ The `main` function sets up a new `Application` that runs the event loop.
 ## TUI / Term
 
 TODO: document Component and rendering related stuff
+
+## Event
+
+The `helix-event` crate defines primitives for defining and acting on events
+within the editor. "Events" cover things like opening, changing and closing of
+documents, starting and stopping of language servers and more.
+
+`helix-event` has tools for defining events and registering _hooks_ which run
+any time an event is emitted. `helix-event` also provides `AsyncHook` - a tool
+for running cancellable tasks which run after events with _debouncing_.
+
+See the `AsyncHook` type for more information. Events can be created within the
+`events!` macro. Synchronous hooks can be created with `register_hook!`. And
+editor-wide events can be sent to hooks with `helix_event::dispatch`.

docs/releases.md

@@ -5,19 +5,18 @@ Helix releases are versioned in the Calendar Versioning scheme:
 `22.05.1`. In these instructions we'll use `<tag>` as a placeholder for the tag
 being published.
 
-* Merge the changelog PR
-* Add new `<release>` entry in `contrib/Helix.appdata.xml` with release information according to the [AppStream spec](https://www.freedesktop.org/software/appstream/docs/sect-Metadata-Releases.html)
+* Merge the PR with the release updates. That branch should:
+    * Update the version:
+        * Update the `workspace.package.version` key in `Cargo.toml`. Cargo only accepts
+          SemVer versions so a CalVer version of `22.07` for example must be formatted
+          as `22.7.0`. Patch/bugfix releases should increment the SemVer patch number. A
+          patch release for 22.07 would be `22.7.1`.
+        * Run `cargo check` and commit the resulting change to `Cargo.lock`
+    * Add changelog notes to `CHANGELOG.md`
+    * Add new `<release>` entry in `contrib/Helix.appdata.xml` with release information according to the [AppStream spec](https://www.freedesktop.org/software/appstream/docs/sect-Metadata-Releases.html)
 * Tag and push
-    * `git tag -s -m "<tag>" -a <tag> && git push`
-    * Make sure to switch to master and pull first
-* Edit the `Cargo.toml` file and change the date in the `version` field to the next planned release
-    * Due to Cargo having a strict requirement on SemVer with 3 or more version
-      numbers, a `0` is required in the micro version; however, unless we are
-      publishing a patch release after a major release, the `.0` is dropped in
-      the user facing version.
-    * Releases are planned to happen every two months, so `22.05.0` would change to `22.07.0`
-    * If we are pushing a patch/bugfix release in the same month as the previous
-      release, bump the micro version, e.g. `22.07.0` to `22.07.1`
+    * Switch to master and pull
+    * `git tag -s -m "<tag>" -a <tag> && git push origin <tag>` (note the `-s` which signs the tag)
 * Wait for the Release CI to finish
     * It will automatically turn the git tag into a GitHub release when it uploads artifacts
 * Edit the new release

flake.lock

@@ -1,50 +1,12 @@
 {
   "nodes": {
-    "crane": {
-      "inputs": {
-        "nixpkgs": [
-          "nixpkgs"
-        ]
-      },
-      "locked": {
-        "lastModified": 1709610799,
-        "narHash": "sha256-5jfLQx0U9hXbi2skYMGodDJkIgffrjIOgMRjZqms2QE=",
-        "owner": "ipetkov",
-        "repo": "crane",
-        "rev": "81c393c776d5379c030607866afef6406ca1be57",
-        "type": "github"
-      },
-      "original": {
-        "owner": "ipetkov",
-        "repo": "crane",
-        "type": "github"
-      }
-    },
-    "flake-utils": {
-      "inputs": {
-        "systems": "systems"
-      },
-      "locked": {
-        "lastModified": 1709126324,
-        "narHash": "sha256-q6EQdSeUZOG26WelxqkmR7kArjgWCdw5sfJVHPH/7j8=",
-        "owner": "numtide",
-        "repo": "flake-utils",
-        "rev": "d465f4819400de7c8d874d50b982301f28a84605",
-        "type": "github"
-      },
-      "original": {
-        "owner": "numtide",
-        "repo": "flake-utils",
-        "type": "github"
-      }
-    },
     "nixpkgs": {
       "locked": {
-        "lastModified": 1709479366,
-        "narHash": "sha256-n6F0n8UV6lnTZbYPl1A9q1BS0p4hduAv1mGAP17CVd0=",
+        "lastModified": 1740560979,
+        "narHash": "sha256-Vr3Qi346M+8CjedtbyUevIGDZW8LcA1fTG0ugPY/Hic=",
         "owner": "nixos",
         "repo": "nixpkgs",
-        "rev": "b8697e57f10292a6165a20f03d2f42920dfaf973",
+        "rev": "5135c59491985879812717f4c9fea69604e7f26f",
         "type": "github"
       },
       "original": {
@@ -56,27 +18,22 @@
     },
     "root": {
       "inputs": {
-        "crane": "crane",
-        "flake-utils": "flake-utils",
         "nixpkgs": "nixpkgs",
         "rust-overlay": "rust-overlay"
       }
     },
     "rust-overlay": {
       "inputs": {
-        "flake-utils": [
-          "flake-utils"
-        ],
         "nixpkgs": [
           "nixpkgs"
         ]
       },
       "locked": {
-        "lastModified": 1709604635,
-        "narHash": "sha256-le4fwmWmjGRYWwkho0Gr7mnnZndOOe4XGbLw68OvF40=",
+        "lastModified": 1740623427,
+        "narHash": "sha256-3SdPQrZoa4odlScFDUHd4CUPQ/R1gtH4Mq9u8CBiK8M=",
         "owner": "oxalica",
         "repo": "rust-overlay",
-        "rev": "e86c0fb5d3a22a5f30d7f64ecad88643fe26449d",
+        "rev": "d342e8b5fd88421ff982f383c853f0fc78a847ab",
         "type": "github"
       },
       "original": {
@@ -84,21 +41,6 @@
         "repo": "rust-overlay",
         "type": "github"
       }
-    },
-    "systems": {
-      "locked": {
-        "lastModified": 1681028828,
-        "narHash": "sha256-Vy1rq5AaRuLzOxct8nz4T6wlgyUR7zLU309k9mBC768=",
-        "owner": "nix-systems",
-        "repo": "default",
-        "rev": "da67096a3b9bf56a91d16901293e51ba5b49a27e",
-        "type": "github"
-      },
-      "original": {
-        "owner": "nix-systems",
-        "repo": "default",
-        "type": "github"
-      }
     }
   },
   "root": "root",

flake.nix

@@ -3,16 +3,8 @@
 
   inputs = {
     nixpkgs.url = "github:nixos/nixpkgs/nixos-unstable";
-    flake-utils.url = "github:numtide/flake-utils";
     rust-overlay = {
       url = "github:oxalica/rust-overlay";
-      inputs = {
-        nixpkgs.follows = "nixpkgs";
-        flake-utils.follows = "flake-utils";
-      };
-    };
-    crane = {
-      url = "github:ipetkov/crane";
       inputs.nixpkgs.follows = "nixpkgs";
     };
   };
@@ -20,176 +12,89 @@
   outputs = {
     self,
     nixpkgs,
-    crane,
-    flake-utils,
     rust-overlay,
     ...
-  }:
-    flake-utils.lib.eachDefaultSystem (system: let
-      pkgs = import nixpkgs {
-        inherit system;
-        overlays = [(import rust-overlay)];
-      };
-      mkRootPath = rel:
-        builtins.path {
-          path = "${toString ./.}/${rel}";
-          name = rel;
+  }: let
+    inherit (nixpkgs) lib;
+    systems = [
+      "x86_64-linux"
+      "aarch64-linux"
+      "x86_64-darwin"
+      "aarch64-darwin"
+    ];
+    eachSystem = lib.genAttrs systems;
+    pkgsFor = eachSystem (system:
+      import nixpkgs {
+        localSystem.system = system;
+        overlays = [(import rust-overlay) self.overlays.helix];
+      });
+    gitRev = self.rev or self.dirtyRev or null;
+  in {
+    packages = eachSystem (system: {
+      inherit (pkgsFor.${system}) helix;
+      /*
+      The default Helix build. Uses the latest stable Rust toolchain, and unstable
+      nixpkgs.
+
+      The build inputs can be overridden with the following:
+
+      packages.${system}.default.override { rustPlatform = newPlatform; };
+
+      Overriding a derivation attribute can be done as well:
+
+      packages.${system}.default.overrideAttrs { buildType = "debug"; };
+      */
+      default = self.packages.${system}.helix;
+    });
+    checks =
+      lib.mapAttrs (system: pkgs: let
+        # Get Helix's MSRV toolchain to build with by default.
+        msrvToolchain = pkgs.pkgsBuildHost.rust-bin.fromRustupToolchainFile ./rust-toolchain.toml;
+        msrvPlatform = pkgs.makeRustPlatform {
+          cargo = msrvToolchain;
+          rustc = msrvToolchain;
         };
-      filteredSource = let
-        pathsToIgnore = [
-          ".envrc"
-          ".ignore"
-          ".github"
-          ".gitignore"
-          "logo_dark.svg"
-          "logo_light.svg"
-          "rust-toolchain.toml"
-          "rustfmt.toml"
-          "runtime"
-          "screenshot.png"
-          "book"
-          "docs"
-          "README.md"
-          "CHANGELOG.md"
-          "shell.nix"
-          "default.nix"
-          "grammars.nix"
-          "flake.nix"
-          "flake.lock"
-        ];
-        ignorePaths = path: type: let
-          inherit (nixpkgs) lib;
-          # split the nix store path into its components
-          components = lib.splitString "/" path;
-          # drop off the `/nix/hash-source` section from the path
-          relPathComponents = lib.drop 4 components;
-          # reassemble the path components
-          relPath = lib.concatStringsSep "/" relPathComponents;
+      in {
+        helix = self.packages.${system}.helix.override {
+          rustPlatform = msrvPlatform;
+        };
+      })
+      pkgsFor;
+
+    # Devshell behavior is preserved.
+    devShells =
+      lib.mapAttrs (system: pkgs: {
+        default = let
+          commonRustFlagsEnv = "-C link-arg=-fuse-ld=lld -C target-cpu=native --cfg tokio_unstable";
+          platformRustFlagsEnv = lib.optionalString pkgs.stdenv.isLinux "-Clink-arg=-Wl,--no-rosegment";
         in
-          lib.all (p: ! (lib.hasPrefix p relPath)) pathsToIgnore;
-      in
-        builtins.path {
-          name = "helix-source";
-          path = toString ./.;
-          # filter out unnecessary paths
-          filter = ignorePaths;
-        };
-      makeOverridableHelix = old: config: let
-        grammars = pkgs.callPackage ./grammars.nix config;
-        runtimeDir = pkgs.runCommand "helix-runtime" {} ''
-          mkdir -p $out
-          ln -s ${mkRootPath "runtime"}/* $out
-          rm -r $out/grammars
-          ln -s ${grammars} $out/grammars
-        '';
-        helix-wrapped =
-          pkgs.runCommand
-          old.name
-          {
-            inherit (old) pname version;
-            meta = old.meta or {};
-            passthru =
-              (old.passthru or {})
-              // {
-                unwrapped = old;
-              };
-            nativeBuildInputs = [pkgs.makeWrapper];
-            makeWrapperArgs = config.makeWrapperArgs or [];
-          }
-          ''
-            cp -rs --no-preserve=mode,ownership ${old} $out
-            wrapProgram "$out/bin/hx" ''${makeWrapperArgs[@]} --set HELIX_RUNTIME "${runtimeDir}"
-          '';
-      in
-        helix-wrapped
-        // {
-          override = makeOverridableHelix old;
-          passthru =
-            helix-wrapped.passthru
-            // {
-              wrapper = old: makeOverridableHelix old config;
-            };
-        };
-      stdenv =
-        if pkgs.stdenv.isLinux
-        then pkgs.stdenv
-        else pkgs.clangStdenv;
-      rustFlagsEnv = pkgs.lib.optionalString stdenv.isLinux "-C link-arg=-fuse-ld=lld -C target-cpu=native -Clink-arg=-Wl,--no-rosegment";
-      rustToolchain = pkgs.pkgsBuildHost.rust-bin.fromRustupToolchainFile ./rust-toolchain.toml;
-      craneLibMSRV = (crane.mkLib pkgs).overrideToolchain rustToolchain;
-      craneLibStable = (crane.mkLib pkgs).overrideToolchain pkgs.pkgsBuildHost.rust-bin.stable.latest.default;
-      commonArgs = {
-        inherit stdenv;
-        inherit (craneLibMSRV.crateNameFromCargoToml {cargoToml = ./helix-term/Cargo.toml;}) pname;
-        inherit (craneLibMSRV.crateNameFromCargoToml {cargoToml = ./Cargo.toml;}) version;
-        src = filteredSource;
-        # disable fetching and building of tree-sitter grammars in the helix-term build.rs
-        HELIX_DISABLE_AUTO_GRAMMAR_BUILD = "1";
-        buildInputs = [stdenv.cc.cc.lib];
-        # disable tests
-        doCheck = false;
-        meta.mainProgram = "hx";
-      };
-      cargoArtifacts = craneLibMSRV.buildDepsOnly commonArgs;
-    in {
-      packages = {
-        helix-unwrapped = craneLibStable.buildPackage (commonArgs
-          // {
-            cargoArtifacts = craneLibStable.buildDepsOnly commonArgs;
-            postInstall = ''
-              mkdir -p $out/share/applications $out/share/icons/hicolor/scalable/apps $out/share/icons/hicolor/256x256/apps
-              cp contrib/Helix.desktop $out/share/applications
-              cp logo.svg $out/share/icons/hicolor/scalable/apps/helix.svg
-              cp contrib/helix.png $out/share/icons/hicolor/256x256/apps
+          pkgs.mkShell {
+            inputsFrom = [self.checks.${system}.helix];
+            nativeBuildInputs = with pkgs;
+              [
+                lld
+                cargo-flamegraph
+                rust-bin.nightly.latest.rust-analyzer
+              ]
+              ++ (lib.optional (stdenv.isx86_64 && stdenv.isLinux) cargo-tarpaulin)
+              ++ (lib.optional stdenv.isLinux lldb)
+              ++ (lib.optional stdenv.isDarwin darwin.apple_sdk.frameworks.CoreFoundation);
+            shellHook = ''
+              export RUST_BACKTRACE="1"
+              export RUSTFLAGS="''${RUSTFLAGS:-""} ${commonRustFlagsEnv} ${platformRustFlagsEnv}"
             '';
-          });
-        helix = makeOverridableHelix self.packages.${system}.helix-unwrapped {};
-        default = self.packages.${system}.helix;
+          };
+      })
+      pkgsFor;
+
+    overlays = {
+      helix = final: prev: {
+        helix = final.callPackage ./default.nix {inherit gitRev;};
       };
 
-      checks = {
-        # Build the crate itself
-        inherit (self.packages.${system}) helix;
-
-        clippy = craneLibMSRV.cargoClippy (commonArgs
-          // {
-            inherit cargoArtifacts;
-            cargoClippyExtraArgs = "--all-targets -- --deny warnings";
-          });
-
-        fmt = craneLibMSRV.cargoFmt commonArgs;
-
-        doc = craneLibMSRV.cargoDoc (commonArgs
-          // {
-            inherit cargoArtifacts;
-          });
-
-        test = craneLibMSRV.cargoTest (commonArgs
-          // {
-            inherit cargoArtifacts;
-          });
-      };
-
-      devShells.default = pkgs.mkShell {
-        inputsFrom = builtins.attrValues self.checks.${system};
-        nativeBuildInputs = with pkgs;
-          [lld_13 cargo-flamegraph rust-analyzer]
-          ++ (lib.optional (stdenv.isx86_64 && stdenv.isLinux) pkgs.cargo-tarpaulin)
-          ++ (lib.optional stdenv.isLinux pkgs.lldb)
-          ++ (lib.optional stdenv.isDarwin pkgs.darwin.apple_sdk.frameworks.CoreFoundation);
-        shellHook = ''
-          export HELIX_RUNTIME="$PWD/runtime"
-          export RUST_BACKTRACE="1"
-          export RUSTFLAGS="''${RUSTFLAGS:-""} ${rustFlagsEnv}"
-        '';
-      };
-    })
-    // {
-      overlays.default = final: prev: {
-        inherit (self.packages.${final.system}) helix;
-      };
+      default = self.overlays.helix;
     };
-
+  };
   nixConfig = {
     extra-substituters = ["https://helix.cachix.org"];
     extra-trusted-public-keys = ["helix.cachix.org-1:ejp9KQpR1FBI2onstMQ34yogDm4OgU2ru6lIwPvuCVs="];

grammars.nix

@@ -1,22 +1,13 @@
 {
   stdenv,
   lib,
-  runCommandLocal,
   runCommand,
-  yj,
   includeGrammarIf ? _: true,
   grammarOverlays ? [],
   ...
 }: let
-  # HACK: nix < 2.6 has a bug in the toml parser, so we convert to JSON
-  # before parsing
-  languages-json = runCommandLocal "languages-toml-to-json" {} ''
-    ${yj}/bin/yj -t < ${./languages.toml} > $out
-  '';
   languagesConfig =
-    if lib.versionAtLeast builtins.nixVersion "2.6.0"
-    then builtins.fromTOML (builtins.readFile ./languages.toml)
-    else builtins.fromJSON (builtins.readFile (builtins.toPath languages-json));
+    builtins.fromTOML (builtins.readFile ./languages.toml);
   isGitGrammar = grammar:
     builtins.hasAttr "source" grammar
     && builtins.hasAttr "git" grammar.source
@@ -32,10 +23,10 @@
   # If `use-grammars.except` is set, use all other grammars.
   # Otherwise use all grammars.
   useGrammar = grammar:
-    if languagesConfig?use-grammars.only then
-      builtins.elem grammar.name languagesConfig.use-grammars.only
-    else if languagesConfig?use-grammars.except then
-      !(builtins.elem grammar.name languagesConfig.use-grammars.except)
+    if languagesConfig ? use-grammars.only
+    then builtins.elem grammar.name languagesConfig.use-grammars.only
+    else if languagesConfig ? use-grammars.except
+    then !(builtins.elem grammar.name languagesConfig.use-grammars.except)
     else true;
   grammarsToUse = builtins.filter useGrammar languagesConfig.grammar;
   gitGrammars = builtins.filter isGitGrammar grammarsToUse;
@@ -66,10 +57,10 @@
       version = grammar.source.rev;
 
       src = source;
-      sourceRoot = if builtins.hasAttr "subpath" grammar.source then
-        "source/${grammar.source.subpath}"
-      else
-        "source";
+      sourceRoot =
+        if builtins.hasAttr "subpath" grammar.source
+        then "source/${grammar.source.subpath}"
+        else "source";
 
       dontConfigure = true;
 
@@ -96,8 +87,6 @@
         $CC -c src/parser.c -o parser.o $FLAGS
         $CXX -shared -o $NAME.so *.o
 
-        ls -al
-
         runHook postBuild
       '';
 
@@ -116,17 +105,21 @@
       '';
     };
   grammarsToBuild = builtins.filter includeGrammarIf gitGrammars;
-  builtGrammars = builtins.map (grammar: {
-    inherit (grammar) name;
-    value = buildGrammar grammar;
-  }) grammarsToBuild;
+  builtGrammars =
+    builtins.map (grammar: {
+      inherit (grammar) name;
+      value = buildGrammar grammar;
+    })
+    grammarsToBuild;
   extensibleGrammars =
     lib.makeExtensible (self: builtins.listToAttrs builtGrammars);
-  overlayedGrammars = lib.pipe extensibleGrammars
+  overlaidGrammars =
+    lib.pipe extensibleGrammars
     (builtins.map (overlay: grammar: grammar.extend overlay) grammarOverlays);
-  grammarLinks = lib.mapAttrsToList
+  grammarLinks =
+    lib.mapAttrsToList
     (name: artifact: "ln -s ${artifact}/${name}.so $out/${name}.so")
-    (lib.filterAttrs (n: v: lib.isDerivation v) overlayedGrammars);
+    (lib.filterAttrs (n: v: lib.isDerivation v) overlaidGrammars);
 in
   runCommand "consolidated-helix-grammars" {} ''
     mkdir -p $out

helix-core/Cargo.toml

@@ -18,42 +18,46 @@ integration = []
 [dependencies]
 helix-stdx = { path = "../helix-stdx" }
 helix-loader = { path = "../helix-loader" }
+helix-parsec = { path = "../helix-parsec" }
 
-ropey = { version = "1.6.1", default-features = false, features = ["simd"] }
-smallvec = "1.13"
+ropey.workspace = true
+smallvec = "1.15"
 smartstring = "1.0.1"
-unicode-segmentation = "1.11"
-unicode-width = "0.1"
-unicode-general-category = "0.6"
+unicode-segmentation.workspace = true
+# unicode-width is changing width definitions
+# that both break our logic and disagree with common
+# width definitions in terminals, we need to replace it.
+# For now lets lock the version to avoid rendering glitches
+# when installing without `--locked`
+unicode-width = "=0.1.12"
+unicode-general-category = "1.0"
 slotmap.workspace = true
-tree-sitter.workspace = true
-once_cell = "1.19"
+tree-house.workspace = true
+once_cell = "1.21"
 arc-swap = "1"
 regex = "1"
-bitflags = "2.6"
-ahash = "0.8.11"
-hashbrown = { version = "0.14.5", features = ["raw"] }
-dunce = "1.0"
-url = "2.5.0"
+bitflags.workspace = true
+foldhash.workspace = true
+url = "2.5.4"
 
 log = "0.4"
+anyhow = "1.0"
 serde = { version = "1.0", features = ["derive"] }
 serde_json = "1.0"
-toml = "0.8"
-
-imara-diff = "0.1.6"
+toml.workspace = true
 
+imara-diff =  "0.2.0"
 encoding_rs = "0.8"
 
 chrono = { version = "0.4", default-features = false, features = ["alloc", "std"] }
 
-etcetera = "0.8"
-textwrap = "0.16.1"
+textwrap = "0.16.2"
 
 nucleo.workspace = true
-parking_lot = "0.12"
-globset = "0.4.14"
+parking_lot.workspace = true
+globset = "0.4.16"
+regex-cursor = "0.1.5"
 
 [dev-dependencies]
 quickcheck = { version = "1", default-features = false }
-indoc = "2.0.5"
+indoc = "2.0.6"

helix-core/src/auto_pairs.rs

@@ -75,9 +75,9 @@ impl From<(&char, &char)> for Pair {
 
 impl AutoPairs {
     /// Make a new AutoPairs set with the given pairs and default conditions.
-    pub fn new<'a, V: 'a, A>(pairs: V) -> Self
+    pub fn new<'a, V, A>(pairs: V) -> Self
     where
-        V: IntoIterator<Item = A>,
+        V: IntoIterator<Item = A> + 'a,
         A: Into<Pair>,
     {
         let mut auto_pairs = HashMap::new();

helix-core/src/case_conversion.rs

@@ -0,0 +1,69 @@
+use crate::Tendril;
+
+// todo: should this be grapheme aware?
+
+pub fn to_pascal_case(text: impl Iterator<Item = char>) -> Tendril {
+    let mut res = Tendril::new();
+    to_pascal_case_with(text, &mut res);
+    res
+}
+
+pub fn to_pascal_case_with(text: impl Iterator<Item = char>, buf: &mut Tendril) {
+    let mut at_word_start = true;
+    for c in text {
+        // we don't count _ as a word char here so case conversions work well
+        if !c.is_alphanumeric() {
+            at_word_start = true;
+            continue;
+        }
+        if at_word_start {
+            at_word_start = false;
+            buf.extend(c.to_uppercase());
+        } else {
+            buf.push(c)
+        }
+    }
+}
+
+pub fn to_upper_case_with(text: impl Iterator<Item = char>, buf: &mut Tendril) {
+    for c in text {
+        for c in c.to_uppercase() {
+            buf.push(c)
+        }
+    }
+}
+
+pub fn to_lower_case_with(text: impl Iterator<Item = char>, buf: &mut Tendril) {
+    for c in text {
+        for c in c.to_lowercase() {
+            buf.push(c)
+        }
+    }
+}
+
+pub fn to_camel_case(text: impl Iterator<Item = char>) -> Tendril {
+    let mut res = Tendril::new();
+    to_camel_case_with(text, &mut res);
+    res
+}
+pub fn to_camel_case_with(mut text: impl Iterator<Item = char>, buf: &mut Tendril) {
+    for c in &mut text {
+        if c.is_alphanumeric() {
+            buf.extend(c.to_lowercase())
+        }
+    }
+    let mut at_word_start = false;
+    for c in text {
+        // we don't count _ as a word char here so case conversions work well
+        if !c.is_alphanumeric() {
+            at_word_start = true;
+            continue;
+        }
+        if at_word_start {
+            at_word_start = false;
+            buf.extend(c.to_uppercase());
+        } else {
+            buf.push(c)
+        }
+    }
+}

helix-core/src/comment.rs

@@ -4,11 +4,30 @@
 use smallvec::SmallVec;
 
 use crate::{
-    syntax::BlockCommentToken, Change, Range, Rope, RopeSlice, Selection, Tendril, Transaction,
+    syntax::config::BlockCommentToken, Change, Range, Rope, RopeSlice, Selection, Tendril,
+    Transaction,
 };
 use helix_stdx::rope::RopeSliceExt;
 use std::borrow::Cow;
 
+pub const DEFAULT_COMMENT_TOKEN: &str = "#";
+
+/// Returns the longest matching comment token of the given line (if it exists).
+pub fn get_comment_token<'a, S: AsRef<str>>(
+    text: RopeSlice,
+    tokens: &'a [S],
+    line_num: usize,
+) -> Option<&'a str> {
+    let line = text.line(line_num);
+    let start = line.first_non_whitespace_char()?;
+
+    tokens
+        .iter()
+        .map(AsRef::as_ref)
+        .filter(|token| line.slice(start..).starts_with(token))
+        .max_by_key(|token| token.len())
+}
+
 /// Given text, a comment token, and a set of line indices, returns the following:
 /// - Whether the given lines should be considered commented
 ///     - If any of the lines are uncommented, all lines are considered as such.
@@ -28,21 +47,20 @@ fn find_line_comment(
     let mut min = usize::MAX; // minimum col for first_non_whitespace_char
     let mut margin = 1;
     let token_len = token.chars().count();
+
     for line in lines {
         let line_slice = text.line(line);
         if let Some(pos) = line_slice.first_non_whitespace_char() {
             let len = line_slice.len_chars();
 
-            if pos < min {
-                min = pos;
-            }
+            min = std::cmp::min(min, pos);
 
             // line can be shorter than pos + token len
             let fragment = Cow::from(line_slice.slice(pos..std::cmp::min(pos + token.len(), len)));
 
+            // as soon as one of the non-blank lines doesn't have a comment, the whole block is
+            // considered uncommented.
             if fragment != token {
-                // as soon as one of the non-blank lines doesn't have a comment, the whole block is
-                // considered uncommented.
                 commented = false;
             }
 
@@ -56,6 +74,7 @@ fn find_line_comment(
             to_change.push(line);
         }
     }
+
     (commented, to_change, min, margin)
 }
 
@@ -63,7 +82,7 @@ fn find_line_comment(
 pub fn toggle_line_comments(doc: &Rope, selection: &Selection, token: Option<&str>) -> Transaction {
     let text = doc.slice(..);
 
-    let token = token.unwrap_or("//");
+    let token = token.unwrap_or(DEFAULT_COMMENT_TOKEN);
     let comment = Tendril::from(format!("{} ", token));
 
     let mut lines: Vec<usize> = Vec::with_capacity(selection.len());
@@ -129,10 +148,7 @@ pub fn find_block_comments(
     let mut only_whitespace = true;
     let mut comment_changes = Vec::with_capacity(selection.len());
     let default_tokens = tokens.first().cloned().unwrap_or_default();
-    // TODO: check if this can be removed on MSRV bump
-    #[allow(clippy::redundant_clone)]
     let mut start_token = default_tokens.start.clone();
-    #[allow(clippy::redundant_clone)]
     let mut end_token = default_tokens.end.clone();
 
     let mut tokens = tokens.to_vec();
@@ -189,13 +205,9 @@ pub fn find_block_comments(
                     range: *range,
                     start_pos,
                     end_pos,
-                    start_margin: selection_slice
-                        .get_char(after_start)
-                        .map_or(false, |c| c == ' '),
+                    start_margin: selection_slice.get_char(after_start) == Some(' '),
                     end_margin: after_start != before_end
-                        && selection_slice
-                            .get_char(before_end)
-                            .map_or(false, |c| c == ' '),
+                        && (selection_slice.get_char(before_end) == Some(' ')),
                     start_token: start_token.to_string(),
                     end_token: end_token.to_string(),
                 });
@@ -317,56 +329,87 @@ pub fn split_lines_of_selection(text: RopeSlice, selection: &Selection) -> Selec
 mod test {
     use super::*;
 
-    #[test]
-    fn test_find_line_comment() {
-        // four lines, two space indented, except for line 1 which is blank.
-        let mut doc = Rope::from("  1\n\n  2\n  3");
-        // select whole document
-        let mut selection = Selection::single(0, doc.len_chars() - 1);
+    mod find_line_comment {
+        use super::*;
 
-        let text = doc.slice(..);
+        #[test]
+        fn not_commented() {
+            // four lines, two space indented, except for line 1 which is blank.
+            let doc = Rope::from("  1\n\n  2\n  3");
 
-        let res = find_line_comment("//", text, 0..3);
-        // (commented = true, to_change = [line 0, line 2], min = col 2, margin = 0)
-        assert_eq!(res, (false, vec![0, 2], 2, 0));
+            let text = doc.slice(..);
 
-        // comment
-        let transaction = toggle_line_comments(&doc, &selection, None);
-        transaction.apply(&mut doc);
-        selection = selection.map(transaction.changes());
+            let res = find_line_comment("//", text, 0..3);
+            // (commented = false, to_change = [line 0, line 2], min = col 2, margin = 0)
+            assert_eq!(res, (false, vec![0, 2], 2, 0));
+        }
 
-        assert_eq!(doc, "  // 1\n\n  // 2\n  // 3");
+        #[test]
+        fn is_commented() {
+            // three lines where the second line is empty.
+            let doc = Rope::from("// hello\n\n// there");
 
-        // uncomment
-        let transaction = toggle_line_comments(&doc, &selection, None);
-        transaction.apply(&mut doc);
-        selection = selection.map(transaction.changes());
-        assert_eq!(doc, "  1\n\n  2\n  3");
-        assert!(selection.len() == 1); // to ignore the selection unused warning
+            let res = find_line_comment("//", doc.slice(..), 0..3);
 
-        // 0 margin comments
-        doc = Rope::from("  //1\n\n  //2\n  //3");
-        // reset the selection.
-        selection = Selection::single(0, doc.len_chars() - 1);
+            // (commented = true, to_change = [line 0, line 2], min = col 0, margin = 1)
+            assert_eq!(res, (true, vec![0, 2], 0, 1));
+        }
+    }
 
-        let transaction = toggle_line_comments(&doc, &selection, None);
-        transaction.apply(&mut doc);
-        selection = selection.map(transaction.changes());
-        assert_eq!(doc, "  1\n\n  2\n  3");
-        assert!(selection.len() == 1); // to ignore the selection unused warning
+    // TODO: account for uncommenting with uneven comment indentation
+    mod toggle_line_comment {
+        use super::*;
 
-        // 0 margin comments, with no space
-        doc = Rope::from("//");
-        // reset the selection.
-        selection = Selection::single(0, doc.len_chars() - 1);
+        #[test]
+        fn comment() {
+            // four lines, two space indented, except for line 1 which is blank.
+            let mut doc = Rope::from("  1\n\n  2\n  3");
+            // select whole document
+            let selection = Selection::single(0, doc.len_chars() - 1);
 
-        let transaction = toggle_line_comments(&doc, &selection, None);
-        transaction.apply(&mut doc);
-        selection = selection.map(transaction.changes());
-        assert_eq!(doc, "");
-        assert!(selection.len() == 1); // to ignore the selection unused warning
+            let transaction = toggle_line_comments(&doc, &selection, None);
+            transaction.apply(&mut doc);
 
-        // TODO: account for uncommenting with uneven comment indentation
+            assert_eq!(doc, "  # 1\n\n  # 2\n  # 3");
+        }
+
+        #[test]
+        fn uncomment() {
+            let mut doc = Rope::from("  # 1\n\n  # 2\n  # 3");
+            let mut selection = Selection::single(0, doc.len_chars() - 1);
+
+            let transaction = toggle_line_comments(&doc, &selection, None);
+            transaction.apply(&mut doc);
+            selection = selection.map(transaction.changes());
+
+            assert_eq!(doc, "  1\n\n  2\n  3");
+            assert!(selection.len() == 1); // to ignore the selection unused warning
+        }
+
+        #[test]
+        fn uncomment_0_margin_comments() {
+            let mut doc = Rope::from("  #1\n\n  #2\n  #3");
+            let mut selection = Selection::single(0, doc.len_chars() - 1);
+
+            let transaction = toggle_line_comments(&doc, &selection, None);
+            transaction.apply(&mut doc);
+            selection = selection.map(transaction.changes());
+
+            assert_eq!(doc, "  1\n\n  2\n  3");
+            assert!(selection.len() == 1); // to ignore the selection unused warning
+        }
+
+        #[test]
+        fn uncomment_0_margin_comments_with_no_space() {
+            let mut doc = Rope::from("#");
+            let mut selection = Selection::single(0, doc.len_chars() - 1);
+
+            let transaction = toggle_line_comments(&doc, &selection, None);
+            transaction.apply(&mut doc);
+            selection = selection.map(transaction.changes());
+            assert_eq!(doc, "");
+            assert!(selection.len() == 1); // to ignore the selection unused warning
+        }
     }
 
     #[test]
@@ -413,4 +456,32 @@ mod test {
         transaction.apply(&mut doc);
         assert_eq!(doc, "");
     }
+
+    /// Test, if `get_comment_tokens` works, even if the content of the file includes chars, whose
+    /// byte size unequal the amount of chars
+    #[test]
+    fn test_get_comment_with_char_boundaries() {
+        let rope = Rope::from("··");
+        let tokens = ["//", "///"];
+
+        assert_eq!(
+            super::get_comment_token(rope.slice(..), tokens.as_slice(), 0),
+            None
+        );
+    }
+
+    /// Test for `get_comment_token`.
+    ///
+    /// Assuming the comment tokens are stored as `["///", "//"]`, `get_comment_token` should still
+    /// return `///` instead of `//` if the user is in a doc-comment section.
+    #[test]
+    fn test_use_longest_comment() {
+        let text = Rope::from("    /// amogus");
+        let tokens = ["///", "//"];
+
+        assert_eq!(
+            super::get_comment_token(text.slice(..), tokens.as_slice(), 0),
+            Some("///")
+        );
+    }
 }

helix-core/src/completion.rs

@@ -0,0 +1,26 @@
+use std::borrow::Cow;
+
+use crate::{diagnostic::LanguageServerId, Transaction};
+
+#[derive(Debug, PartialEq, Clone)]
+pub struct CompletionItem {
+    pub transaction: Transaction,
+    pub label: Cow<'static, str>,
+    pub kind: Cow<'static, str>,
+    /// Containing Markdown
+    pub documentation: Option<String>,
+    pub provider: CompletionProvider,
+}
+
+#[derive(Debug, PartialEq, Eq, Hash, Clone, Copy)]
+pub enum CompletionProvider {
+    Lsp(LanguageServerId),
+    Path,
+    Word,
+}
+
+impl From<LanguageServerId> for CompletionProvider {
+    fn from(id: LanguageServerId) -> Self {
+        CompletionProvider::Lsp(id)
+    }
+}

helix-core/src/config.rs

@@ -1,4 +1,4 @@
-use crate::syntax::{Configuration, Loader, LoaderError};
+use crate::syntax::{config::Configuration, Loader, LoaderError};
 
 /// Language configuration based on built-in languages.toml.
 pub fn default_lang_config() -> Configuration {

helix-core/src/diagnostic.rs

@@ -1,10 +1,12 @@
 //! LSP diagnostic utility types.
-use std::fmt;
+use std::{fmt, sync::Arc};
 
+pub use helix_stdx::range::Range;
 use serde::{Deserialize, Serialize};
 
 /// Describes the severity level of a [`Diagnostic`].
-#[derive(Debug, Clone, Copy, Eq, PartialEq, PartialOrd, Ord, Deserialize, Serialize)]
+#[derive(Debug, Clone, Copy, Eq, PartialEq, PartialOrd, Ord, Serialize, Deserialize)]
+#[serde(rename_all = "lowercase")]
 pub enum Severity {
     Hint,
     Info,
@@ -18,13 +20,6 @@ impl Default for Severity {
     }
 }
 
-/// A range of `char`s within the text.
-#[derive(Debug, Clone, Copy, PartialOrd, Ord, PartialEq, Eq)]
-pub struct Range {
-    pub start: usize,
-    pub end: usize,
-}
-
 #[derive(Debug, Eq, Hash, PartialEq, Clone, Deserialize, Serialize)]
 pub enum NumberOrString {
     Number(i32),
@@ -55,8 +50,35 @@ pub struct Diagnostic {
     pub data: Option<serde_json::Value>,
 }
 
-// TODO turn this into an enum + feature flag when lsp becomes optional
-pub type DiagnosticProvider = LanguageServerId;
+/// The source of a diagnostic.
+///
+/// This type is cheap to clone: all data is either `Copy` or wrapped in an `Arc`.
+#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord)]
+pub enum DiagnosticProvider {
+    Lsp {
+        /// The ID of the language server which sent the diagnostic.
+        server_id: LanguageServerId,
+        /// An optional identifier under which diagnostics are managed by the client.
+        ///
+        /// `identifier` is a field from the LSP "Pull Diagnostics" feature meant to provide an
+        /// optional "namespace" for diagnostics: a language server can respond to a diagnostics
+        /// pull request with an identifier and these diagnostics should be treated as separate
+        /// from push diagnostics. Rust-analyzer uses this feature for example to provide Cargo
+        /// diagnostics with push and internal diagnostics with pull. The push diagnostics should
+        /// not clear the pull diagnostics and vice-versa.
+        identifier: Option<Arc<str>>,
+    },
+    // Future internal features can go here...
+}
+
+impl DiagnosticProvider {
+    pub fn language_server_id(&self) -> Option<LanguageServerId> {
+        match self {
+            Self::Lsp { server_id, .. } => Some(*server_id),
+            // _ => None,
+        }
+    }
+}
 
 // while I would prefer having this in helix-lsp that necessitates a bunch of
 // conversions I would rather not add. I think its fine since this just a very
@@ -71,3 +93,10 @@ impl fmt::Display for LanguageServerId {
         write!(f, "{:?}", self.0)
     }
 }
+
+impl Diagnostic {
+    #[inline]
+    pub fn severity(&self) -> Severity {
+        self.severity.unwrap_or(Severity::Warning)
+    }
+}

helix-core/src/diff.rs

@@ -1,51 +1,22 @@
 use std::ops::Range;
 use std::time::Instant;
 
-use imara_diff::intern::InternedInput;
-use imara_diff::Algorithm;
+use imara_diff::{Algorithm, Diff, Hunk, IndentHeuristic, IndentLevel, InternedInput};
 use ropey::RopeSlice;
 
 use crate::{ChangeSet, Rope, Tendril, Transaction};
 
-/// A `imara_diff::Sink` that builds a `ChangeSet` for a character diff of a hunk
-struct CharChangeSetBuilder<'a> {
-    res: &'a mut ChangeSet,
-    hunk: &'a InternedInput<char>,
-    pos: u32,
-}
-
-impl imara_diff::Sink for CharChangeSetBuilder<'_> {
-    type Out = ();
-    fn process_change(&mut self, before: Range<u32>, after: Range<u32>) {
-        self.res.retain((before.start - self.pos) as usize);
-        self.res.delete(before.len());
-        self.pos = before.end;
-
-        let res = self.hunk.after[after.start as usize..after.end as usize]
-            .iter()
-            .map(|&token| self.hunk.interner[token])
-            .collect();
-
-        self.res.insert(res);
-    }
-
-    fn finish(self) -> Self::Out {
-        self.res.retain(self.hunk.before.len() - self.pos as usize);
-    }
-}
-
-struct LineChangeSetBuilder<'a> {
+struct ChangeSetBuilder<'a> {
     res: ChangeSet,
     after: RopeSlice<'a>,
     file: &'a InternedInput<RopeSlice<'a>>,
     current_hunk: InternedInput<char>,
+    char_diff: Diff,
     pos: u32,
 }
 
-impl imara_diff::Sink for LineChangeSetBuilder<'_> {
-    type Out = ChangeSet;
-
-    fn process_change(&mut self, before: Range<u32>, after: Range<u32>) {
+impl ChangeSetBuilder<'_> {
+    fn process_hunk(&mut self, before: Range<u32>, after: Range<u32>) {
         let len = self.file.before[self.pos as usize..before.start as usize]
             .iter()
             .map(|&it| self.file.interner[it].len_chars())
@@ -109,25 +80,36 @@ impl imara_diff::Sink for LineChangeSetBuilder<'_> {
                 .flat_map(|&it| self.file.interner[it].chars());
             self.current_hunk.update_before(hunk_before);
             self.current_hunk.update_after(hunk_after);
-
             // the histogram heuristic does not work as well
             // for characters because the same characters often reoccur
             // use myer diff instead
-            imara_diff::diff(
+            self.char_diff.compute_with(
                 Algorithm::Myers,
-                &self.current_hunk,
-                CharChangeSetBuilder {
-                    res: &mut self.res,
-                    hunk: &self.current_hunk,
-                    pos: 0,
-                },
+                &self.current_hunk.before,
+                &self.current_hunk.after,
+                self.current_hunk.interner.num_tokens(),
             );
+            let mut pos = 0;
+            for Hunk { before, after } in self.char_diff.hunks() {
+                self.res.retain((before.start - pos) as usize);
+                self.res.delete(before.len());
+                pos = before.end;
 
+                let res = self.current_hunk.after[after.start as usize..after.end as usize]
+                    .iter()
+                    .map(|&token| self.current_hunk.interner[token])
+                    .collect();
+
+                self.res.insert(res);
+            }
+            self.res
+                .retain(self.current_hunk.before.len() - pos as usize);
+            // reuse allocations
             self.current_hunk.clear();
         }
     }
 
-    fn finish(mut self) -> Self::Out {
+    fn finish(mut self) -> ChangeSet {
         let len = self.file.before[self.pos as usize..]
             .iter()
             .map(|&it| self.file.interner[it].len_chars())
@@ -140,7 +122,7 @@ impl imara_diff::Sink for LineChangeSetBuilder<'_> {
 
 struct RopeLines<'a>(RopeSlice<'a>);
 
-impl<'a> imara_diff::intern::TokenSource for RopeLines<'a> {
+impl<'a> imara_diff::TokenSource for RopeLines<'a> {
     type Token = RopeSlice<'a>;
     type Tokenizer = ropey::iter::Lines<'a>;
 
@@ -161,15 +143,23 @@ pub fn compare_ropes(before: &Rope, after: &Rope) -> Transaction {
     let res = ChangeSet::with_capacity(32);
     let after = after.slice(..);
     let file = InternedInput::new(RopeLines(before.slice(..)), RopeLines(after));
-    let builder = LineChangeSetBuilder {
+    let mut builder = ChangeSetBuilder {
         res,
         file: &file,
         after,
         pos: 0,
         current_hunk: InternedInput::default(),
+        char_diff: Diff::default(),
     };
-
-    let res = imara_diff::diff(Algorithm::Histogram, &file, builder).into();
+    let mut diff = Diff::compute(Algorithm::Histogram, &file);
+    diff.postprocess_with_heuristic(
+        &file,
+        IndentHeuristic::new(|token| IndentLevel::for_ascii_line(file.interner[token].bytes(), 4)),
+    );
+    for hunk in diff.hunks() {
+        builder.process_hunk(hunk.before, hunk.after)
+    }
+    let res = builder.finish().into();
 
     log::debug!(
         "rope diff took {}s",

helix-core/src/doc_formatter.rs

@@ -10,18 +10,21 @@
 //! called a "block" and the caller must advance it as needed.
 
 use std::borrow::Cow;
+use std::cmp::Ordering;
 use std::fmt::Debug;
-use std::mem::{replace, take};
+use std::mem::replace;
 
 #[cfg(test)]
 mod test;
 
 use unicode_segmentation::{Graphemes, UnicodeSegmentation};
 
+use helix_stdx::rope::{RopeGraphemes, RopeSliceExt};
+
 use crate::graphemes::{Grapheme, GraphemeStr};
 use crate::syntax::Highlight;
 use crate::text_annotations::TextAnnotations;
-use crate::{Position, RopeGraphemes, RopeSlice};
+use crate::{Position, RopeSlice};
 
 /// TODO make Highlight a u32 to reduce the size of this enum to a single word.
 #[derive(Debug, Clone, Copy)]
@@ -37,52 +40,104 @@ pub enum GraphemeSource {
     },
 }
 
-#[derive(Debug, Clone)]
-pub struct FormattedGrapheme<'a> {
-    pub grapheme: Grapheme<'a>,
-    pub source: GraphemeSource,
+impl GraphemeSource {
+    /// Returns whether this grapheme is virtual inline text
+    pub fn is_virtual(self) -> bool {
+        matches!(self, GraphemeSource::VirtualText { .. })
+    }
+
+    pub fn is_eof(self) -> bool {
+        // all doc chars except the EOF char have non-zero codepoints
+        matches!(self, GraphemeSource::Document { codepoints: 0 })
+    }
+
+    pub fn doc_chars(self) -> usize {
+        match self {
+            GraphemeSource::Document { codepoints } => codepoints as usize,
+            GraphemeSource::VirtualText { .. } => 0,
+        }
+    }
 }
 
-impl<'a> FormattedGrapheme<'a> {
-    pub fn new(
+#[derive(Debug, Clone)]
+pub struct FormattedGrapheme<'a> {
+    pub raw: Grapheme<'a>,
+    pub source: GraphemeSource,
+    pub visual_pos: Position,
+    /// Document line at the start of the grapheme
+    pub line_idx: usize,
+    /// Document char position at the start of the grapheme
+    pub char_idx: usize,
+}
+
+impl FormattedGrapheme<'_> {
+    pub fn is_virtual(&self) -> bool {
+        self.source.is_virtual()
+    }
+
+    pub fn doc_chars(&self) -> usize {
+        self.source.doc_chars()
+    }
+
+    pub fn is_whitespace(&self) -> bool {
+        self.raw.is_whitespace()
+    }
+
+    pub fn width(&self) -> usize {
+        self.raw.width()
+    }
+
+    pub fn is_word_boundary(&self) -> bool {
+        self.raw.is_word_boundary()
+    }
+}
+
+#[derive(Debug, Clone)]
+struct GraphemeWithSource<'a> {
+    grapheme: Grapheme<'a>,
+    source: GraphemeSource,
+}
+
+impl<'a> GraphemeWithSource<'a> {
+    fn new(
         g: GraphemeStr<'a>,
         visual_x: usize,
         tab_width: u16,
         source: GraphemeSource,
-    ) -> FormattedGrapheme<'a> {
-        FormattedGrapheme {
+    ) -> GraphemeWithSource<'a> {
+        GraphemeWithSource {
             grapheme: Grapheme::new(g, visual_x, tab_width),
             source,
         }
     }
-    /// Returns whether this grapheme is virtual inline text
-    pub fn is_virtual(&self) -> bool {
-        matches!(self.source, GraphemeSource::VirtualText { .. })
-    }
-
-    pub fn placeholder() -> Self {
-        FormattedGrapheme {
+    fn placeholder() -> Self {
+        GraphemeWithSource {
             grapheme: Grapheme::Other { g: " ".into() },
             source: GraphemeSource::Document { codepoints: 0 },
         }
     }
 
-    pub fn doc_chars(&self) -> usize {
-        match self.source {
-            GraphemeSource::Document { codepoints } => codepoints as usize,
-            GraphemeSource::VirtualText { .. } => 0,
-        }
+    fn doc_chars(&self) -> usize {
+        self.source.doc_chars()
     }
 
-    pub fn is_whitespace(&self) -> bool {
+    fn is_whitespace(&self) -> bool {
         self.grapheme.is_whitespace()
     }
 
-    pub fn width(&self) -> usize {
+    fn is_newline(&self) -> bool {
+        matches!(self.grapheme, Grapheme::Newline)
+    }
+
+    fn is_eof(&self) -> bool {
+        self.source.is_eof()
+    }
+
+    fn width(&self) -> usize {
         self.grapheme.width()
     }
 
-    pub fn is_word_boundary(&self) -> bool {
+    fn is_word_boundary(&self) -> bool {
         self.grapheme.is_word_boundary()
     }
 }
@@ -96,6 +151,7 @@ pub struct TextFormat {
     pub wrap_indicator: Box<str>,
     pub wrap_indicator_highlight: Option<Highlight>,
     pub viewport_width: u16,
+    pub soft_wrap_at_text_width: bool,
 }
 
 // test implementation is basically only used for testing or when softwrap is always disabled
@@ -109,6 +165,7 @@ impl Default for TextFormat {
             wrap_indicator: Box::from(" "),
             viewport_width: 17,
             wrap_indicator_highlight: None,
+            soft_wrap_at_text_width: false,
         }
     }
 }
@@ -127,10 +184,7 @@ pub struct DocumentFormatter<'t> {
     line_pos: usize,
     exhausted: bool,
 
-    /// Line breaks to be reserved for virtual text
-    /// at the next line break
-    virtual_lines: usize,
-    inline_anntoation_graphemes: Option<(Graphemes<'t>, Option<Highlight>)>,
+    inline_annotation_graphemes: Option<(Graphemes<'t>, Option<Highlight>)>,
 
     // softwrap specific
     /// The indentation of the current line
@@ -139,9 +193,9 @@ pub struct DocumentFormatter<'t> {
     indent_level: Option<usize>,
     /// In case a long word needs to be split a single grapheme might need to be wrapped
     /// while the rest of the word stays on the same line
-    peeked_grapheme: Option<(FormattedGrapheme<'t>, usize)>,
+    peeked_grapheme: Option<GraphemeWithSource<'t>>,
     /// A first-in first-out (fifo) buffer for the Graphemes of any given word
-    word_buf: Vec<FormattedGrapheme<'t>>,
+    word_buf: Vec<GraphemeWithSource<'t>>,
     /// The index of the next grapheme that will be yielded from the `word_buf`
     word_i: usize,
 }
@@ -157,35 +211,35 @@ impl<'t> DocumentFormatter<'t> {
         text_fmt: &'t TextFormat,
         annotations: &'t TextAnnotations,
         char_idx: usize,
-    ) -> (Self, usize) {
+    ) -> Self {
         // TODO divide long lines into blocks to avoid bad performance for long lines
         let block_line_idx = text.char_to_line(char_idx.min(text.len_chars()));
         let block_char_idx = text.line_to_char(block_line_idx);
         annotations.reset_pos(block_char_idx);
-        (
-            DocumentFormatter {
-                text_fmt,
-                annotations,
-                visual_pos: Position { row: 0, col: 0 },
-                graphemes: RopeGraphemes::new(text.slice(block_char_idx..)),
-                char_pos: block_char_idx,
-                exhausted: false,
-                virtual_lines: 0,
-                indent_level: None,
-                peeked_grapheme: None,
-                word_buf: Vec::with_capacity(64),
-                word_i: 0,
-                line_pos: block_line_idx,
-                inline_anntoation_graphemes: None,
-            },
-            block_char_idx,
-        )
+
+        DocumentFormatter {
+            text_fmt,
+            annotations,
+            visual_pos: Position { row: 0, col: 0 },
+            graphemes: text.slice(block_char_idx..).graphemes(),
+            char_pos: block_char_idx,
+            exhausted: false,
+            indent_level: None,
+            peeked_grapheme: None,
+            word_buf: Vec::with_capacity(64),
+            word_i: 0,
+            line_pos: block_line_idx,
+            inline_annotation_graphemes: None,
+        }
     }
 
-    fn next_inline_annotation_grapheme(&mut self) -> Option<(&'t str, Option<Highlight>)> {
+    fn next_inline_annotation_grapheme(
+        &mut self,
+        char_pos: usize,
+    ) -> Option<(&'t str, Option<Highlight>)> {
         loop {
             if let Some(&mut (ref mut annotation, highlight)) =
-                self.inline_anntoation_graphemes.as_mut()
+                self.inline_annotation_graphemes.as_mut()
             {
                 if let Some(grapheme) = annotation.next() {
                     return Some((grapheme, highlight));
@@ -193,9 +247,9 @@ impl<'t> DocumentFormatter<'t> {
             }
 
             if let Some((annotation, highlight)) =
-                self.annotations.next_inline_annotation_at(self.char_pos)
+                self.annotations.next_inline_annotation_at(char_pos)
             {
-                self.inline_anntoation_graphemes = Some((
+                self.inline_annotation_graphemes = Some((
                     UnicodeSegmentation::graphemes(&*annotation.text, true),
                     highlight,
                 ))
@@ -205,21 +259,19 @@ impl<'t> DocumentFormatter<'t> {
         }
     }
 
-    fn advance_grapheme(&mut self, col: usize) -> Option<FormattedGrapheme<'t>> {
+    fn advance_grapheme(&mut self, col: usize, char_pos: usize) -> Option<GraphemeWithSource<'t>> {
         let (grapheme, source) =
-            if let Some((grapheme, highlight)) = self.next_inline_annotation_grapheme() {
+            if let Some((grapheme, highlight)) = self.next_inline_annotation_grapheme(char_pos) {
                 (grapheme.into(), GraphemeSource::VirtualText { highlight })
             } else if let Some(grapheme) = self.graphemes.next() {
-                self.virtual_lines += self.annotations.annotation_lines_at(self.char_pos);
                 let codepoints = grapheme.len_chars() as u32;
 
-                let overlay = self.annotations.overlay_at(self.char_pos);
+                let overlay = self.annotations.overlay_at(char_pos);
                 let grapheme = match overlay {
                     Some((overlay, _)) => overlay.grapheme.as_str().into(),
                     None => Cow::from(grapheme).into(),
                 };
 
-                self.char_pos += codepoints as usize;
                 (grapheme, GraphemeSource::Document { codepoints })
             } else {
                 if self.exhausted {
@@ -228,19 +280,19 @@ impl<'t> DocumentFormatter<'t> {
                 self.exhausted = true;
                 // EOF grapheme is required for rendering
                 // and correct position computations
-                return Some(FormattedGrapheme {
+                return Some(GraphemeWithSource {
                     grapheme: Grapheme::Other { g: " ".into() },
                     source: GraphemeSource::Document { codepoints: 0 },
                 });
             };
 
-        let grapheme = FormattedGrapheme::new(grapheme, col, self.text_fmt.tab_width, source);
+        let grapheme = GraphemeWithSource::new(grapheme, col, self.text_fmt.tab_width, source);
 
         Some(grapheme)
     }
 
     /// Move a word to the next visual line
-    fn wrap_word(&mut self, virtual_lines_before_word: usize) -> usize {
+    fn wrap_word(&mut self) -> usize {
         // softwrap this word to the next line
         let indent_carry_over = if let Some(indent) = self.indent_level {
             if indent as u16 <= self.text_fmt.max_indent_retain {
@@ -254,15 +306,17 @@ impl<'t> DocumentFormatter<'t> {
             0
         };
 
+        let virtual_lines =
+            self.annotations
+                .virtual_lines_at(self.char_pos, self.visual_pos, self.line_pos);
         self.visual_pos.col = indent_carry_over as usize;
-        self.virtual_lines -= virtual_lines_before_word;
-        self.visual_pos.row += 1 + virtual_lines_before_word;
+        self.visual_pos.row += 1 + virtual_lines;
         let mut i = 0;
         let mut word_width = 0;
         let wrap_indicator = UnicodeSegmentation::graphemes(&*self.text_fmt.wrap_indicator, true)
             .map(|g| {
                 i += 1;
-                let grapheme = FormattedGrapheme::new(
+                let grapheme = GraphemeWithSource::new(
                     g.into(),
                     self.visual_pos.col + word_width,
                     self.text_fmt.tab_width,
@@ -282,46 +336,72 @@ impl<'t> DocumentFormatter<'t> {
                 .change_position(visual_x, self.text_fmt.tab_width);
             word_width += grapheme.width();
         }
+        if let Some(grapheme) = &mut self.peeked_grapheme {
+            let visual_x = self.visual_pos.col + word_width;
+            grapheme
+                .grapheme
+                .change_position(visual_x, self.text_fmt.tab_width);
+        }
         word_width
     }
 
+    fn peek_grapheme(&mut self, col: usize, char_pos: usize) -> Option<&GraphemeWithSource<'t>> {
+        if self.peeked_grapheme.is_none() {
+            self.peeked_grapheme = self.advance_grapheme(col, char_pos);
+        }
+        self.peeked_grapheme.as_ref()
+    }
+
+    fn next_grapheme(&mut self, col: usize, char_pos: usize) -> Option<GraphemeWithSource<'t>> {
+        self.peek_grapheme(col, char_pos);
+        self.peeked_grapheme.take()
+    }
+
     fn advance_to_next_word(&mut self) {
         self.word_buf.clear();
         let mut word_width = 0;
-        let virtual_lines_before_word = self.virtual_lines;
-        let mut virtual_lines_before_grapheme = self.virtual_lines;
+        let mut word_chars = 0;
+
+        if self.exhausted {
+            return;
+        }
 
         loop {
-            // softwrap word if necessary
-            if word_width + self.visual_pos.col >= self.text_fmt.viewport_width as usize {
-                // wrapping this word would move too much text to the next line
-                // split the word at the line end instead
-                if word_width > self.text_fmt.max_wrap as usize {
-                    // Usually we stop accomulating graphemes as soon as softwrapping becomes necessary.
-                    // However if the last grapheme is multiple columns wide it might extend beyond the EOL.
-                    // The condition below ensures that this grapheme is not cutoff and instead wrapped to the next line
-                    if word_width + self.visual_pos.col > self.text_fmt.viewport_width as usize {
-                        self.peeked_grapheme = self.word_buf.pop().map(|grapheme| {
-                            (grapheme, self.virtual_lines - virtual_lines_before_grapheme)
-                        });
-                        self.virtual_lines = virtual_lines_before_grapheme;
-                    }
+            let mut col = self.visual_pos.col + word_width;
+            let char_pos = self.char_pos + word_chars;
+            match col.cmp(&(self.text_fmt.viewport_width as usize)) {
+                // The EOF char and newline chars are always selectable in helix. That means
+                // that wrapping happens "too-early" if a word fits a line perfectly. This
+                // is intentional so that all selectable graphemes are always visible (and
+                // therefore the cursor never disappears). However if the user manually set a
+                // lower softwrap width then this is undesirable. Just increasing the viewport-
+                // width by one doesn't work because if a line is wrapped multiple times then
+                // some words may extend past the specified width.
+                //
+                // So we special case a word that ends exactly at line bounds and is followed
+                // by a newline/eof character here.
+                Ordering::Equal
+                    if self.text_fmt.soft_wrap_at_text_width
+                        && self
+                            .peek_grapheme(col, char_pos)
+                            .is_some_and(|grapheme| grapheme.is_newline() || grapheme.is_eof()) => {
+                }
+                Ordering::Equal if word_width > self.text_fmt.max_wrap as usize => return,
+                Ordering::Greater if word_width > self.text_fmt.max_wrap as usize => {
+                    self.peeked_grapheme = self.word_buf.pop();
                     return;
                 }
-
-                word_width = self.wrap_word(virtual_lines_before_word);
+                Ordering::Equal | Ordering::Greater => {
+                    word_width = self.wrap_word();
+                    col = self.visual_pos.col + word_width;
+                }
+                Ordering::Less => (),
             }
 
-            virtual_lines_before_grapheme = self.virtual_lines;
-
-            let grapheme = if let Some((grapheme, virtual_lines)) = self.peeked_grapheme.take() {
-                self.virtual_lines += virtual_lines;
-                grapheme
-            } else if let Some(grapheme) = self.advance_grapheme(self.visual_pos.col + word_width) {
-                grapheme
-            } else {
+            let Some(grapheme) = self.next_grapheme(col, char_pos) else {
                 return;
             };
+            word_chars += grapheme.doc_chars();
 
             // Track indentation
             if !grapheme.is_whitespace() && self.indent_level.is_none() {
@@ -340,19 +420,18 @@ impl<'t> DocumentFormatter<'t> {
         }
     }
 
-    /// returns the document line pos of the **next** grapheme that will be yielded
-    pub fn line_pos(&self) -> usize {
-        self.line_pos
+    /// returns the char index at the end of the last yielded grapheme
+    pub fn next_char_pos(&self) -> usize {
+        self.char_pos
     }
-
-    /// returns the visual pos of the **next** grapheme that will be yielded
-    pub fn visual_pos(&self) -> Position {
+    /// returns the visual position at the end of the last yielded grapheme
+    pub fn next_visual_pos(&self) -> Position {
         self.visual_pos
     }
 }
 
 impl<'t> Iterator for DocumentFormatter<'t> {
-    type Item = (FormattedGrapheme<'t>, Position);
+    type Item = FormattedGrapheme<'t>;
 
     fn next(&mut self) -> Option<Self::Item> {
         let grapheme = if self.text_fmt.soft_wrap {
@@ -362,23 +441,40 @@ impl<'t> Iterator for DocumentFormatter<'t> {
             }
             let grapheme = replace(
                 self.word_buf.get_mut(self.word_i)?,
-                FormattedGrapheme::placeholder(),
+                GraphemeWithSource::placeholder(),
             );
             self.word_i += 1;
             grapheme
         } else {
-            self.advance_grapheme(self.visual_pos.col)?
+            self.advance_grapheme(self.visual_pos.col, self.char_pos)?
         };
 
-        let pos = self.visual_pos;
-        if grapheme.grapheme == Grapheme::Newline {
-            self.visual_pos.row += 1;
-            self.visual_pos.row += take(&mut self.virtual_lines);
+        let grapheme = FormattedGrapheme {
+            raw: grapheme.grapheme,
+            source: grapheme.source,
+            visual_pos: self.visual_pos,
+            line_idx: self.line_pos,
+            char_idx: self.char_pos,
+        };
+
+        self.char_pos += grapheme.doc_chars();
+        if !grapheme.is_virtual() {
+            self.annotations.process_virtual_text_anchors(&grapheme);
+        }
+        if grapheme.raw == Grapheme::Newline {
+            // move to end of newline char
+            self.visual_pos.col += 1;
+            let virtual_lines =
+                self.annotations
+                    .virtual_lines_at(self.char_pos, self.visual_pos, self.line_pos);
+            self.visual_pos.row += 1 + virtual_lines;
             self.visual_pos.col = 0;
-            self.line_pos += 1;
+            if !grapheme.is_virtual() {
+                self.line_pos += 1;
+            }
         } else {
             self.visual_pos.col += grapheme.width();
         }
-        Some((grapheme, pos))
+        Some(grapheme)
     }
 }

helix-core/src/doc_formatter/test.rs

@@ -12,6 +12,7 @@ impl TextFormat {
             wrap_indicator_highlight: None,
             // use a prime number to allow lining up too often with repeat
             viewport_width: 17,
+            soft_wrap_at_text_width: false,
         }
     }
 }
@@ -21,20 +22,23 @@ impl<'t> DocumentFormatter<'t> {
         use std::fmt::Write;
         let mut res = String::new();
         let viewport_width = self.text_fmt.viewport_width;
+        let soft_wrap_at_text_width = self.text_fmt.soft_wrap_at_text_width;
         let mut line = 0;
 
-        for (grapheme, pos) in self {
-            if pos.row != line {
+        for grapheme in self {
+            if grapheme.visual_pos.row != line {
                 line += 1;
-                assert_eq!(pos.row, line);
-                write!(res, "\n{}", ".".repeat(pos.col)).unwrap();
+                assert_eq!(grapheme.visual_pos.row, line);
+                write!(res, "\n{}", ".".repeat(grapheme.visual_pos.col)).unwrap();
+            }
+            if !soft_wrap_at_text_width {
                 assert!(
-                    pos.col <= viewport_width as usize,
+                    grapheme.visual_pos.col <= viewport_width as usize,
                     "softwrapped failed {}<={viewport_width}",
-                    pos.col
+                    grapheme.visual_pos.col
                 );
             }
-            write!(res, "{}", grapheme.grapheme).unwrap();
+            write!(res, "{}", grapheme.raw).unwrap();
         }
 
         res
@@ -48,7 +52,6 @@ fn softwrap_text(text: &str) -> String {
         &TextAnnotations::default(),
         0,
     )
-    .0
     .collect_to_str()
 }
 
@@ -99,6 +102,30 @@ fn long_word_softwrap() {
     );
 }
 
+#[test]
+fn softwrap_multichar_grapheme() {
+    assert_eq!(
+        softwrap_text("xxxx xxxx xxx a\u{0301}bc\n"),
+        "xxxx xxxx xxx \n.ábc \n "
+    )
+}
+
+fn softwrap_text_at_text_width(text: &str) -> String {
+    let mut text_fmt = TextFormat::new_test(true);
+    text_fmt.soft_wrap_at_text_width = true;
+    let annotations = TextAnnotations::default();
+    let mut formatter =
+        DocumentFormatter::new_at_prev_checkpoint(text.into(), &text_fmt, &annotations, 0);
+    formatter.collect_to_str()
+}
+#[test]
+fn long_word_softwrap_text_width() {
+    assert_eq!(
+        softwrap_text_at_text_width("xxxxxxxx1xxxx2xxx\nxxxxxxxx1xxxx2xxx"),
+        "xxxxxxxx1xxxx2xxx \nxxxxxxxx1xxxx2xxx "
+    );
+}
+
 fn overlay_text(text: &str, char_pos: usize, softwrap: bool, overlays: &[Overlay]) -> String {
     DocumentFormatter::new_at_prev_checkpoint(
         text.into(),
@@ -106,7 +133,6 @@ fn overlay_text(text: &str, char_pos: usize, softwrap: bool, overlays: &[Overlay
         TextAnnotations::default().add_overlay(overlays, None),
         char_pos,
     )
-    .0
     .collect_to_str()
 }
 
@@ -143,7 +169,6 @@ fn annotate_text(text: &str, softwrap: bool, annotations: &[InlineAnnotation]) -
         TextAnnotations::default().add_inline_annotations(annotations, None),
         0,
     )
-    .0
     .collect_to_str()
 }
 
@@ -182,7 +207,6 @@ fn annotation_and_overlay() {
                 .add_overlay(overlay.as_slice(), None),
             0,
         )
-        .0
         .collect_to_str(),
         "fooo  bar "
     );

helix-core/src/editor_config.rs

@@ -0,0 +1,334 @@
+//! Support for [EditorConfig](https://EditorConfig.org) configuration loading.
+//!
+//! EditorConfig is an editor-agnostic format for specifying configuration in an INI-like, human
+//! friendly syntax in `.editorconfig` files (which are intended to be checked into VCS). This
+//! module provides functions to search for all `.editorconfig` files that apply to a given path
+//! and returns an `EditorConfig` type containing any specified configuration options.
+//!
+//! At time of writing, this module follows the [spec](https://spec.editorconfig.org/) at
+//! version 0.17.2.
+
+use std::{
+    collections::HashMap,
+    fs,
+    num::{NonZeroU16, NonZeroU8},
+    path::Path,
+    str::FromStr,
+};
+
+use encoding_rs::Encoding;
+use globset::{GlobBuilder, GlobMatcher};
+
+use crate::{
+    indent::{IndentStyle, MAX_INDENT},
+    LineEnding,
+};
+
+/// Configuration declared for a path in `.editorconfig` files.
+#[derive(Debug, Default, PartialEq, Eq)]
+pub struct EditorConfig {
+    pub indent_style: Option<IndentStyle>,
+    pub tab_width: Option<NonZeroU8>,
+    pub line_ending: Option<LineEnding>,
+    pub encoding: Option<&'static Encoding>,
+    // pub spelling_language: Option<SpellingLanguage>,
+    pub trim_trailing_whitespace: Option<bool>,
+    pub insert_final_newline: Option<bool>,
+    pub max_line_length: Option<NonZeroU16>,
+}
+
+impl EditorConfig {
+    /// Finds any configuration in `.editorconfig` files which applies to the given path.
+    ///
+    /// If no configuration applies then `EditorConfig::default()` is returned.
+    pub fn find(path: &Path) -> Self {
+        let mut configs = Vec::new();
+        // <https://spec.editorconfig.org/#file-processing>
+        for ancestor in path.ancestors() {
+            let editor_config_file = ancestor.join(".editorconfig");
+            let Ok(contents) = fs::read_to_string(&editor_config_file) else {
+                continue;
+            };
+            let ini = match contents.parse::<Ini>() {
+                Ok(ini) => ini,
+                Err(err) => {
+                    log::warn!("Ignoring EditorConfig file at '{editor_config_file:?}' because a glob failed to compile: {err}");
+                    continue;
+                }
+            };
+            let is_root = ini.pairs.get("root").map(AsRef::as_ref) == Some("true");
+            configs.push((ini, ancestor));
+            // > The search shall stop if an EditorConfig file is found with the `root` key set to
+            // > `true` in the preamble or when reaching the root filesystem directory.
+            if is_root {
+                break;
+            }
+        }
+
+        let mut pairs = Pairs::new();
+        // Reverse the configuration stack so that the `.editorconfig` files closest to `path`
+        // are applied last and overwrite settings in files closer to the search ceiling.
+        //
+        // > If multiple EditorConfig files have matching sections, the pairs from the closer
+        // > EditorConfig file are read last, so pairs in closer files take precedence.
+        for (config, dir) in configs.into_iter().rev() {
+            let relative_path = path.strip_prefix(dir).expect("dir is an ancestor of path");
+
+            for section in config.sections {
+                if section.glob.is_match(relative_path) {
+                    log::info!(
+                        "applying EditorConfig from section '{}' in file {:?}",
+                        section.glob.glob(),
+                        dir.join(".editorconfig")
+                    );
+                    pairs.extend(section.pairs);
+                }
+            }
+        }
+
+        Self::from_pairs(pairs)
+    }
+
+    fn from_pairs(pairs: Pairs) -> Self {
+        enum IndentSize {
+            Tab,
+            Spaces(NonZeroU8),
+        }
+
+        // <https://spec.editorconfig.org/#supported-pairs>
+        let indent_size = pairs.get("indent_size").and_then(|value| {
+            if value.as_ref() == "tab" {
+                Some(IndentSize::Tab)
+            } else if let Ok(spaces) = value.parse::<NonZeroU8>() {
+                Some(IndentSize::Spaces(spaces))
+            } else {
+                None
+            }
+        });
+        let tab_width = pairs
+            .get("tab_width")
+            .and_then(|value| value.parse::<NonZeroU8>().ok())
+            .or(match indent_size {
+                Some(IndentSize::Spaces(spaces)) => Some(spaces),
+                _ => None,
+            });
+        let indent_style = pairs
+            .get("indent_style")
+            .and_then(|value| match value.as_ref() {
+                "tab" => Some(IndentStyle::Tabs),
+                "space" => {
+                    let spaces = match indent_size {
+                        Some(IndentSize::Spaces(spaces)) => spaces.get(),
+                        Some(IndentSize::Tab) => tab_width.map(|n| n.get()).unwrap_or(4),
+                        None => 4,
+                    };
+                    Some(IndentStyle::Spaces(spaces.clamp(1, MAX_INDENT)))
+                }
+                _ => None,
+            });
+        let line_ending = pairs
+            .get("end_of_line")
+            .and_then(|value| match value.as_ref() {
+                "lf" => Some(LineEnding::LF),
+                "crlf" => Some(LineEnding::Crlf),
+                #[cfg(feature = "unicode-lines")]
+                "cr" => Some(LineEnding::CR),
+                _ => None,
+            });
+        let encoding = pairs.get("charset").and_then(|value| match value.as_ref() {
+            "latin1" => Some(encoding_rs::WINDOWS_1252),
+            "utf-8" => Some(encoding_rs::UTF_8),
+            // `utf-8-bom` is intentionally ignored.
+            // > `utf-8-bom` is discouraged.
+            "utf-16le" => Some(encoding_rs::UTF_16LE),
+            "utf-16be" => Some(encoding_rs::UTF_16BE),
+            _ => None,
+        });
+        let trim_trailing_whitespace =
+            pairs
+                .get("trim_trailing_whitespace")
+                .and_then(|value| match value.as_ref() {
+                    "true" => Some(true),
+                    "false" => Some(false),
+                    _ => None,
+                });
+        let insert_final_newline = pairs
+            .get("insert_final_newline")
+            .and_then(|value| match value.as_ref() {
+                "true" => Some(true),
+                "false" => Some(false),
+                _ => None,
+            });
+        // This option is not in the spec but is supported by some editors.
+        // <https://github.com/editorconfig/editorconfig/wiki/EditorConfig-Properties#max_line_length>
+        let max_line_length = pairs
+            .get("max_line_length")
+            .and_then(|value| value.parse::<NonZeroU16>().ok());
+
+        Self {
+            indent_style,
+            tab_width,
+            line_ending,
+            encoding,
+            trim_trailing_whitespace,
+            insert_final_newline,
+            max_line_length,
+        }
+    }
+}
+
+type Pairs = HashMap<Box<str>, Box<str>>;
+
+#[derive(Debug)]
+struct Section {
+    glob: GlobMatcher,
+    pairs: Pairs,
+}
+
+#[derive(Debug, Default)]
+struct Ini {
+    pairs: Pairs,
+    sections: Vec<Section>,
+}
+
+impl FromStr for Ini {
+    type Err = globset::Error;
+
+    fn from_str(source: &str) -> Result<Self, Self::Err> {
+        // <https://spec.editorconfig.org/#file-format>
+        let mut ini = Ini::default();
+        // > EditorConfig files are in an INI-like file format. To read an EditorConfig file, take
+        // > one line at a time, from beginning to end. For each line:
+        for full_line in source.lines() {
+            // > 1. Remove all leading and trailing whitespace.
+            let line = full_line.trim();
+            // > 2. Process the remaining text as specified for its type below.
+            // > The types of lines are:
+            // > * Blank: contains nothing. Blank lines are ignored.
+            if line.is_empty() {
+                continue;
+            }
+            // > * Comment: starts with a ';' or '#'. Comment lines are ignored.
+            if line.starts_with([';', '#']) {
+                continue;
+            }
+            if let Some(section) = line.strip_prefix('[').and_then(|s| s.strip_suffix(']')) {
+                // > * Section Header: starts with a `[` and ends with a `]`. These lines define
+                // >   globs...
+
+                // <https://spec.editorconfig.org/#glob-expressions>
+                // We need to modify the glob string slightly since EditorConfig's glob flavor
+                // doesn't match `globset`'s exactly. `globset` only allows '**' at the beginning
+                // or end of a glob or between two '/'s. (This replacement is not very fancy but
+                // should cover most practical cases.)
+                let mut glob_str = section.replace("**.", "**/*.");
+                if !is_glob_relative(section) {
+                    glob_str.insert_str(0, "**/");
+                }
+                let glob = GlobBuilder::new(&glob_str)
+                    .literal_separator(true)
+                    .backslash_escape(true)
+                    .empty_alternates(true)
+                    .build()?;
+                ini.sections.push(Section {
+                    glob: glob.compile_matcher(),
+                    pairs: Pairs::new(),
+                });
+            } else if let Some((key, value)) = line.split_once('=') {
+                // > * Key-Value Pair (or Pair): contains a key and a value, separated by an `=`.
+                // >     * Key: The part before the first `=` on the line.
+                // >     * Value: The part, if any, after the first `=` on the line.
+                // >     * Keys and values are trimmed of leading and trailing whitespace, but
+                // >       include any whitespace that is between non-whitespace characters.
+                // >     * If a value is not provided, then the value is an empty string.
+                let key = key.trim().to_lowercase().into_boxed_str();
+                let value = value.trim().to_lowercase().into_boxed_str();
+                if let Some(section) = ini.sections.last_mut() {
+                    section.pairs.insert(key, value);
+                } else {
+                    ini.pairs.insert(key, value);
+                }
+            }
+        }
+        Ok(ini)
+    }
+}
+
+/// Determines whether a glob is relative to the directory of the config file.
+fn is_glob_relative(source: &str) -> bool {
+    // > If the glob contains a path separator (a `/` not inside square brackets), then the
+    // > glob is relative to the directory level of the particular `.editorconfig` file itself.
+    let mut idx = 0;
+    while let Some(open) = source[idx..].find('[').map(|open| idx + open) {
+        if source[..open].contains('/') {
+            return true;
+        }
+        idx = source[open..]
+            .find(']')
+            .map_or(source.len(), |close| idx + close);
+    }
+    source[idx..].contains('/')
+}
+
+#[cfg(test)]
+mod test {
+    use super::*;
+
+    #[test]
+    fn is_glob_relative_test() {
+        assert!(is_glob_relative("subdir/*.c"));
+        assert!(!is_glob_relative("*.txt"));
+        assert!(!is_glob_relative("[a/b].c"));
+    }
+
+    fn editor_config(path: impl AsRef<Path>, source: &str) -> EditorConfig {
+        let path = path.as_ref();
+        let ini = source.parse::<Ini>().unwrap();
+        let pairs = ini
+            .sections
+            .into_iter()
+            .filter(|section| section.glob.is_match(path))
+            .fold(Pairs::new(), |mut acc, section| {
+                acc.extend(section.pairs);
+                acc
+            });
+        EditorConfig::from_pairs(pairs)
+    }
+
+    #[test]
+    fn parse_test() {
+        let source = r#"
+        [*]
+        indent_style = space
+
+        [Makefile]
+        indent_style = tab
+
+        [docs/**.txt]
+        insert_final_newline = true
+        "#;
+
+        assert_eq!(
+            editor_config("a.txt", source),
+            EditorConfig {
+                indent_style: Some(IndentStyle::Spaces(4)),
+                ..Default::default()
+            }
+        );
+        assert_eq!(
+            editor_config("pkg/Makefile", source),
+            EditorConfig {
+                indent_style: Some(IndentStyle::Tabs),
+                ..Default::default()
+            }
+        );
+        assert_eq!(
+            editor_config("docs/config/editor.txt", source),
+            EditorConfig {
+                indent_style: Some(IndentStyle::Spaces(4)),
+                insert_final_newline: Some(true),
+                ..Default::default()
+            }
+        );
+    }
+}

helix-core/src/graphemes.rs

@@ -1,7 +1,7 @@
 //! Utility functions to traverse the unicode graphemes of a `Rope`'s text contents.
 //!
 //! Based on <https://github.com/cessen/led/blob/c4fa72405f510b7fd16052f90a598c429b3104a6/src/graphemes.rs>
-use ropey::{iter::Chunks, str_utils::byte_to_char_idx, RopeSlice};
+use ropey::{str_utils::byte_to_char_idx, RopeSlice};
 use unicode_segmentation::{GraphemeCursor, GraphemeIncomplete};
 use unicode_width::UnicodeWidthStr;
 
@@ -28,6 +28,11 @@ pub enum Grapheme<'a> {
 }
 
 impl<'a> Grapheme<'a> {
+    pub fn new_decoration(g: &'static str) -> Grapheme<'a> {
+        assert_ne!(g, "\t");
+        Grapheme::new(g.into(), 0, 0)
+    }
+
     pub fn new(g: GraphemeStr<'a>, visual_x: usize, tab_width: u16) -> Grapheme<'a> {
         match g {
             g if g == "\t" => Grapheme::Tab {
@@ -59,7 +64,7 @@ impl<'a> Grapheme<'a> {
     }
 
     pub fn is_whitespace(&self) -> bool {
-        !matches!(&self, Grapheme::Other { g } if !g.chars().all(char_is_whitespace))
+        !matches!(&self, Grapheme::Other { g } if !g.chars().next().is_some_and(char_is_whitespace))
     }
 
     // TODO currently word boundaries are used for softwrapping.
@@ -67,7 +72,7 @@ impl<'a> Grapheme<'a> {
     // This could however be improved in the future by considering unicode
     // character classes but
     pub fn is_word_boundary(&self) -> bool {
-        !matches!(&self, Grapheme::Other { g,.. } if g.chars().all(char_is_word))
+        !matches!(&self, Grapheme::Other { g,.. } if g.chars().next().is_some_and(char_is_word))
     }
 }
 
@@ -114,6 +119,9 @@ pub fn grapheme_width(g: &str) -> usize {
     }
 }
 
+// NOTE: for byte indexing versions of these functions see `RopeSliceExt`'s
+// `floor_grapheme_boundary` and `ceil_grapheme_boundary` and the rope grapheme iterators.
+
 #[must_use]
 pub fn nth_prev_grapheme_boundary(slice: RopeSlice, char_idx: usize, n: usize) -> usize {
     // Bounds check
@@ -203,43 +211,6 @@ pub fn nth_next_grapheme_boundary(slice: RopeSlice, char_idx: usize, n: usize) -
     chunk_char_idx + tmp
 }
 
-#[must_use]
-pub fn nth_next_grapheme_boundary_byte(slice: RopeSlice, mut byte_idx: usize, n: usize) -> usize {
-    // Bounds check
-    debug_assert!(byte_idx <= slice.len_bytes());
-
-    // Get the chunk with our byte index in it.
-    let (mut chunk, mut chunk_byte_idx, mut _chunk_char_idx, _) = slice.chunk_at_byte(byte_idx);
-
-    // Set up the grapheme cursor.
-    let mut gc = GraphemeCursor::new(byte_idx, slice.len_bytes(), true);
-
-    // Find the nth next grapheme cluster boundary.
-    for _ in 0..n {
-        loop {
-            match gc.next_boundary(chunk, chunk_byte_idx) {
-                Ok(None) => return slice.len_bytes(),
-                Ok(Some(n)) => {
-                    byte_idx = n;
-                    break;
-                }
-                Err(GraphemeIncomplete::NextChunk) => {
-                    chunk_byte_idx += chunk.len();
-                    let (a, _, _c, _) = slice.chunk_at_byte(chunk_byte_idx);
-                    chunk = a;
-                    // chunk_char_idx = c;
-                }
-                Err(GraphemeIncomplete::PreContext(n)) => {
-                    let ctx_chunk = slice.chunk_at_byte(n - 1).0;
-                    gc.provide_context(ctx_chunk, n - ctx_chunk.len());
-                }
-                _ => unreachable!(),
-            }
-        }
-    }
-    byte_idx
-}
-
 /// Finds the next grapheme boundary after the given char position.
 #[must_use]
 #[inline(always)]
@@ -247,13 +218,6 @@ pub fn next_grapheme_boundary(slice: RopeSlice, char_idx: usize) -> usize {
     nth_next_grapheme_boundary(slice, char_idx, 1)
 }
 
-/// Finds the next grapheme boundary after the given byte position.
-#[must_use]
-#[inline(always)]
-pub fn next_grapheme_boundary_byte(slice: RopeSlice, byte_idx: usize) -> usize {
-    nth_next_grapheme_boundary_byte(slice, byte_idx, 1)
-}
-
 /// Returns the passed char index if it's already a grapheme boundary,
 /// or the next grapheme boundary char index if not.
 #[must_use]
@@ -278,215 +242,6 @@ pub fn ensure_grapheme_boundary_prev(slice: RopeSlice, char_idx: usize) -> usize
     }
 }
 
-/// Returns whether the given char position is a grapheme boundary.
-#[must_use]
-pub fn is_grapheme_boundary(slice: RopeSlice, char_idx: usize) -> bool {
-    // Bounds check
-    debug_assert!(char_idx <= slice.len_chars());
-
-    // We work with bytes for this, so convert.
-    let byte_idx = slice.char_to_byte(char_idx);
-
-    // Get the chunk with our byte index in it.
-    let (chunk, chunk_byte_idx, _, _) = slice.chunk_at_byte(byte_idx);
-
-    // Set up the grapheme cursor.
-    let mut gc = GraphemeCursor::new(byte_idx, slice.len_bytes(), true);
-
-    // Determine if the given position is a grapheme cluster boundary.
-    loop {
-        match gc.is_boundary(chunk, chunk_byte_idx) {
-            Ok(n) => return n,
-            Err(GraphemeIncomplete::PreContext(n)) => {
-                let (ctx_chunk, ctx_byte_start, _, _) = slice.chunk_at_byte(n - 1);
-                gc.provide_context(ctx_chunk, ctx_byte_start);
-            }
-            Err(_) => unreachable!(),
-        }
-    }
-}
-
-/// Returns whether the given byte position is a grapheme boundary.
-#[must_use]
-pub fn is_grapheme_boundary_byte(slice: RopeSlice, byte_idx: usize) -> bool {
-    // Bounds check
-    debug_assert!(byte_idx <= slice.len_bytes());
-
-    // Get the chunk with our byte index in it.
-    let (chunk, chunk_byte_idx, _, _) = slice.chunk_at_byte(byte_idx);
-
-    // Set up the grapheme cursor.
-    let mut gc = GraphemeCursor::new(byte_idx, slice.len_bytes(), true);
-
-    // Determine if the given position is a grapheme cluster boundary.
-    loop {
-        match gc.is_boundary(chunk, chunk_byte_idx) {
-            Ok(n) => return n,
-            Err(GraphemeIncomplete::PreContext(n)) => {
-                let (ctx_chunk, ctx_byte_start, _, _) = slice.chunk_at_byte(n - 1);
-                gc.provide_context(ctx_chunk, ctx_byte_start);
-            }
-            Err(_) => unreachable!(),
-        }
-    }
-}
-
-/// An iterator over the graphemes of a `RopeSlice`.
-#[derive(Clone)]
-pub struct RopeGraphemes<'a> {
-    text: RopeSlice<'a>,
-    chunks: Chunks<'a>,
-    cur_chunk: &'a str,
-    cur_chunk_start: usize,
-    cursor: GraphemeCursor,
-}
-
-impl<'a> fmt::Debug for RopeGraphemes<'a> {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
-        f.debug_struct("RopeGraphemes")
-            .field("text", &self.text)
-            .field("chunks", &self.chunks)
-            .field("cur_chunk", &self.cur_chunk)
-            .field("cur_chunk_start", &self.cur_chunk_start)
-            // .field("cursor", &self.cursor)
-            .finish()
-    }
-}
-
-impl<'a> RopeGraphemes<'a> {
-    #[must_use]
-    pub fn new(slice: RopeSlice) -> RopeGraphemes {
-        let mut chunks = slice.chunks();
-        let first_chunk = chunks.next().unwrap_or("");
-        RopeGraphemes {
-            text: slice,
-            chunks,
-            cur_chunk: first_chunk,
-            cur_chunk_start: 0,
-            cursor: GraphemeCursor::new(0, slice.len_bytes(), true),
-        }
-    }
-}
-
-impl<'a> Iterator for RopeGraphemes<'a> {
-    type Item = RopeSlice<'a>;
-
-    fn next(&mut self) -> Option<RopeSlice<'a>> {
-        let a = self.cursor.cur_cursor();
-        let b;
-        loop {
-            match self
-                .cursor
-                .next_boundary(self.cur_chunk, self.cur_chunk_start)
-            {
-                Ok(None) => {
-                    return None;
-                }
-                Ok(Some(n)) => {
-                    b = n;
-                    break;
-                }
-                Err(GraphemeIncomplete::NextChunk) => {
-                    self.cur_chunk_start += self.cur_chunk.len();
-                    self.cur_chunk = self.chunks.next().unwrap_or("");
-                }
-                Err(GraphemeIncomplete::PreContext(idx)) => {
-                    let (chunk, byte_idx, _, _) = self.text.chunk_at_byte(idx.saturating_sub(1));
-                    self.cursor.provide_context(chunk, byte_idx);
-                }
-                _ => unreachable!(),
-            }
-        }
-
-        if a < self.cur_chunk_start {
-            Some(self.text.byte_slice(a..b))
-        } else {
-            let a2 = a - self.cur_chunk_start;
-            let b2 = b - self.cur_chunk_start;
-            Some((&self.cur_chunk[a2..b2]).into())
-        }
-    }
-}
-
-/// An iterator over the graphemes of a `RopeSlice` in reverse.
-#[derive(Clone)]
-pub struct RevRopeGraphemes<'a> {
-    text: RopeSlice<'a>,
-    chunks: Chunks<'a>,
-    cur_chunk: &'a str,
-    cur_chunk_start: usize,
-    cursor: GraphemeCursor,
-}
-
-impl<'a> fmt::Debug for RevRopeGraphemes<'a> {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
-        f.debug_struct("RevRopeGraphemes")
-            .field("text", &self.text)
-            .field("chunks", &self.chunks)
-            .field("cur_chunk", &self.cur_chunk)
-            .field("cur_chunk_start", &self.cur_chunk_start)
-            // .field("cursor", &self.cursor)
-            .finish()
-    }
-}
-
-impl<'a> RevRopeGraphemes<'a> {
-    #[must_use]
-    pub fn new(slice: RopeSlice) -> RevRopeGraphemes {
-        let (mut chunks, mut cur_chunk_start, _, _) = slice.chunks_at_byte(slice.len_bytes());
-        chunks.reverse();
-        let first_chunk = chunks.next().unwrap_or("");
-        cur_chunk_start -= first_chunk.len();
-        RevRopeGraphemes {
-            text: slice,
-            chunks,
-            cur_chunk: first_chunk,
-            cur_chunk_start,
-            cursor: GraphemeCursor::new(slice.len_bytes(), slice.len_bytes(), true),
-        }
-    }
-}
-
-impl<'a> Iterator for RevRopeGraphemes<'a> {
-    type Item = RopeSlice<'a>;
-
-    fn next(&mut self) -> Option<RopeSlice<'a>> {
-        let a = self.cursor.cur_cursor();
-        let b;
-        loop {
-            match self
-                .cursor
-                .prev_boundary(self.cur_chunk, self.cur_chunk_start)
-            {
-                Ok(None) => {
-                    return None;
-                }
-                Ok(Some(n)) => {
-                    b = n;
-                    break;
-                }
-                Err(GraphemeIncomplete::PrevChunk) => {
-                    self.cur_chunk = self.chunks.next().unwrap_or("");
-                    self.cur_chunk_start -= self.cur_chunk.len();
-                }
-                Err(GraphemeIncomplete::PreContext(idx)) => {
-                    let (chunk, byte_idx, _, _) = self.text.chunk_at_byte(idx.saturating_sub(1));
-                    self.cursor.provide_context(chunk, byte_idx);
-                }
-                _ => unreachable!(),
-            }
-        }
-
-        if a >= self.cur_chunk_start + self.cur_chunk.len() {
-            Some(self.text.byte_slice(b..a))
-        } else {
-            let a2 = a - self.cur_chunk_start;
-            let b2 = b - self.cur_chunk_start;
-            Some((&self.cur_chunk[b2..a2]).into())
-        }
-    }
-}
-
 /// A highly compressed Cow<'a, str> that holds
 /// atmost u31::MAX bytes and is readonly
 pub struct GraphemeStr<'a> {
@@ -537,7 +292,7 @@ impl<'a> From<&'a str> for GraphemeStr<'a> {
     }
 }
 
-impl<'a> From<String> for GraphemeStr<'a> {
+impl From<String> for GraphemeStr<'_> {
     fn from(g: String) -> Self {
         let len = g.len();
         let ptr = Box::into_raw(g.into_bytes().into_boxed_slice()) as *mut u8;

helix-core/src/indent.rs

@@ -1,14 +1,18 @@
-use std::{borrow::Cow, collections::HashMap};
+use std::{borrow::Cow, collections::HashMap, iter};
 
 use helix_stdx::rope::RopeSliceExt;
-use tree_sitter::{Query, QueryCursor, QueryPredicateArg};
+use tree_house::TREE_SITTER_MATCH_LIMIT;
 
 use crate::{
     chars::{char_is_line_ending, char_is_whitespace},
     graphemes::{grapheme_width, tab_width_at},
-    syntax::{IndentationHeuristic, LanguageConfiguration, RopeProvider, Syntax},
-    tree_sitter::Node,
-    Position, Rope, RopeGraphemes, RopeSlice,
+    syntax::{self, config::IndentationHeuristic},
+    tree_sitter::{
+        self,
+        query::{InvalidPredicateError, UserPredicate},
+        Capture, Grammar, InactiveQueryCursor, Node, Pattern, Query, QueryMatch, RopeInput,
+    },
+    Position, Rope, RopeSlice, Syntax, Tendril,
 };
 
 /// Enum representing indentation style.
@@ -149,6 +153,12 @@ pub fn auto_detect_indent_style(document_text: &Rope) -> Option<IndentStyle> {
         // Give more weight to tabs, because their presence is a very
         // strong indicator.
         histogram[0] *= 2;
+        // Gives less weight to single indent, as single spaces are
+        // often used in certain languages' comment systems and rarely
+        // used as the actual document indentation.
+        if histogram[1] > 1 {
+            histogram[1] /= 2;
+        }
 
         histogram
     };
@@ -200,7 +210,7 @@ pub fn indent_level_for_line(line: RopeSlice, tab_width: usize, indent_width: us
 /// Create a string of tabs & spaces that has the same visual width as the given RopeSlice (independent of the tab width).
 fn whitespace_with_same_width(text: RopeSlice) -> String {
     let mut s = String::new();
-    for grapheme in RopeGraphemes::new(text) {
+    for grapheme in text.graphemes() {
         if grapheme == "\t" {
             s.push('\t');
         } else {
@@ -210,6 +220,36 @@ fn whitespace_with_same_width(text: RopeSlice) -> String {
     s
 }
 
+/// normalizes indentation to tabs/spaces based on user configuration
+/// This function does not change the actual indentation width, just the character
+/// composition.
+pub fn normalize_indentation(
+    prefix: RopeSlice<'_>,
+    line: RopeSlice<'_>,
+    dst: &mut Tendril,
+    indent_style: IndentStyle,
+    tab_width: usize,
+) -> usize {
+    #[allow(deprecated)]
+    let off = crate::visual_coords_at_pos(prefix, prefix.len_chars(), tab_width).col;
+    let mut len = 0;
+    let mut original_len = 0;
+    for ch in line.chars() {
+        match ch {
+            '\t' => len += tab_width_at(len + off, tab_width as u16),
+            ' ' => len += 1,
+            _ => break,
+        }
+        original_len += 1;
+    }
+    if indent_style == IndentStyle::Tabs {
+        dst.extend(iter::repeat('\t').take(len / tab_width));
+        len %= tab_width;
+    }
+    dst.extend(iter::repeat(' ').take(len));
+    original_len
+}
+
 fn add_indent_level(
     mut base_indent: String,
     added_indent_level: isize,
@@ -249,23 +289,169 @@ fn add_indent_level(
 
 /// Return true if only whitespace comes before the node on its line.
 /// If given, new_line_byte_pos is treated the same way as any existing newline.
-fn is_first_in_line(node: Node, text: RopeSlice, new_line_byte_pos: Option<usize>) -> bool {
-    let mut line_start_byte_pos = text.line_to_byte(node.start_position().row);
+fn is_first_in_line(node: &Node, text: RopeSlice, new_line_byte_pos: Option<u32>) -> bool {
+    let line = text.byte_to_line(node.start_byte() as usize);
+    let mut line_start_byte_pos = text.line_to_byte(line) as u32;
     if let Some(pos) = new_line_byte_pos {
         if line_start_byte_pos < pos && pos <= node.start_byte() {
             line_start_byte_pos = pos;
         }
     }
-    text.byte_slice(line_start_byte_pos..node.start_byte())
+    text.byte_slice(line_start_byte_pos as usize..node.start_byte() as usize)
         .chars()
         .all(|c| c.is_whitespace())
 }
 
+#[derive(Debug, Default)]
+pub struct IndentQueryPredicates {
+    not_kind_eq: Vec<(Capture, Box<str>)>,
+    same_line: Option<(Capture, Capture, bool)>,
+    one_line: Option<(Capture, bool)>,
+}
+
+impl IndentQueryPredicates {
+    fn are_satisfied(
+        &self,
+        match_: &QueryMatch,
+        text: RopeSlice,
+        new_line_byte_pos: Option<u32>,
+    ) -> bool {
+        for (capture, not_expected_kind) in self.not_kind_eq.iter() {
+            let node = match_.nodes_for_capture(*capture).next();
+            if node.is_some_and(|n| n.kind() == not_expected_kind.as_ref()) {
+                return false;
+            }
+        }
+
+        if let Some((capture1, capture2, negated)) = self.same_line {
+            let n1 = match_.nodes_for_capture(capture1).next();
+            let n2 = match_.nodes_for_capture(capture2).next();
+            let satisfied = n1.zip(n2).is_some_and(|(n1, n2)| {
+                let n1_line = get_node_start_line(text, n1, new_line_byte_pos);
+                let n2_line = get_node_start_line(text, n2, new_line_byte_pos);
+                let same_line = n1_line == n2_line;
+                same_line != negated
+            });
+
+            if !satisfied {
+                return false;
+            }
+        }
+
+        if let Some((capture, negated)) = self.one_line {
+            let node = match_.nodes_for_capture(capture).next();
+            let satisfied = node.is_some_and(|node| {
+                let start_line = get_node_start_line(text, node, new_line_byte_pos);
+                let end_line = get_node_end_line(text, node, new_line_byte_pos);
+                let one_line = end_line == start_line;
+                one_line != negated
+            });
+
+            if !satisfied {
+                return false;
+            }
+        }
+
+        true
+    }
+}
+
+#[derive(Debug)]
+pub struct IndentQuery {
+    query: Query,
+    properties: HashMap<Pattern, IndentScope>,
+    predicates: HashMap<Pattern, IndentQueryPredicates>,
+    indent_capture: Option<Capture>,
+    indent_always_capture: Option<Capture>,
+    outdent_capture: Option<Capture>,
+    outdent_always_capture: Option<Capture>,
+    align_capture: Option<Capture>,
+    anchor_capture: Option<Capture>,
+    extend_capture: Option<Capture>,
+    extend_prevent_once_capture: Option<Capture>,
+}
+
+impl IndentQuery {
+    pub fn new(grammar: Grammar, source: &str) -> Result<Self, tree_sitter::query::ParseError> {
+        let mut properties = HashMap::new();
+        let mut predicates: HashMap<Pattern, IndentQueryPredicates> = HashMap::new();
+        let query = Query::new(grammar, source, |pattern, predicate| match predicate {
+            UserPredicate::SetProperty { key: "scope", val } => {
+                let scope = match val {
+                    Some("all") => IndentScope::All,
+                    Some("tail") => IndentScope::Tail,
+                    Some(other) => {
+                        return Err(format!("unknown scope (#set! scope \"{other}\")").into())
+                    }
+                    None => return Err("missing scope value (#set! scope ...)".into()),
+                };
+
+                properties.insert(pattern, scope);
+
+                Ok(())
+            }
+            UserPredicate::Other(predicate) => {
+                let name = predicate.name();
+                match name {
+                    "not-kind-eq?" => {
+                        predicate.check_arg_count(2)?;
+                        let capture = predicate.capture_arg(0)?;
+                        let not_expected_kind = predicate.str_arg(1)?;
+
+                        predicates
+                            .entry(pattern)
+                            .or_default()
+                            .not_kind_eq
+                            .push((capture, not_expected_kind.into()));
+                        Ok(())
+                    }
+                    "same-line?" | "not-same-line?" => {
+                        predicate.check_arg_count(2)?;
+                        let capture1 = predicate.capture_arg(0)?;
+                        let capture2 = predicate.capture_arg(1)?;
+                        let negated = name == "not-same-line?";
+
+                        predicates.entry(pattern).or_default().same_line =
+                            Some((capture1, capture2, negated));
+                        Ok(())
+                    }
+                    "one-line?" | "not-one-line?" => {
+                        predicate.check_arg_count(1)?;
+                        let capture = predicate.capture_arg(0)?;
+                        let negated = name == "not-one-line?";
+
+                        predicates.entry(pattern).or_default().one_line = Some((capture, negated));
+                        Ok(())
+                    }
+                    _ => Err(InvalidPredicateError::unknown(UserPredicate::Other(
+                        predicate,
+                    ))),
+                }
+            }
+            _ => Err(InvalidPredicateError::unknown(predicate)),
+        })?;
+
+        Ok(Self {
+            properties,
+            predicates,
+            indent_capture: query.get_capture("indent"),
+            indent_always_capture: query.get_capture("indent.always"),
+            outdent_capture: query.get_capture("outdent"),
+            outdent_always_capture: query.get_capture("outdent.always"),
+            align_capture: query.get_capture("align"),
+            anchor_capture: query.get_capture("anchor"),
+            extend_capture: query.get_capture("extend"),
+            extend_prevent_once_capture: query.get_capture("extend.prevent-once"),
+            query,
+        })
+    }
+}
+
 /// The total indent for some line of code.
 /// This is usually constructed in one of 2 ways:
 /// - Successively add indent captures to get the (added) indent from a single line
 /// - Successively add the indent results for each line
-/// The string that this indentation defines starts with the string contained in the align field (unless it is None), followed by:
+///   The string that this indentation defines starts with the string contained in the align field (unless it is None), followed by:
 /// - max(0, indent - outdent) tabs, if tabs are used for indentation
 /// - max(0, indent - outdent)*indent_width spaces, if spaces are used for indentation
 #[derive(Default, Debug, PartialEq, Eq, Clone)]
@@ -386,7 +572,7 @@ enum IndentCaptureType<'a> {
     Align(RopeSlice<'a>),
 }
 
-impl<'a> IndentCaptureType<'a> {
+impl IndentCaptureType<'_> {
     fn default_scope(&self) -> IndentScope {
         match self {
             IndentCaptureType::Indent | IndentCaptureType::IndentAlways => IndentScope::Tail,
@@ -423,193 +609,114 @@ struct IndentQueryResult<'a> {
     extend_captures: HashMap<usize, Vec<ExtendCapture>>,
 }
 
-fn get_node_start_line(node: Node, new_line_byte_pos: Option<usize>) -> usize {
-    let mut node_line = node.start_position().row;
+fn get_node_start_line(text: RopeSlice, node: &Node, new_line_byte_pos: Option<u32>) -> usize {
+    let mut node_line = text.byte_to_line(node.start_byte() as usize);
     // Adjust for the new line that will be inserted
-    if new_line_byte_pos.map_or(false, |pos| node.start_byte() >= pos) {
+    if new_line_byte_pos.is_some_and(|pos| node.start_byte() >= pos) {
         node_line += 1;
     }
     node_line
 }
-fn get_node_end_line(node: Node, new_line_byte_pos: Option<usize>) -> usize {
-    let mut node_line = node.end_position().row;
+fn get_node_end_line(text: RopeSlice, node: &Node, new_line_byte_pos: Option<u32>) -> usize {
+    let mut node_line = text.byte_to_line(node.end_byte() as usize);
     // Adjust for the new line that will be inserted (with a strict inequality since end_byte is exclusive)
-    if new_line_byte_pos.map_or(false, |pos| node.end_byte() > pos) {
+    if new_line_byte_pos.is_some_and(|pos| node.end_byte() > pos) {
         node_line += 1;
     }
     node_line
 }
 
 fn query_indents<'a>(
-    query: &Query,
+    query: &IndentQuery,
     syntax: &Syntax,
-    cursor: &mut QueryCursor,
     text: RopeSlice<'a>,
-    range: std::ops::Range<usize>,
-    new_line_byte_pos: Option<usize>,
+    range: std::ops::Range<u32>,
+    new_line_byte_pos: Option<u32>,
 ) -> IndentQueryResult<'a> {
     let mut indent_captures: HashMap<usize, Vec<IndentCapture>> = HashMap::new();
     let mut extend_captures: HashMap<usize, Vec<ExtendCapture>> = HashMap::new();
-    cursor.set_byte_range(range);
+
+    let mut cursor = InactiveQueryCursor::new(range, TREE_SITTER_MATCH_LIMIT).execute_query(
+        &query.query,
+        &syntax.tree().root_node(),
+        RopeInput::new(text),
+    );
 
     // Iterate over all captures from the query
-    for m in cursor.matches(query, syntax.tree().root_node(), RopeProvider(text)) {
+    while let Some(m) = cursor.next_match() {
         // Skip matches where not all custom predicates are fulfilled
-        if !query.general_predicates(m.pattern_index).iter().all(|pred| {
-            match pred.operator.as_ref() {
-                "not-kind-eq?" => match (pred.args.get(0), pred.args.get(1)) {
-                    (
-                        Some(QueryPredicateArg::Capture(capture_idx)),
-                        Some(QueryPredicateArg::String(kind)),
-                    ) => {
-                        let node = m.nodes_for_capture_index(*capture_idx).next();
-                        match node {
-                            Some(node) => node.kind()!=kind.as_ref(),
-                            _ => true,
-                        }
-                    }
-                    _ => {
-                        panic!("Invalid indent query: Arguments to \"not-kind-eq?\" must be a capture and a string");
-                    }
-                },
-                "same-line?" | "not-same-line?" => {
-                    match (pred.args.get(0), pred.args.get(1)) {
-                        (
-                            Some(QueryPredicateArg::Capture(capt1)),
-                            Some(QueryPredicateArg::Capture(capt2))
-                        ) => {
-                            let n1 = m.nodes_for_capture_index(*capt1).next();
-                            let n2 = m.nodes_for_capture_index(*capt2).next();
-                            match (n1, n2) {
-                                (Some(n1), Some(n2)) => {
-                                    let n1_line = get_node_start_line(n1, new_line_byte_pos);
-                                    let n2_line = get_node_start_line(n2, new_line_byte_pos);
-                                    let same_line = n1_line == n2_line;
-                                    same_line==(pred.operator.as_ref()=="same-line?")
-                                }
-                                _ => true,
-                            }
-                        }
-                        _ => {
-                            panic!("Invalid indent query: Arguments to \"{}\" must be 2 captures", pred.operator);
-                        }
-                    }
-                }
-                "one-line?" | "not-one-line?" => match pred.args.get(0) {
-                    Some(QueryPredicateArg::Capture(capture_idx)) => {
-                        let node = m.nodes_for_capture_index(*capture_idx).next();
-
-                        match node {
-                            Some(node) => {
-                                let (start_line, end_line) = (get_node_start_line(node,new_line_byte_pos), get_node_end_line(node, new_line_byte_pos));
-                                let one_line = end_line == start_line;
-                                one_line != (pred.operator.as_ref() == "not-one-line?")
-                            },
-                            _ => true,
-                        }
-                    }
-                    _ => {
-                        panic!("Invalid indent query: Arguments to \"not-kind-eq?\" must be a capture and a string");
-                    }
-                },
-                _ => {
-                    panic!(
-                        "Invalid indent query: Unknown predicate (\"{}\")",
-                        pred.operator
-                    );
-                }
-            }
-        }) {
+        if query
+            .predicates
+            .get(&m.pattern())
+            .is_some_and(|preds| !preds.are_satisfied(&m, text, new_line_byte_pos))
+        {
             continue;
         }
         // A list of pairs (node_id, indent_capture) that are added by this match.
         // They cannot be added to indent_captures immediately since they may depend on other captures (such as an @anchor).
         let mut added_indent_captures: Vec<(usize, IndentCapture)> = Vec::new();
         // The row/column position of the optional anchor in this query
-        let mut anchor: Option<tree_sitter::Node> = None;
-        for capture in m.captures {
-            let capture_name = query.capture_names()[capture.index as usize];
-            let capture_type = match capture_name {
-                "indent" => IndentCaptureType::Indent,
-                "indent.always" => IndentCaptureType::IndentAlways,
-                "outdent" => IndentCaptureType::Outdent,
-                "outdent.always" => IndentCaptureType::OutdentAlways,
-                // The alignment will be updated to the correct value at the end, when the anchor is known.
-                "align" => IndentCaptureType::Align(RopeSlice::from("")),
-                "anchor" => {
-                    if anchor.is_some() {
-                        log::error!("Invalid indent query: Encountered more than one @anchor in the same match.")
-                    } else {
-                        anchor = Some(capture.node);
-                    }
-                    continue;
-                }
-                "extend" => {
-                    extend_captures
-                        .entry(capture.node.id())
-                        .or_insert_with(|| Vec::with_capacity(1))
-                        .push(ExtendCapture::Extend);
-                    continue;
-                }
-                "extend.prevent-once" => {
-                    extend_captures
-                        .entry(capture.node.id())
-                        .or_insert_with(|| Vec::with_capacity(1))
-                        .push(ExtendCapture::PreventOnce);
-                    continue;
-                }
-                _ => {
-                    // Ignore any unknown captures (these may be needed for predicates such as #match?)
-                    continue;
+        let mut anchor: Option<&Node> = None;
+        for matched_node in m.matched_nodes() {
+            let node_id = matched_node.node.id();
+            let capture = Some(matched_node.capture);
+            let capture_type = if capture == query.indent_capture {
+                IndentCaptureType::Indent
+            } else if capture == query.indent_always_capture {
+                IndentCaptureType::IndentAlways
+            } else if capture == query.outdent_capture {
+                IndentCaptureType::Outdent
+            } else if capture == query.outdent_always_capture {
+                IndentCaptureType::OutdentAlways
+            } else if capture == query.align_capture {
+                IndentCaptureType::Align(RopeSlice::from(""))
+            } else if capture == query.anchor_capture {
+                if anchor.is_some() {
+                    log::error!("Invalid indent query: Encountered more than one @anchor in the same match.")
+                } else {
+                    anchor = Some(&matched_node.node);
                 }
+                continue;
+            } else if capture == query.extend_capture {
+                extend_captures
+                    .entry(node_id)
+                    .or_insert_with(|| Vec::with_capacity(1))
+                    .push(ExtendCapture::Extend);
+                continue;
+            } else if capture == query.extend_prevent_once_capture {
+                extend_captures
+                    .entry(node_id)
+                    .or_insert_with(|| Vec::with_capacity(1))
+                    .push(ExtendCapture::PreventOnce);
+                continue;
+            } else {
+                // Ignore any unknown captures (these may be needed for predicates such as #match?)
+                continue;
             };
-            let scope = capture_type.default_scope();
-            let mut indent_capture = IndentCapture {
+
+            // Apply additional settings for this capture
+            let scope = query
+                .properties
+                .get(&m.pattern())
+                .copied()
+                .unwrap_or_else(|| capture_type.default_scope());
+            let indent_capture = IndentCapture {
                 capture_type,
                 scope,
             };
-            // Apply additional settings for this capture
-            for property in query.property_settings(m.pattern_index) {
-                match property.key.as_ref() {
-                    "scope" => {
-                        indent_capture.scope = match property.value.as_deref() {
-                            Some("all") => IndentScope::All,
-                            Some("tail") => IndentScope::Tail,
-                            Some(s) => {
-                                panic!("Invalid indent query: Unknown value for \"scope\" property (\"{}\")", s);
-                            }
-                            None => {
-                                panic!(
-                                    "Invalid indent query: Missing value for \"scope\" property"
-                                );
-                            }
-                        }
-                    }
-                    _ => {
-                        panic!(
-                            "Invalid indent query: Unknown property \"{}\"",
-                            property.key
-                        );
-                    }
-                }
-            }
-            added_indent_captures.push((capture.node.id(), indent_capture))
+            added_indent_captures.push((node_id, indent_capture))
         }
         for (node_id, mut capture) in added_indent_captures {
             // Set the anchor for all align queries.
             if let IndentCaptureType::Align(_) = capture.capture_type {
-                let anchor = match anchor {
-                    None => {
-                        log::error!(
-                            "Invalid indent query: @align requires an accompanying @anchor."
-                        );
-                        continue;
-                    }
-                    Some(anchor) => anchor,
+                let Some(anchor) = anchor else {
+                    log::error!("Invalid indent query: @align requires an accompanying @anchor.");
+                    continue;
                 };
+                let line = text.byte_to_line(anchor.start_byte() as usize);
+                let line_start = text.line_to_byte(line);
                 capture.capture_type = IndentCaptureType::Align(
-                    text.line(anchor.start_position().row)
-                        .byte_slice(0..anchor.start_position().column),
+                    text.byte_slice(line_start..anchor.start_byte() as usize),
                 );
             }
             indent_captures
@@ -661,13 +768,15 @@ fn extend_nodes<'a>(
                         // - the cursor is on the same line as the end of the node OR
                         // - the line that the cursor is on is more indented than the
                         //   first line of the node
-                        if deepest_preceding.end_position().row == line {
+                        if text.byte_to_line(deepest_preceding.end_byte() as usize) == line {
                             extend_node = true;
                         } else {
                             let cursor_indent =
                                 indent_level_for_line(text.line(line), tab_width, indent_width);
                             let node_indent = indent_level_for_line(
-                                text.line(deepest_preceding.start_position().row),
+                                text.line(
+                                    text.byte_to_line(deepest_preceding.start_byte() as usize),
+                                ),
                                 tab_width,
                                 indent_width,
                             );
@@ -684,7 +793,7 @@ fn extend_nodes<'a>(
         if node_captured && stop_extend {
             stop_extend = false;
         } else if extend_node && !stop_extend {
-            *node = deepest_preceding;
+            *node = deepest_preceding.clone();
             break;
         }
         // If the tree contains a syntax error, `deepest_preceding` may not
@@ -701,17 +810,17 @@ fn extend_nodes<'a>(
 /// - The indent captures for all relevant nodes.
 #[allow(clippy::too_many_arguments)]
 fn init_indent_query<'a, 'b>(
-    query: &Query,
+    query: &IndentQuery,
     syntax: &'a Syntax,
     text: RopeSlice<'b>,
     tab_width: usize,
     indent_width: usize,
     line: usize,
-    byte_pos: usize,
-    new_line_byte_pos: Option<usize>,
+    byte_pos: u32,
+    new_line_byte_pos: Option<u32>,
 ) -> Option<(Node<'a>, HashMap<usize, Vec<IndentCapture<'b>>>)> {
     // The innermost tree-sitter node which is considered for the indent
-    // computation. It may change if some predeceding node is extended
+    // computation. It may change if some preceding node is extended
     let mut node = syntax
         .tree()
         .root_node()
@@ -721,37 +830,25 @@ fn init_indent_query<'a, 'b>(
         // The query range should intersect with all nodes directly preceding
         // the position of the indent query in case one of them is extended.
         let mut deepest_preceding = None; // The deepest node preceding the indent query position
-        let mut tree_cursor = node.walk();
-        for child in node.children(&mut tree_cursor) {
+        for child in node.children() {
             if child.byte_range().end <= byte_pos {
-                deepest_preceding = Some(child);
+                deepest_preceding = Some(child.clone());
             }
         }
         deepest_preceding = deepest_preceding.map(|mut prec| {
             // Get the deepest directly preceding node
             while prec.child_count() > 0 {
-                prec = prec.child(prec.child_count() - 1).unwrap();
+                prec = prec.child(prec.child_count() - 1).unwrap().clone();
             }
             prec
         });
         let query_range = deepest_preceding
+            .as_ref()
             .map(|prec| prec.byte_range().end - 1..byte_pos + 1)
             .unwrap_or(byte_pos..byte_pos + 1);
 
-        crate::syntax::PARSER.with(|ts_parser| {
-            let mut ts_parser = ts_parser.borrow_mut();
-            let mut cursor = ts_parser.cursors.pop().unwrap_or_default();
-            let query_result = query_indents(
-                query,
-                syntax,
-                &mut cursor,
-                text,
-                query_range,
-                new_line_byte_pos,
-            );
-            ts_parser.cursors.push(cursor);
-            (query_result, deepest_preceding)
-        })
+        let query_result = query_indents(query, syntax, text, query_range, new_line_byte_pos);
+        (query_result, deepest_preceding)
     };
     let extend_captures = query_result.extend_captures;
 
@@ -786,6 +883,7 @@ fn init_indent_query<'a, 'b>(
 /// - The line after the node. This is defined by:
 ///   - The scope `tail`.
 ///   - The scope `all` if this node is not the first node on its line.
+///
 /// Intuitively, `all` applies to everything contained in this node while `tail` applies to everything except for the first line of the node.
 /// The indents from different nodes for the same line are then combined.
 /// The result [Indentation] is simply the sum of the [Indentation] for all lines.
@@ -808,7 +906,7 @@ fn init_indent_query<'a, 'b>(
 /// ```
 #[allow(clippy::too_many_arguments)]
 pub fn treesitter_indent_for_pos<'a>(
-    query: &Query,
+    query: &IndentQuery,
     syntax: &Syntax,
     tab_width: usize,
     indent_width: usize,
@@ -817,7 +915,7 @@ pub fn treesitter_indent_for_pos<'a>(
     pos: usize,
     new_line: bool,
 ) -> Option<Indentation<'a>> {
-    let byte_pos = text.char_to_byte(pos);
+    let byte_pos = text.char_to_byte(pos) as u32;
     let new_line_byte_pos = new_line.then_some(byte_pos);
     let (mut node, mut indent_captures) = init_indent_query(
         query,
@@ -837,7 +935,7 @@ pub fn treesitter_indent_for_pos<'a>(
     let mut indent_for_line_below = Indentation::default();
 
     loop {
-        let is_first = is_first_in_line(node, text, new_line_byte_pos);
+        let is_first = is_first_in_line(&node, text, new_line_byte_pos);
 
         // Apply all indent definitions for this node.
         // Since we only iterate over each node once, we can remove the
@@ -860,8 +958,8 @@ pub fn treesitter_indent_for_pos<'a>(
         }
 
         if let Some(parent) = node.parent() {
-            let node_line = get_node_start_line(node, new_line_byte_pos);
-            let parent_line = get_node_start_line(parent, new_line_byte_pos);
+            let node_line = get_node_start_line(text, &node, new_line_byte_pos);
+            let parent_line = get_node_start_line(text, &parent, new_line_byte_pos);
 
             if node_line != parent_line {
                 // Don't add indent for the line below the line of the query
@@ -883,8 +981,9 @@ pub fn treesitter_indent_for_pos<'a>(
         } else {
             // Only add the indentation for the line below if that line
             // is not after the line that the indentation is calculated for.
-            if (node.start_position().row < line)
-                || (new_line && node.start_position().row == line && node.start_byte() < byte_pos)
+            let node_start_line = text.byte_to_line(node.start_byte() as usize);
+            if node_start_line < line
+                || (new_line && node_start_line == line && node.start_byte() < byte_pos)
             {
                 result.add_line(indent_for_line_below);
             }
@@ -899,7 +998,7 @@ pub fn treesitter_indent_for_pos<'a>(
 /// This is done either using treesitter, or if that's not available by copying the indentation from the current line
 #[allow(clippy::too_many_arguments)]
 pub fn indent_for_newline(
-    language_config: Option<&LanguageConfiguration>,
+    loader: &syntax::Loader,
     syntax: Option<&Syntax>,
     indent_heuristic: &IndentationHeuristic,
     indent_style: &IndentStyle,
@@ -916,7 +1015,7 @@ pub fn indent_for_newline(
         Some(syntax),
     ) = (
         indent_heuristic,
-        language_config.and_then(|config| config.indent_query()),
+        syntax.and_then(|syntax| loader.indent_query(syntax.root_language())),
         syntax,
     ) {
         if let Some(indent) = treesitter_indent_for_pos(
@@ -984,10 +1083,10 @@ pub fn indent_for_newline(
     indent_style.as_str().repeat(indent_level)
 }
 
-pub fn get_scopes(syntax: Option<&Syntax>, text: RopeSlice, pos: usize) -> Vec<&'static str> {
+pub fn get_scopes<'a>(syntax: Option<&'a Syntax>, text: RopeSlice, pos: usize) -> Vec<&'a str> {
     let mut scopes = Vec::new();
     if let Some(syntax) = syntax {
-        let pos = text.char_to_byte(pos);
+        let pos = text.char_to_byte(pos) as u32;
         let mut node = match syntax
             .tree()
             .root_node()

helix-core/src/lib.rs

@@ -1,12 +1,16 @@
 pub use encoding_rs as encoding;
 
 pub mod auto_pairs;
+pub mod case_conversion;
 pub mod chars;
+pub mod command_line;
 pub mod comment;
+pub mod completion;
 pub mod config;
 pub mod diagnostic;
 pub mod diff;
 pub mod doc_formatter;
+pub mod editor_config;
 pub mod fuzzy;
 pub mod graphemes;
 pub mod history;
@@ -20,7 +24,7 @@ pub mod object;
 mod position;
 pub mod search;
 pub mod selection;
-pub mod shellwords;
+pub mod snippets;
 pub mod surround;
 pub mod syntax;
 pub mod test;
@@ -49,12 +53,11 @@ pub use smartstring::SmartString;
 pub type Tendril = SmartString<smartstring::LazyCompact>;
 
 #[doc(inline)]
-pub use {regex, tree_sitter};
+pub use {regex, tree_house::tree_sitter};
 
-pub use graphemes::RopeGraphemes;
 pub use position::{
-    char_idx_at_visual_offset, coords_at_pos, pos_at_coords, visual_offset_from_anchor,
-    visual_offset_from_block, Position, VisualOffsetError,
+    char_idx_at_visual_offset, coords_at_pos, pos_at_coords, softwrapped_dimensions,
+    visual_offset_from_anchor, visual_offset_from_block, Position, VisualOffsetError,
 };
 #[allow(deprecated)]
 pub use position::{pos_at_visual_coords, visual_coords_at_pos};
@@ -63,9 +66,12 @@ pub use selection::{Range, Selection};
 pub use smallvec::{smallvec, SmallVec};
 pub use syntax::Syntax;
 
+pub use completion::CompletionItem;
 pub use diagnostic::Diagnostic;
 
 pub use line_ending::{LineEnding, NATIVE_LINE_ENDING};
 pub use transaction::{Assoc, Change, ChangeSet, Deletion, Operation, Transaction};
 
 pub use uri::Uri;
+
+pub use tree_house::Language;

helix-core/src/match_brackets.rs

@@ -1,7 +1,7 @@
 use std::iter;
 
+use crate::tree_sitter::Node;
 use ropey::RopeSlice;
-use tree_sitter::Node;
 
 use crate::movement::Direction::{self, Backward, Forward};
 use crate::Syntax;
@@ -75,7 +75,7 @@ fn find_pair(
     pos_: usize,
     traverse_parents: bool,
 ) -> Option<usize> {
-    let pos = doc.char_to_byte(pos_);
+    let pos = doc.char_to_byte(pos_) as u32;
 
     let root = syntax.tree_for_byte_range(pos, pos).root_node();
     let mut node = root.descendant_for_byte_range(pos, pos)?;
@@ -128,7 +128,7 @@ fn find_pair(
                 if find_pair_end(doc, sibling.prev_sibling(), start_char, end_char, Backward)
                     .is_some()
                 {
-                    return doc.try_byte_to_char(sibling.start_byte()).ok();
+                    return doc.try_byte_to_char(sibling.start_byte() as usize).ok();
                 }
             }
         } else if node.is_named() {
@@ -144,9 +144,9 @@ fn find_pair(
     if node.child_count() != 0 {
         return None;
     }
-    let node_start = doc.byte_to_char(node.start_byte());
-    find_matching_bracket_plaintext(doc.byte_slice(node.byte_range()), pos_ - node_start)
-        .map(|pos| pos + node_start)
+    let node_start = doc.byte_to_char(node.start_byte() as usize);
+    let node_text = doc.byte_slice(node.start_byte() as usize..node.end_byte() as usize);
+    find_matching_bracket_plaintext(node_text, pos_ - node_start).map(|pos| pos + node_start)
 }
 
 /// Returns the position of the matching bracket under cursor.
@@ -304,7 +304,7 @@ fn as_char(doc: RopeSlice, node: &Node) -> Option<(usize, char)> {
     if node.byte_range().len() != 1 {
         return None;
     }
-    let pos = doc.try_byte_to_char(node.start_byte()).ok()?;
+    let pos = doc.try_byte_to_char(node.start_byte() as usize).ok()?;
     Some((pos, doc.char(pos)))
 }
 

helix-core/src/movement.rs

@@ -1,7 +1,6 @@
-use std::{cmp::Reverse, iter};
+use std::{borrow::Cow, cmp::Reverse, iter};
 
 use ropey::iter::Chars;
-use tree_sitter::{Node, QueryCursor};
 
 use crate::{
     char_idx_at_visual_offset,
@@ -13,9 +12,10 @@ use crate::{
     },
     line_ending::rope_is_line_ending,
     position::char_idx_at_visual_block_offset,
-    syntax::LanguageConfiguration,
+    syntax,
     text_annotations::TextAnnotations,
     textobject::TextObject,
+    tree_sitter::Node,
     visual_offset_from_block, Range, RopeSlice, Selection, Syntax,
 };
 
@@ -79,19 +79,19 @@ pub fn move_vertically_visual(
         Direction::Backward => -(count as isize),
     };
 
-    // TODO how to handle inline annotations that span an entire visual line (very unlikely).
-
     // Compute visual offset relative to block start to avoid trasversing the block twice
     row_off += visual_pos.row as isize;
-    let new_pos = char_idx_at_visual_offset(
+    let (mut new_pos, virtual_rows) = char_idx_at_visual_offset(
         slice,
         block_off,
         row_off,
         new_col as usize,
         text_fmt,
         annotations,
-    )
-    .0;
+    );
+    if dir == Direction::Forward {
+        new_pos += (virtual_rows != 0) as usize;
+    }
 
     // Special-case to avoid moving to the end of the last non-empty line.
     if behaviour == Movement::Extend && slice.line(slice.char_to_line(new_pos)).len_chars() == 0 {
@@ -197,13 +197,31 @@ pub fn move_prev_long_word_end(slice: RopeSlice, range: Range, count: usize) ->
     word_move(slice, range, count, WordMotionTarget::PrevLongWordEnd)
 }
 
+pub fn move_next_sub_word_start(slice: RopeSlice, range: Range, count: usize) -> Range {
+    word_move(slice, range, count, WordMotionTarget::NextSubWordStart)
+}
+
+pub fn move_next_sub_word_end(slice: RopeSlice, range: Range, count: usize) -> Range {
+    word_move(slice, range, count, WordMotionTarget::NextSubWordEnd)
+}
+
+pub fn move_prev_sub_word_start(slice: RopeSlice, range: Range, count: usize) -> Range {
+    word_move(slice, range, count, WordMotionTarget::PrevSubWordStart)
+}
+
+pub fn move_prev_sub_word_end(slice: RopeSlice, range: Range, count: usize) -> Range {
+    word_move(slice, range, count, WordMotionTarget::PrevSubWordEnd)
+}
+
 fn word_move(slice: RopeSlice, range: Range, count: usize, target: WordMotionTarget) -> Range {
     let is_prev = matches!(
         target,
         WordMotionTarget::PrevWordStart
             | WordMotionTarget::PrevLongWordStart
+            | WordMotionTarget::PrevSubWordStart
             | WordMotionTarget::PrevWordEnd
             | WordMotionTarget::PrevLongWordEnd
+            | WordMotionTarget::PrevSubWordEnd
     );
 
     // Special-case early-out.
@@ -383,6 +401,12 @@ pub enum WordMotionTarget {
     NextLongWordEnd,
     PrevLongWordStart,
     PrevLongWordEnd,
+    // A sub word is similar to a regular word, except it is also delimited by
+    // underscores and transitions from lowercase to uppercase.
+    NextSubWordStart,
+    NextSubWordEnd,
+    PrevSubWordStart,
+    PrevSubWordEnd,
 }
 
 pub trait CharHelpers {
@@ -398,8 +422,10 @@ impl CharHelpers for Chars<'_> {
             target,
             WordMotionTarget::PrevWordStart
                 | WordMotionTarget::PrevLongWordStart
+                | WordMotionTarget::PrevSubWordStart
                 | WordMotionTarget::PrevWordEnd
                 | WordMotionTarget::PrevLongWordEnd
+                | WordMotionTarget::PrevSubWordEnd
         );
 
         // Reverse the iterator if needed for the motion direction.
@@ -476,6 +502,25 @@ fn is_long_word_boundary(a: char, b: char) -> bool {
     }
 }
 
+fn is_sub_word_boundary(a: char, b: char, dir: Direction) -> bool {
+    match (categorize_char(a), categorize_char(b)) {
+        (CharCategory::Word, CharCategory::Word) => {
+            if (a == '_') != (b == '_') {
+                return true;
+            }
+
+            // Subword boundaries are directional: in 'fooBar', there is a
+            // boundary between 'o' and 'B', but not between 'B' and 'a'.
+            match dir {
+                Direction::Forward => a.is_lowercase() && b.is_uppercase(),
+                Direction::Backward => a.is_uppercase() && b.is_lowercase(),
+            }
+        }
+        (a, b) if a != b => true,
+        _ => false,
+    }
+}
+
 fn reached_target(target: WordMotionTarget, prev_ch: char, next_ch: char) -> bool {
     match target {
         WordMotionTarget::NextWordStart | WordMotionTarget::PrevWordEnd => {
@@ -494,26 +539,44 @@ fn reached_target(target: WordMotionTarget, prev_ch: char, next_ch: char) -> boo
             is_long_word_boundary(prev_ch, next_ch)
                 && (!prev_ch.is_whitespace() || char_is_line_ending(next_ch))
         }
+        WordMotionTarget::NextSubWordStart => {
+            is_sub_word_boundary(prev_ch, next_ch, Direction::Forward)
+                && (char_is_line_ending(next_ch) || !(next_ch.is_whitespace() || next_ch == '_'))
+        }
+        WordMotionTarget::PrevSubWordEnd => {
+            is_sub_word_boundary(prev_ch, next_ch, Direction::Backward)
+                && (char_is_line_ending(next_ch) || !(next_ch.is_whitespace() || next_ch == '_'))
+        }
+        WordMotionTarget::NextSubWordEnd => {
+            is_sub_word_boundary(prev_ch, next_ch, Direction::Forward)
+                && (!(prev_ch.is_whitespace() || prev_ch == '_') || char_is_line_ending(next_ch))
+        }
+        WordMotionTarget::PrevSubWordStart => {
+            is_sub_word_boundary(prev_ch, next_ch, Direction::Backward)
+                && (!(prev_ch.is_whitespace() || prev_ch == '_') || char_is_line_ending(next_ch))
+        }
     }
 }
 
 /// Finds the range of the next or previous textobject in the syntax sub-tree of `node`.
 /// Returns the range in the forwards direction.
+#[allow(clippy::too_many_arguments)]
 pub fn goto_treesitter_object(
     slice: RopeSlice,
     range: Range,
     object_name: &str,
     dir: Direction,
-    slice_tree: Node,
-    lang_config: &LanguageConfiguration,
+    slice_tree: &Node,
+    syntax: &Syntax,
+    loader: &syntax::Loader,
     count: usize,
 ) -> Range {
+    let textobject_query = loader.textobject_query(syntax.root_language());
     let get_range = move |range: Range| -> Option<Range> {
         let byte_pos = slice.char_to_byte(range.cursor(slice));
 
         let cap_name = |t: TextObject| format!("{}.{}", object_name, t);
-        let mut cursor = QueryCursor::new();
-        let nodes = lang_config.textobject_query()?.capture_nodes_any(
+        let nodes = textobject_query?.capture_nodes_any(
             &[
                 &cap_name(TextObject::Movement),
                 &cap_name(TextObject::Around),
@@ -521,7 +584,6 @@ pub fn goto_treesitter_object(
             ],
             slice_tree,
             slice,
-            &mut cursor,
         )?;
 
         let node = match dir {
@@ -556,14 +618,15 @@ pub fn goto_treesitter_object(
     last_range
 }
 
-fn find_parent_start(mut node: Node) -> Option<Node> {
+fn find_parent_start<'tree>(node: &Node<'tree>) -> Option<Node<'tree>> {
     let start = node.start_byte();
+    let mut node = Cow::Borrowed(node);
 
     while node.start_byte() >= start || !node.is_named() {
-        node = node.parent()?;
+        node = Cow::Owned(node.parent()?);
     }
 
-    Some(node)
+    Some(node.into_owned())
 }
 
 pub fn move_parent_node_end(
@@ -574,8 +637,8 @@ pub fn move_parent_node_end(
     movement: Movement,
 ) -> Selection {
     selection.transform(|range| {
-        let start_from = text.char_to_byte(range.from());
-        let start_to = text.char_to_byte(range.to());
+        let start_from = text.char_to_byte(range.from()) as u32;
+        let start_to = text.char_to_byte(range.to()) as u32;
 
         let mut node = match syntax.named_descendant_for_byte_range(start_from, start_to) {
             Some(node) => node,
@@ -593,18 +656,18 @@ pub fn move_parent_node_end(
             // moving forward, we always want to move one past the end of the
             // current node, so use the end byte of the current node, which is an exclusive
             // end of the range
-            Direction::Forward => text.byte_to_char(node.end_byte()),
+            Direction::Forward => text.byte_to_char(node.end_byte() as usize),
 
             // moving backward, we want the cursor to land on the start char of
             // the current node, or if it is already at the start of a node, to traverse up to
             // the parent
             Direction::Backward => {
-                let end_head = text.byte_to_char(node.start_byte());
+                let end_head = text.byte_to_char(node.start_byte() as usize);
 
                 // if we're already on the beginning, look up to the parent
                 if end_head == range.cursor(text) {
-                    node = find_parent_start(node).unwrap_or(node);
-                    text.byte_to_char(node.start_byte())
+                    node = find_parent_start(&node).unwrap_or(node);
+                    text.byte_to_char(node.start_byte() as usize)
                 } else {
                     end_head
                 }
@@ -1012,6 +1075,178 @@ mod test {
         }
     }
 
+    #[test]
+    fn test_behaviour_when_moving_to_start_of_next_sub_words() {
+        let tests = [
+            (
+                "NextSubwordStart",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 11)),
+                ],
+            ),
+            (
+                "next_subword_start",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 5)),
+                    (1, Range::new(4, 4), Range::new(5, 13)),
+                ],
+            ),
+            (
+                "Next_Subword_Start",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 5)),
+                    (1, Range::new(4, 4), Range::new(5, 13)),
+                ],
+            ),
+            (
+                "NEXT_SUBWORD_START",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 5)),
+                    (1, Range::new(4, 4), Range::new(5, 13)),
+                ],
+            ),
+            (
+                "next subword start",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 5)),
+                    (1, Range::new(4, 4), Range::new(5, 13)),
+                ],
+            ),
+            (
+                "Next Subword Start",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 5)),
+                    (1, Range::new(4, 4), Range::new(5, 13)),
+                ],
+            ),
+            (
+                "NEXT SUBWORD START",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 5)),
+                    (1, Range::new(4, 4), Range::new(5, 13)),
+                ],
+            ),
+            (
+                "next__subword__start",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 6)),
+                    (1, Range::new(4, 4), Range::new(4, 6)),
+                    (1, Range::new(5, 5), Range::new(6, 15)),
+                ],
+            ),
+            (
+                "Next__Subword__Start",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 6)),
+                    (1, Range::new(4, 4), Range::new(4, 6)),
+                    (1, Range::new(5, 5), Range::new(6, 15)),
+                ],
+            ),
+            (
+                "NEXT__SUBWORD__START",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 6)),
+                    (1, Range::new(4, 4), Range::new(4, 6)),
+                    (1, Range::new(5, 5), Range::new(6, 15)),
+                ],
+            ),
+        ];
+
+        for (sample, scenario) in tests {
+            for (count, begin, expected_end) in scenario.into_iter() {
+                let range = move_next_sub_word_start(Rope::from(sample).slice(..), begin, count);
+                assert_eq!(range, expected_end, "Case failed: [{}]", sample);
+            }
+        }
+    }
+
+    #[test]
+    fn test_behaviour_when_moving_to_end_of_next_sub_words() {
+        let tests = [
+            (
+                "NextSubwordEnd",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 11)),
+                ],
+            ),
+            (
+                "next subword end",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 12)),
+                ],
+            ),
+            (
+                "Next Subword End",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 12)),
+                ],
+            ),
+            (
+                "NEXT SUBWORD END",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 12)),
+                ],
+            ),
+            (
+                "next_subword_end",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 12)),
+                ],
+            ),
+            (
+                "Next_Subword_End",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 12)),
+                ],
+            ),
+            (
+                "NEXT_SUBWORD_END",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 12)),
+                ],
+            ),
+            (
+                "next__subword__end",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 13)),
+                    (1, Range::new(5, 5), Range::new(5, 13)),
+                ],
+            ),
+            (
+                "Next__Subword__End",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 13)),
+                    (1, Range::new(5, 5), Range::new(5, 13)),
+                ],
+            ),
+            (
+                "NEXT__SUBWORD__END",
+                vec![
+                    (1, Range::new(0, 0), Range::new(0, 4)),
+                    (1, Range::new(4, 4), Range::new(4, 13)),
+                    (1, Range::new(5, 5), Range::new(5, 13)),
+                ],
+            ),
+        ];
+
+        for (sample, scenario) in tests {
+            for (count, begin, expected_end) in scenario.into_iter() {
+                let range = move_next_sub_word_end(Rope::from(sample).slice(..), begin, count);
+                assert_eq!(range, expected_end, "Case failed: [{}]", sample);
+            }
+        }
+    }
+
     #[test]
     fn test_behaviour_when_moving_to_start_of_next_long_words() {
         let tests = [
@@ -1181,6 +1416,92 @@ mod test {
         }
     }
 
+    #[test]
+    fn test_behaviour_when_moving_to_start_of_previous_sub_words() {
+        let tests = [
+            (
+                "PrevSubwordEnd",
+                vec![
+                    (1, Range::new(13, 13), Range::new(14, 11)),
+                    (1, Range::new(11, 11), Range::new(11, 4)),
+                ],
+            ),
+            (
+                "prev subword end",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 13)),
+                    (1, Range::new(12, 12), Range::new(13, 5)),
+                ],
+            ),
+            (
+                "Prev Subword End",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 13)),
+                    (1, Range::new(12, 12), Range::new(13, 5)),
+                ],
+            ),
+            (
+                "PREV SUBWORD END",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 13)),
+                    (1, Range::new(12, 12), Range::new(13, 5)),
+                ],
+            ),
+            (
+                "prev_subword_end",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 13)),
+                    (1, Range::new(12, 12), Range::new(13, 5)),
+                ],
+            ),
+            (
+                "Prev_Subword_End",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 13)),
+                    (1, Range::new(12, 12), Range::new(13, 5)),
+                ],
+            ),
+            (
+                "PREV_SUBWORD_END",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 13)),
+                    (1, Range::new(12, 12), Range::new(13, 5)),
+                ],
+            ),
+            (
+                "prev__subword__end",
+                vec![
+                    (1, Range::new(17, 17), Range::new(18, 15)),
+                    (1, Range::new(13, 13), Range::new(14, 6)),
+                    (1, Range::new(14, 14), Range::new(15, 6)),
+                ],
+            ),
+            (
+                "Prev__Subword__End",
+                vec![
+                    (1, Range::new(17, 17), Range::new(18, 15)),
+                    (1, Range::new(13, 13), Range::new(14, 6)),
+                    (1, Range::new(14, 14), Range::new(15, 6)),
+                ],
+            ),
+            (
+                "PREV__SUBWORD__END",
+                vec![
+                    (1, Range::new(17, 17), Range::new(18, 15)),
+                    (1, Range::new(13, 13), Range::new(14, 6)),
+                    (1, Range::new(14, 14), Range::new(15, 6)),
+                ],
+            ),
+        ];
+
+        for (sample, scenario) in tests {
+            for (count, begin, expected_end) in scenario.into_iter() {
+                let range = move_prev_sub_word_start(Rope::from(sample).slice(..), begin, count);
+                assert_eq!(range, expected_end, "Case failed: [{}]", sample);
+            }
+        }
+    }
+
     #[test]
     fn test_behaviour_when_moving_to_start_of_previous_long_words() {
         let tests = [
@@ -1444,6 +1765,92 @@ mod test {
         }
     }
 
+    #[test]
+    fn test_behaviour_when_moving_to_end_of_previous_sub_words() {
+        let tests = [
+            (
+                "PrevSubwordEnd",
+                vec![
+                    (1, Range::new(13, 13), Range::new(14, 11)),
+                    (1, Range::new(11, 11), Range::new(11, 4)),
+                ],
+            ),
+            (
+                "prev subword end",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 12)),
+                    (1, Range::new(12, 12), Range::new(12, 4)),
+                ],
+            ),
+            (
+                "Prev Subword End",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 12)),
+                    (1, Range::new(12, 12), Range::new(12, 4)),
+                ],
+            ),
+            (
+                "PREV SUBWORD END",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 12)),
+                    (1, Range::new(12, 12), Range::new(12, 4)),
+                ],
+            ),
+            (
+                "prev_subword_end",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 12)),
+                    (1, Range::new(12, 12), Range::new(12, 4)),
+                ],
+            ),
+            (
+                "Prev_Subword_End",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 12)),
+                    (1, Range::new(12, 12), Range::new(12, 4)),
+                ],
+            ),
+            (
+                "PREV_SUBWORD_END",
+                vec![
+                    (1, Range::new(15, 15), Range::new(16, 12)),
+                    (1, Range::new(12, 12), Range::new(12, 4)),
+                ],
+            ),
+            (
+                "prev__subword__end",
+                vec![
+                    (1, Range::new(17, 17), Range::new(18, 13)),
+                    (1, Range::new(13, 13), Range::new(13, 4)),
+                    (1, Range::new(14, 14), Range::new(15, 13)),
+                ],
+            ),
+            (
+                "Prev__Subword__End",
+                vec![
+                    (1, Range::new(17, 17), Range::new(18, 13)),
+                    (1, Range::new(13, 13), Range::new(13, 4)),
+                    (1, Range::new(14, 14), Range::new(15, 13)),
+                ],
+            ),
+            (
+                "PREV__SUBWORD__END",
+                vec![
+                    (1, Range::new(17, 17), Range::new(18, 13)),
+                    (1, Range::new(13, 13), Range::new(13, 4)),
+                    (1, Range::new(14, 14), Range::new(15, 13)),
+                ],
+            ),
+        ];
+
+        for (sample, scenario) in tests {
+            for (count, begin, expected_end) in scenario.into_iter() {
+                let range = move_prev_sub_word_end(Rope::from(sample).slice(..), begin, count);
+                assert_eq!(range, expected_end, "Case failed: [{}]", sample);
+            }
+        }
+    }
+
     #[test]
     fn test_behaviour_when_moving_to_end_of_next_long_words() {
         let tests = [

helix-core/src/object.rs

@@ -4,8 +4,8 @@ pub fn expand_selection(syntax: &Syntax, text: RopeSlice, selection: Selection)
     let cursor = &mut syntax.walk();
 
     selection.transform(|range| {
-        let from = text.char_to_byte(range.from());
-        let to = text.char_to_byte(range.to());
+        let from = text.char_to_byte(range.from()) as u32;
+        let to = text.char_to_byte(range.to()) as u32;
 
         let byte_range = from..to;
         cursor.reset_to_byte_range(from, to);
@@ -17,8 +17,8 @@ pub fn expand_selection(syntax: &Syntax, text: RopeSlice, selection: Selection)
         }
 
         let node = cursor.node();
-        let from = text.byte_to_char(node.start_byte());
-        let to = text.byte_to_char(node.end_byte());
+        let from = text.byte_to_char(node.start_byte() as usize);
+        let to = text.byte_to_char(node.end_byte() as usize);
 
         Range::new(to, from).with_direction(range.direction())
     })
@@ -53,10 +53,10 @@ pub fn select_next_sibling(syntax: &Syntax, text: RopeSlice, selection: Selectio
 }
 
 pub fn select_all_siblings(syntax: &Syntax, text: RopeSlice, selection: Selection) -> Selection {
-    selection.transform_iter(|range| {
-        let mut cursor = syntax.walk();
+    let mut cursor = syntax.walk();
+    selection.transform_iter(move |range| {
         let (from, to) = range.into_byte_range(text);
-        cursor.reset_to_byte_range(from, to);
+        cursor.reset_to_byte_range(from as u32, to as u32);
 
         if !cursor.goto_parent_with(|parent| parent.child_count() > 1) {
             return vec![range].into_iter();
@@ -67,21 +67,18 @@ pub fn select_all_siblings(syntax: &Syntax, text: RopeSlice, selection: Selectio
 }
 
 pub fn select_all_children(syntax: &Syntax, text: RopeSlice, selection: Selection) -> Selection {
-    selection.transform_iter(|range| {
-        let mut cursor = syntax.walk();
+    let mut cursor = syntax.walk();
+    selection.transform_iter(move |range| {
         let (from, to) = range.into_byte_range(text);
-        cursor.reset_to_byte_range(from, to);
+        cursor.reset_to_byte_range(from as u32, to as u32);
         select_children(&mut cursor, text, range).into_iter()
     })
 }
 
-fn select_children<'n>(
-    cursor: &'n mut TreeCursor<'n>,
-    text: RopeSlice,
-    range: Range,
-) -> Vec<Range> {
+fn select_children(cursor: &mut TreeCursor, text: RopeSlice, range: Range) -> Vec<Range> {
     let children = cursor
-        .named_children()
+        .children()
+        .filter(|child| child.is_named())
         .map(|child| Range::from_node(child, text, range.direction()))
         .collect::<Vec<_>>();
 
@@ -98,7 +95,7 @@ pub fn select_prev_sibling(syntax: &Syntax, text: RopeSlice, selection: Selectio
         text,
         selection,
         |cursor| {
-            while !cursor.goto_prev_sibling() {
+            while !cursor.goto_previous_sibling() {
                 if !cursor.goto_parent() {
                     break;
                 }
@@ -121,16 +118,16 @@ where
     let cursor = &mut syntax.walk();
 
     selection.transform(|range| {
-        let from = text.char_to_byte(range.from());
-        let to = text.char_to_byte(range.to());
+        let from = text.char_to_byte(range.from()) as u32;
+        let to = text.char_to_byte(range.to()) as u32;
 
         cursor.reset_to_byte_range(from, to);
 
         motion(cursor);
 
         let node = cursor.node();
-        let from = text.byte_to_char(node.start_byte());
-        let to = text.byte_to_char(node.end_byte());
+        let from = text.byte_to_char(node.start_byte() as usize);
+        let to = text.byte_to_char(node.end_byte() as usize);
 
         Range::new(from, to).with_direction(direction.unwrap_or_else(|| range.direction()))
     })

helix-core/src/position.rs

@@ -1,9 +1,15 @@
-use std::{borrow::Cow, cmp::Ordering};
+use std::{
+    borrow::Cow,
+    cmp::Ordering,
+    ops::{Add, AddAssign, Sub, SubAssign},
+};
+
+use helix_stdx::rope::RopeSliceExt;
 
 use crate::{
     chars::char_is_line_ending,
     doc_formatter::{DocumentFormatter, TextFormat},
-    graphemes::{ensure_grapheme_boundary_prev, grapheme_width, RopeGraphemes},
+    graphemes::{ensure_grapheme_boundary_prev, grapheme_width},
     line_ending::line_end_char_index,
     text_annotations::TextAnnotations,
     RopeSlice,
@@ -16,6 +22,38 @@ pub struct Position {
     pub col: usize,
 }
 
+impl AddAssign for Position {
+    fn add_assign(&mut self, rhs: Self) {
+        self.row += rhs.row;
+        self.col += rhs.col;
+    }
+}
+
+impl SubAssign for Position {
+    fn sub_assign(&mut self, rhs: Self) {
+        self.row -= rhs.row;
+        self.col -= rhs.col;
+    }
+}
+
+impl Sub for Position {
+    type Output = Position;
+
+    fn sub(mut self, rhs: Self) -> Self::Output {
+        self -= rhs;
+        self
+    }
+}
+
+impl Add for Position {
+    type Output = Position;
+
+    fn add(mut self, rhs: Self) -> Self::Output {
+        self += rhs;
+        self
+    }
+}
+
 impl Position {
     pub const fn new(row: usize, col: usize) -> Self {
         Self { row, col }
@@ -51,11 +89,6 @@ impl From<(usize, usize)> for Position {
     }
 }
 
-impl From<Position> for tree_sitter::Point {
-    fn from(pos: Position) -> Self {
-        Self::new(pos.row, pos.col)
-    }
-}
 /// Convert a character index to (line, column) coordinates.
 ///
 /// column in `char` count which can be used for row:column display in
@@ -65,7 +98,7 @@ pub fn coords_at_pos(text: RopeSlice, pos: usize) -> Position {
 
     let line_start = text.line_to_char(line);
     let pos = ensure_grapheme_boundary_prev(text, pos);
-    let col = RopeGraphemes::new(text.slice(line_start..pos)).count();
+    let col = text.slice(line_start..pos).graphemes().count();
 
     Position::new(line, col)
 }
@@ -90,7 +123,7 @@ pub fn visual_coords_at_pos(text: RopeSlice, pos: usize, tab_width: usize) -> Po
 
     let mut col = 0;
 
-    for grapheme in RopeGraphemes::new(text.slice(line_start..pos)) {
+    for grapheme in text.slice(line_start..pos).graphemes() {
         if grapheme == "\t" {
             col += tab_width - (col % tab_width);
         } else {
@@ -121,22 +154,31 @@ pub fn visual_offset_from_block(
     annotations: &TextAnnotations,
 ) -> (Position, usize) {
     let mut last_pos = Position::default();
-    let (formatter, block_start) =
+    let mut formatter =
         DocumentFormatter::new_at_prev_checkpoint(text, text_fmt, annotations, anchor);
-    let mut char_pos = block_start;
+    let block_start = formatter.next_char_pos();
 
-    for (grapheme, vpos) in formatter {
-        last_pos = vpos;
-        char_pos += grapheme.doc_chars();
-
-        if char_pos > pos {
-            return (last_pos, block_start);
+    while let Some(grapheme) = formatter.next() {
+        last_pos = grapheme.visual_pos;
+        if formatter.next_char_pos() > pos {
+            return (grapheme.visual_pos, block_start);
         }
     }
 
     (last_pos, block_start)
 }
 
+/// Returns the height of the given text when softwrapping
+pub fn softwrapped_dimensions(text: RopeSlice, text_fmt: &TextFormat) -> (usize, u16) {
+    let last_pos =
+        visual_offset_from_block(text, 0, usize::MAX, text_fmt, &TextAnnotations::default()).0;
+    if last_pos.row == 0 {
+        (1, last_pos.col as u16)
+    } else {
+        (last_pos.row + 1, text_fmt.viewport_width)
+    }
+}
+
 #[derive(Debug, PartialEq, Eq, Clone, Copy)]
 pub enum VisualOffsetError {
     PosBeforeAnchorRow,
@@ -153,22 +195,21 @@ pub fn visual_offset_from_anchor(
     annotations: &TextAnnotations,
     max_rows: usize,
 ) -> Result<(Position, usize), VisualOffsetError> {
-    let (formatter, block_start) =
+    let mut formatter =
         DocumentFormatter::new_at_prev_checkpoint(text, text_fmt, annotations, anchor);
-    let mut char_pos = block_start;
     let mut anchor_line = None;
     let mut found_pos = None;
     let mut last_pos = Position::default();
 
+    let block_start = formatter.next_char_pos();
     if pos < block_start {
         return Err(VisualOffsetError::PosBeforeAnchorRow);
     }
 
-    for (grapheme, vpos) in formatter {
-        last_pos = vpos;
-        char_pos += grapheme.doc_chars();
+    while let Some(grapheme) = formatter.next() {
+        last_pos = grapheme.visual_pos;
 
-        if char_pos > pos {
+        if formatter.next_char_pos() > pos {
             if let Some(anchor_line) = anchor_line {
                 last_pos.row -= anchor_line;
                 return Ok((last_pos, block_start));
@@ -176,7 +217,7 @@ pub fn visual_offset_from_anchor(
                 found_pos = Some(last_pos);
             }
         }
-        if char_pos > anchor && anchor_line.is_none() {
+        if formatter.next_char_pos() > anchor && anchor_line.is_none() {
             if let Some(mut found_pos) = found_pos {
                 return if found_pos.row == last_pos.row {
                     found_pos.row = 0;
@@ -190,7 +231,7 @@ pub fn visual_offset_from_anchor(
         }
 
         if let Some(anchor_line) = anchor_line {
-            if vpos.row >= anchor_line + max_rows {
+            if grapheme.visual_pos.row >= anchor_line + max_rows {
                 return Err(VisualOffsetError::PosAfterMaxRow);
             }
         }
@@ -231,7 +272,7 @@ pub fn pos_at_coords(text: RopeSlice, coords: Position, limit_before_line_ending
     };
 
     let mut col_char_offset = 0;
-    for (i, g) in RopeGraphemes::new(text.slice(line_start..line_end)).enumerate() {
+    for (i, g) in text.slice(line_start..line_end).graphemes().enumerate() {
         if i == col {
             break;
         }
@@ -262,7 +303,7 @@ pub fn pos_at_visual_coords(text: RopeSlice, coords: Position, tab_width: usize)
 
     let mut col_char_offset = 0;
     let mut cols_remaining = col;
-    for grapheme in RopeGraphemes::new(text.slice(line_start..line_end)) {
+    for grapheme in text.slice(line_start..line_end).graphemes() {
         let grapheme_width = if grapheme == "\t" {
             tab_width - ((col - cols_remaining) % tab_width)
         } else {
@@ -368,39 +409,43 @@ pub fn char_idx_at_visual_block_offset(
     text_fmt: &TextFormat,
     annotations: &TextAnnotations,
 ) -> (usize, usize) {
-    let (formatter, mut char_idx) =
+    let mut formatter =
         DocumentFormatter::new_at_prev_checkpoint(text, text_fmt, annotations, anchor);
-    let mut last_char_idx = char_idx;
-    let mut last_char_idx_on_line = None;
+    let mut last_char_idx = formatter.next_char_pos();
+    let mut found_non_virtual_on_row = false;
     let mut last_row = 0;
-    for (grapheme, grapheme_pos) in formatter {
-        match grapheme_pos.row.cmp(&row) {
+    for grapheme in &mut formatter {
+        match grapheme.visual_pos.row.cmp(&row) {
             Ordering::Equal => {
-                if grapheme_pos.col + grapheme.width() > column {
+                if grapheme.visual_pos.col + grapheme.width() > column {
                     if !grapheme.is_virtual() {
-                        return (char_idx, 0);
-                    } else if let Some(char_idx) = last_char_idx_on_line {
-                        return (char_idx, 0);
+                        return (grapheme.char_idx, 0);
+                    } else if found_non_virtual_on_row {
+                        return (last_char_idx, 0);
                     }
                 } else if !grapheme.is_virtual() {
-                    last_char_idx_on_line = Some(char_idx)
+                    found_non_virtual_on_row = true;
+                    last_char_idx = grapheme.char_idx;
                 }
             }
+            Ordering::Greater if found_non_virtual_on_row => return (last_char_idx, 0),
             Ordering::Greater => return (last_char_idx, row - last_row),
-            _ => (),
+            Ordering::Less => {
+                if !grapheme.is_virtual() {
+                    last_row = grapheme.visual_pos.row;
+                    last_char_idx = grapheme.char_idx;
+                }
+            }
         }
-
-        last_char_idx = char_idx;
-        last_row = grapheme_pos.row;
-        char_idx += grapheme.doc_chars();
     }
 
-    (char_idx, 0)
+    (formatter.next_char_pos(), 0)
 }
 
 #[cfg(test)]
 mod test {
     use super::*;
+    use crate::text_annotations::InlineAnnotation;
     use crate::Rope;
 
     #[test]
@@ -761,6 +806,30 @@ mod test {
         assert_eq!(pos_at_visual_coords(slice, (10, 10).into(), 4), 0);
     }
 
+    #[test]
+    fn test_char_idx_at_visual_row_offset_inline_annotation() {
+        let text = Rope::from("foo\nbar");
+        let slice = text.slice(..);
+        let mut text_fmt = TextFormat::default();
+        let annotations = [InlineAnnotation {
+            text: "x".repeat(100).into(),
+            char_idx: 3,
+        }];
+        text_fmt.soft_wrap = true;
+
+        assert_eq!(
+            char_idx_at_visual_offset(
+                slice,
+                0,
+                1,
+                0,
+                &text_fmt,
+                TextAnnotations::default().add_inline_annotations(&annotations, None)
+            ),
+            (2, 1)
+        );
+    }
+
     #[test]
     fn test_char_idx_at_visual_row_offset() {
         let text = Rope::from("ḧëḷḷö\nẅöṛḷḋ\nfoo");

helix-core/src/selection.rs

@@ -9,12 +9,13 @@ use crate::{
     },
     line_ending::get_line_ending,
     movement::Direction,
-    Assoc, ChangeSet, RopeGraphemes, RopeSlice,
+    tree_sitter::Node,
+    Assoc, ChangeSet, RopeSlice,
 };
+use helix_stdx::range::is_subset;
 use helix_stdx::rope::{self, RopeSliceExt};
 use smallvec::{smallvec, SmallVec};
 use std::{borrow::Cow, iter, slice};
-use tree_sitter::Node;
 
 /// A single selection range.
 ///
@@ -75,8 +76,8 @@ impl Range {
     }
 
     pub fn from_node(node: Node, text: RopeSlice, direction: Direction) -> Self {
-        let from = text.byte_to_char(node.start_byte());
-        let to = text.byte_to_char(node.end_byte());
+        let from = text.byte_to_char(node.start_byte() as usize);
+        let to = text.byte_to_char(node.end_byte() as usize);
         Range::new(from, to).with_direction(direction)
     }
 
@@ -184,16 +185,16 @@ impl Range {
 
         let positions_to_map = match self.anchor.cmp(&self.head) {
             Ordering::Equal => [
-                (&mut self.anchor, Assoc::After),
-                (&mut self.head, Assoc::After),
+                (&mut self.anchor, Assoc::AfterSticky),
+                (&mut self.head, Assoc::AfterSticky),
             ],
             Ordering::Less => [
-                (&mut self.anchor, Assoc::After),
-                (&mut self.head, Assoc::Before),
+                (&mut self.anchor, Assoc::AfterSticky),
+                (&mut self.head, Assoc::BeforeSticky),
             ],
             Ordering::Greater => [
-                (&mut self.head, Assoc::After),
-                (&mut self.anchor, Assoc::Before),
+                (&mut self.head, Assoc::AfterSticky),
+                (&mut self.anchor, Assoc::BeforeSticky),
             ],
         };
         changes.update_positions(positions_to_map.into_iter());
@@ -378,7 +379,7 @@ impl Range {
 
     /// Returns true if this Range covers a single grapheme in the given text
     pub fn is_single_grapheme(&self, doc: RopeSlice) -> bool {
-        let mut graphemes = RopeGraphemes::new(doc.slice(self.from()..self.to()));
+        let mut graphemes = doc.slice(self.from()..self.to()).graphemes();
         let first = graphemes.next();
         let second = graphemes.next();
         first.is_some() && second.is_none()
@@ -401,6 +402,15 @@ impl From<(usize, usize)> for Range {
     }
 }
 
+impl From<Range> for helix_stdx::Range {
+    fn from(range: Range) -> Self {
+        Self {
+            start: range.from(),
+            end: range.to(),
+        }
+    }
+}
+
 /// A selection consists of one or more selection ranges.
 /// invariant: A selection can never be empty (always contains at least primary range).
 #[derive(Debug, Clone, PartialEq, Eq)]
@@ -482,16 +492,16 @@ impl Selection {
             range.old_visual_position = None;
             match range.anchor.cmp(&range.head) {
                 Ordering::Equal => [
-                    (&mut range.anchor, Assoc::After),
-                    (&mut range.head, Assoc::After),
+                    (&mut range.anchor, Assoc::AfterSticky),
+                    (&mut range.head, Assoc::AfterSticky),
                 ],
                 Ordering::Less => [
-                    (&mut range.anchor, Assoc::After),
-                    (&mut range.head, Assoc::Before),
+                    (&mut range.anchor, Assoc::AfterSticky),
+                    (&mut range.head, Assoc::BeforeSticky),
                 ],
                 Ordering::Greater => [
-                    (&mut range.head, Assoc::After),
-                    (&mut range.anchor, Assoc::Before),
+                    (&mut range.head, Assoc::AfterSticky),
+                    (&mut range.anchor, Assoc::BeforeSticky),
                 ],
             }
         });
@@ -513,6 +523,10 @@ impl Selection {
         }
     }
 
+    pub fn range_bounds(&self) -> impl Iterator<Item = helix_stdx::Range> + '_ {
+        self.ranges.iter().map(|&range| range.into())
+    }
+
     pub fn primary_index(&self) -> usize {
         self.primary_index
     }
@@ -605,7 +619,6 @@ impl Selection {
         self
     }
 
-    // TODO: consume an iterator or a vec to reduce allocations?
     #[must_use]
     pub fn new(ranges: SmallVec<[Range; 1]>, primary_index: usize) -> Self {
         assert!(!ranges.is_empty());
@@ -660,7 +673,7 @@ impl Selection {
     pub fn fragments<'a>(
         &'a self,
         text: RopeSlice<'a>,
-    ) -> impl DoubleEndedIterator<Item = Cow<'a, str>> + ExactSizeIterator<Item = Cow<str>> + 'a
+    ) -> impl DoubleEndedIterator<Item = Cow<'a, str>> + ExactSizeIterator<Item = Cow<'a, str>>
     {
         self.ranges.iter().map(move |range| range.fragment(text))
     }
@@ -683,32 +696,9 @@ impl Selection {
         self.ranges.len()
     }
 
-    // returns true if self ⊇ other
+    /// returns true if self ⊇ other
     pub fn contains(&self, other: &Selection) -> bool {
-        let (mut iter_self, mut iter_other) = (self.iter(), other.iter());
-        let (mut ele_self, mut ele_other) = (iter_self.next(), iter_other.next());
-
-        loop {
-            match (ele_self, ele_other) {
-                (Some(ra), Some(rb)) => {
-                    if !ra.contains_range(rb) {
-                        // `self` doesn't contain next element from `other`, advance `self`, we need to match all from `other`
-                        ele_self = iter_self.next();
-                    } else {
-                        // matched element from `other`, advance `other`
-                        ele_other = iter_other.next();
-                    };
-                }
-                (None, Some(_)) => {
-                    // exhausted `self`, we can't match the reminder of `other`
-                    return false;
-                }
-                (_, None) => {
-                    // no elements from `other` left to match, `self` contains `other`
-                    return true;
-                }
-            }
-        }
+        is_subset::<true>(self.range_bounds(), other.range_bounds())
     }
 }
 
@@ -730,6 +720,12 @@ impl IntoIterator for Selection {
     }
 }
 
+impl FromIterator<Range> for Selection {
+    fn from_iter<T: IntoIterator<Item = Range>>(ranges: T) -> Self {
+        Self::new(ranges.into_iter().collect(), 0)
+    }
+}
+
 impl From<Range> for Selection {
     fn from(range: Range) -> Self {
         Self {
@@ -744,7 +740,7 @@ pub struct LineRangeIter<'a> {
     text: RopeSlice<'a>,
 }
 
-impl<'a> Iterator for LineRangeIter<'a> {
+impl Iterator for LineRangeIter<'_> {
     type Item = (usize, usize);
 
     fn next(&mut self) -> Option<Self::Item> {

helix-core/src/shellwords.rs

@@ -1,350 +0,0 @@
-use std::borrow::Cow;
-
-/// Auto escape for shellwords usage.
-pub fn escape(input: Cow<str>) -> Cow<str> {
-    if !input.chars().any(|x| x.is_ascii_whitespace()) {
-        input
-    } else if cfg!(unix) {
-        Cow::Owned(input.chars().fold(String::new(), |mut buf, c| {
-            if c.is_ascii_whitespace() {
-                buf.push('\\');
-            }
-            buf.push(c);
-            buf
-        }))
-    } else {
-        Cow::Owned(format!("\"{}\"", input))
-    }
-}
-
-enum State {
-    OnWhitespace,
-    Unquoted,
-    UnquotedEscaped,
-    Quoted,
-    QuoteEscaped,
-    Dquoted,
-    DquoteEscaped,
-}
-
-pub struct Shellwords<'a> {
-    state: State,
-    /// Shellwords where whitespace and escapes has been resolved.
-    words: Vec<Cow<'a, str>>,
-    /// The parts of the input that are divided into shellwords. This can be
-    /// used to retrieve the original text for a given word by looking up the
-    /// same index in the Vec as the word in `words`.
-    parts: Vec<&'a str>,
-}
-
-impl<'a> From<&'a str> for Shellwords<'a> {
-    fn from(input: &'a str) -> Self {
-        use State::*;
-
-        let mut state = Unquoted;
-        let mut words = Vec::new();
-        let mut parts = Vec::new();
-        let mut escaped = String::with_capacity(input.len());
-
-        let mut part_start = 0;
-        let mut unescaped_start = 0;
-        let mut end = 0;
-
-        for (i, c) in input.char_indices() {
-            state = match state {
-                OnWhitespace => match c {
-                    '"' => {
-                        end = i;
-                        Dquoted
-                    }
-                    '\'' => {
-                        end = i;
-                        Quoted
-                    }
-                    '\\' => {
-                        if cfg!(unix) {
-                            escaped.push_str(&input[unescaped_start..i]);
-                            unescaped_start = i + 1;
-                            UnquotedEscaped
-                        } else {
-                            OnWhitespace
-                        }
-                    }
-                    c if c.is_ascii_whitespace() => {
-                        end = i;
-                        OnWhitespace
-                    }
-                    _ => Unquoted,
-                },
-                Unquoted => match c {
-                    '\\' => {
-                        if cfg!(unix) {
-                            escaped.push_str(&input[unescaped_start..i]);
-                            unescaped_start = i + 1;
-                            UnquotedEscaped
-                        } else {
-                            Unquoted
-                        }
-                    }
-                    c if c.is_ascii_whitespace() => {
-                        end = i;
-                        OnWhitespace
-                    }
-                    _ => Unquoted,
-                },
-                UnquotedEscaped => Unquoted,
-                Quoted => match c {
-                    '\\' => {
-                        if cfg!(unix) {
-                            escaped.push_str(&input[unescaped_start..i]);
-                            unescaped_start = i + 1;
-                            QuoteEscaped
-                        } else {
-                            Quoted
-                        }
-                    }
-                    '\'' => {
-                        end = i;
-                        OnWhitespace
-                    }
-                    _ => Quoted,
-                },
-                QuoteEscaped => Quoted,
-                Dquoted => match c {
-                    '\\' => {
-                        if cfg!(unix) {
-                            escaped.push_str(&input[unescaped_start..i]);
-                            unescaped_start = i + 1;
-                            DquoteEscaped
-                        } else {
-                            Dquoted
-                        }
-                    }
-                    '"' => {
-                        end = i;
-                        OnWhitespace
-                    }
-                    _ => Dquoted,
-                },
-                DquoteEscaped => Dquoted,
-            };
-
-            let c_len = c.len_utf8();
-            if i == input.len() - c_len && end == 0 {
-                end = i + c_len;
-            }
-
-            if end > 0 {
-                let esc_trim = escaped.trim();
-                let inp = &input[unescaped_start..end];
-
-                if !(esc_trim.is_empty() && inp.trim().is_empty()) {
-                    if esc_trim.is_empty() {
-                        words.push(inp.into());
-                        parts.push(inp);
-                    } else {
-                        words.push([escaped, inp.into()].concat().into());
-                        parts.push(&input[part_start..end]);
-                        escaped = "".to_string();
-                    }
-                }
-                unescaped_start = i + 1;
-                part_start = i + 1;
-                end = 0;
-            }
-        }
-
-        debug_assert!(words.len() == parts.len());
-
-        Self {
-            state,
-            words,
-            parts,
-        }
-    }
-}
-
-impl<'a> Shellwords<'a> {
-    /// Checks that the input ends with a whitespace character which is not escaped.
-    ///
-    /// # Examples
-    ///
-    /// ```rust
-    /// use helix_core::shellwords::Shellwords;
-    /// assert_eq!(Shellwords::from(" ").ends_with_whitespace(), true);
-    /// assert_eq!(Shellwords::from(":open ").ends_with_whitespace(), true);
-    /// assert_eq!(Shellwords::from(":open foo.txt ").ends_with_whitespace(), true);
-    /// assert_eq!(Shellwords::from(":open").ends_with_whitespace(), false);
-    /// #[cfg(unix)]
-    /// assert_eq!(Shellwords::from(":open a\\ ").ends_with_whitespace(), false);
-    /// #[cfg(unix)]
-    /// assert_eq!(Shellwords::from(":open a\\ b.txt").ends_with_whitespace(), false);
-    /// ```
-    pub fn ends_with_whitespace(&self) -> bool {
-        matches!(self.state, State::OnWhitespace)
-    }
-
-    /// Returns the list of shellwords calculated from the input string.
-    pub fn words(&self) -> &[Cow<'a, str>] {
-        &self.words
-    }
-
-    /// Returns a list of strings which correspond to [`Self::words`] but represent the original
-    /// text in the input string - including escape characters - without separating whitespace.
-    pub fn parts(&self) -> &[&'a str] {
-        &self.parts
-    }
-}
-
-#[cfg(test)]
-mod test {
-    use super::*;
-
-    #[test]
-    #[cfg(windows)]
-    fn test_normal() {
-        let input = r#":o single_word twó wörds \three\ \"with\ escaping\\"#;
-        let shellwords = Shellwords::from(input);
-        let result = shellwords.words().to_vec();
-        let expected = vec![
-            Cow::from(":o"),
-            Cow::from("single_word"),
-            Cow::from("twó"),
-            Cow::from("wörds"),
-            Cow::from("\\three\\"),
-            Cow::from("\\"),
-            Cow::from("with\\ escaping\\\\"),
-        ];
-        // TODO test is_owned and is_borrowed, once they get stabilized.
-        assert_eq!(expected, result);
-    }
-
-    #[test]
-    #[cfg(unix)]
-    fn test_normal() {
-        let input = r#":o single_word twó wörds \three\ \"with\ escaping\\"#;
-        let shellwords = Shellwords::from(input);
-        let result = shellwords.words().to_vec();
-        let expected = vec![
-            Cow::from(":o"),
-            Cow::from("single_word"),
-            Cow::from("twó"),
-            Cow::from("wörds"),
-            Cow::from(r#"three "with escaping\"#),
-        ];
-        // TODO test is_owned and is_borrowed, once they get stabilized.
-        assert_eq!(expected, result);
-    }
-
-    #[test]
-    #[cfg(unix)]
-    fn test_quoted() {
-        let quoted =
-            r#":o 'single_word' 'twó wörds' '' ' ''\three\' \"with\ escaping\\' 'quote incomplete"#;
-        let shellwords = Shellwords::from(quoted);
-        let result = shellwords.words().to_vec();
-        let expected = vec![
-            Cow::from(":o"),
-            Cow::from("single_word"),
-            Cow::from("twó wörds"),
-            Cow::from(r#"three' "with escaping\"#),
-            Cow::from("quote incomplete"),
-        ];
-        assert_eq!(expected, result);
-    }
-
-    #[test]
-    #[cfg(unix)]
-    fn test_dquoted() {
-        let dquoted = r#":o "single_word" "twó wörds" "" "  ""\three\' \"with\ escaping\\" "dquote incomplete"#;
-        let shellwords = Shellwords::from(dquoted);
-        let result = shellwords.words().to_vec();
-        let expected = vec![
-            Cow::from(":o"),
-            Cow::from("single_word"),
-            Cow::from("twó wörds"),
-            Cow::from(r#"three' "with escaping\"#),
-            Cow::from("dquote incomplete"),
-        ];
-        assert_eq!(expected, result);
-    }
-
-    #[test]
-    #[cfg(unix)]
-    fn test_mixed() {
-        let dquoted = r#":o single_word 'twó wörds' "\three\' \"with\ escaping\\""no space before"'and after' $#%^@ "%^&(%^" ')(*&^%''a\\\\\b' '"#;
-        let shellwords = Shellwords::from(dquoted);
-        let result = shellwords.words().to_vec();
-        let expected = vec![
-            Cow::from(":o"),
-            Cow::from("single_word"),
-            Cow::from("twó wörds"),
-            Cow::from("three' \"with escaping\\"),
-            Cow::from("no space before"),
-            Cow::from("and after"),
-            Cow::from("$#%^@"),
-            Cow::from("%^&(%^"),
-            Cow::from(")(*&^%"),
-            Cow::from(r#"a\\b"#),
-            //last ' just changes to quoted but since we dont have anything after it, it should be ignored
-        ];
-        assert_eq!(expected, result);
-    }
-
-    #[test]
-    fn test_lists() {
-        let input =
-            r#":set statusline.center ["file-type","file-encoding"] '["list", "in", "quotes"]'"#;
-        let shellwords = Shellwords::from(input);
-        let result = shellwords.words().to_vec();
-        let expected = vec![
-            Cow::from(":set"),
-            Cow::from("statusline.center"),
-            Cow::from(r#"["file-type","file-encoding"]"#),
-            Cow::from(r#"["list", "in", "quotes"]"#),
-        ];
-        assert_eq!(expected, result);
-    }
-
-    #[test]
-    #[cfg(unix)]
-    fn test_escaping_unix() {
-        assert_eq!(escape("foobar".into()), Cow::Borrowed("foobar"));
-        assert_eq!(escape("foo bar".into()), Cow::Borrowed("foo\\ bar"));
-        assert_eq!(escape("foo\tbar".into()), Cow::Borrowed("foo\\\tbar"));
-    }
-
-    #[test]
-    #[cfg(windows)]
-    fn test_escaping_windows() {
-        assert_eq!(escape("foobar".into()), Cow::Borrowed("foobar"));
-        assert_eq!(escape("foo bar".into()), Cow::Borrowed("\"foo bar\""));
-    }
-
-    #[test]
-    #[cfg(unix)]
-    fn test_parts() {
-        assert_eq!(Shellwords::from(":o a").parts(), &[":o", "a"]);
-        assert_eq!(Shellwords::from(":o a\\ ").parts(), &[":o", "a\\ "]);
-    }
-
-    #[test]
-    #[cfg(windows)]
-    fn test_parts() {
-        assert_eq!(Shellwords::from(":o a").parts(), &[":o", "a"]);
-        assert_eq!(Shellwords::from(":o a\\ ").parts(), &[":o", "a\\"]);
-    }
-
-    #[test]
-    fn test_multibyte_at_end() {
-        assert_eq!(Shellwords::from("𒀀").parts(), &["𒀀"]);
-        assert_eq!(
-            Shellwords::from(":sh echo 𒀀").parts(),
-            &[":sh", "echo", "𒀀"]
-        );
-        assert_eq!(
-            Shellwords::from(":sh echo 𒀀 hello world𒀀").parts(),
-            &[":sh", "echo", "𒀀", "hello", "world𒀀"]
-        );
-    }
-}

helix-core/src/snippets.rs

@@ -0,0 +1,13 @@
+mod active;
+mod elaborate;
+mod parser;
+mod render;
+
+#[derive(PartialEq, Eq, Hash, Debug, PartialOrd, Ord, Clone, Copy)]
+pub struct TabstopIdx(usize);
+pub const LAST_TABSTOP_IDX: TabstopIdx = TabstopIdx(usize::MAX);
+
+pub use active::ActiveSnippet;
+pub use elaborate::{Snippet, SnippetElement, Transform};
+pub use render::RenderedSnippet;
+pub use render::SnippetRenderCtx;

helix-core/src/snippets/active.rs

@@ -0,0 +1,272 @@
+use std::ops::{Index, IndexMut};
+
+use foldhash::HashSet;
+use helix_stdx::range::{is_exact_subset, is_subset};
+use helix_stdx::Range;
+use ropey::Rope;
+
+use crate::movement::Direction;
+use crate::snippets::render::{RenderedSnippet, Tabstop};
+use crate::snippets::TabstopIdx;
+use crate::{Assoc, ChangeSet, Selection, Transaction};
+
+pub struct ActiveSnippet {
+    ranges: Vec<Range>,
+    active_tabstops: HashSet<TabstopIdx>,
+    current_tabstop: TabstopIdx,
+    tabstops: Vec<Tabstop>,
+}
+
+impl Index<TabstopIdx> for ActiveSnippet {
+    type Output = Tabstop;
+    fn index(&self, index: TabstopIdx) -> &Tabstop {
+        &self.tabstops[index.0]
+    }
+}
+
+impl IndexMut<TabstopIdx> for ActiveSnippet {
+    fn index_mut(&mut self, index: TabstopIdx) -> &mut Tabstop {
+        &mut self.tabstops[index.0]
+    }
+}
+
+impl ActiveSnippet {
+    pub fn new(snippet: RenderedSnippet) -> Option<Self> {
+        let snippet = Self {
+            ranges: snippet.ranges,
+            tabstops: snippet.tabstops,
+            active_tabstops: HashSet::default(),
+            current_tabstop: TabstopIdx(0),
+        };
+        (snippet.tabstops.len() != 1).then_some(snippet)
+    }
+
+    pub fn is_valid(&self, new_selection: &Selection) -> bool {
+        is_subset::<false>(self.ranges.iter().copied(), new_selection.range_bounds())
+    }
+
+    pub fn tabstops(&self) -> impl Iterator<Item = &Tabstop> {
+        self.tabstops.iter()
+    }
+
+    pub fn delete_placeholder(&self, doc: &Rope) -> Transaction {
+        Transaction::delete(
+            doc,
+            self[self.current_tabstop]
+                .ranges
+                .iter()
+                .map(|range| (range.start, range.end)),
+        )
+    }
+
+    /// maps the active snippets through a `ChangeSet` updating all tabstop ranges
+    pub fn map(&mut self, changes: &ChangeSet) -> bool {
+        let positions_to_map = self.ranges.iter_mut().flat_map(|range| {
+            [
+                (&mut range.start, Assoc::After),
+                (&mut range.end, Assoc::Before),
+            ]
+        });
+        changes.update_positions(positions_to_map);
+
+        for (i, tabstop) in self.tabstops.iter_mut().enumerate() {
+            if self.active_tabstops.contains(&TabstopIdx(i)) {
+                let positions_to_map = tabstop.ranges.iter_mut().flat_map(|range| {
+                    let end_assoc = if range.start == range.end {
+                        Assoc::Before
+                    } else {
+                        Assoc::After
+                    };
+                    [
+                        (&mut range.start, Assoc::Before),
+                        (&mut range.end, end_assoc),
+                    ]
+                });
+                changes.update_positions(positions_to_map);
+            } else {
+                let positions_to_map = tabstop.ranges.iter_mut().flat_map(|range| {
+                    let end_assoc = if range.start == range.end {
+                        Assoc::After
+                    } else {
+                        Assoc::Before
+                    };
+                    [
+                        (&mut range.start, Assoc::After),
+                        (&mut range.end, end_assoc),
+                    ]
+                });
+                changes.update_positions(positions_to_map);
+            }
+            let mut snippet_ranges = self.ranges.iter();
+            let mut snippet_range = snippet_ranges.next().unwrap();
+            let mut tabstop_i = 0;
+            let mut prev = Range { start: 0, end: 0 };
+            let num_ranges = tabstop.ranges.len() / self.ranges.len();
+            tabstop.ranges.retain_mut(|range| {
+                if tabstop_i == num_ranges {
+                    snippet_range = snippet_ranges.next().unwrap();
+                    tabstop_i = 0;
+                }
+                tabstop_i += 1;
+                let retain = snippet_range.start <= snippet_range.end;
+                if retain {
+                    range.start = range.start.max(snippet_range.start);
+                    range.end = range.end.max(range.start).min(snippet_range.end);
+                    // guaranteed by assoc
+                    debug_assert!(prev.start <= range.start);
+                    debug_assert!(range.start <= range.end);
+                    if prev.end > range.start {
+                        // not really sure what to do in this case. It shouldn't
+                        // really occur in practice, the below just ensures
+                        // our invariants hold
+                        range.start = prev.end;
+                        range.end = range.end.max(range.start)
+                    }
+                    prev = *range;
+                }
+                retain
+            });
+        }
+        self.ranges.iter().all(|range| range.end <= range.start)
+    }
+
+    pub fn next_tabstop(&mut self, current_selection: &Selection) -> (Selection, bool) {
+        let primary_idx = self.primary_idx(current_selection);
+        while self.current_tabstop.0 + 1 < self.tabstops.len() {
+            self.current_tabstop.0 += 1;
+            if self.activate_tabstop() {
+                let selection = self.tabstop_selection(primary_idx, Direction::Forward);
+                return (selection, self.current_tabstop.0 + 1 == self.tabstops.len());
+            }
+        }
+
+        (
+            self.tabstop_selection(primary_idx, Direction::Forward),
+            true,
+        )
+    }
+
+    pub fn prev_tabstop(&mut self, current_selection: &Selection) -> Option<Selection> {
+        let primary_idx = self.primary_idx(current_selection);
+        while self.current_tabstop.0 != 0 {
+            self.current_tabstop.0 -= 1;
+            if self.activate_tabstop() {
+                return Some(self.tabstop_selection(primary_idx, Direction::Forward));
+            }
+        }
+        None
+    }
+    // computes the primary idx adjusted for the number of cursors in the current tabstop
+    fn primary_idx(&self, current_selection: &Selection) -> usize {
+        let primary: Range = current_selection.primary().into();
+        let res = self
+            .ranges
+            .iter()
+            .position(|&range| range.contains(primary));
+        res.unwrap_or_else(|| {
+            unreachable!(
+                "active snippet must be valid {current_selection:?} {:?}",
+                self.ranges
+            )
+        })
+    }
+
+    fn activate_tabstop(&mut self) -> bool {
+        let tabstop = &self[self.current_tabstop];
+        if tabstop.has_placeholder() && tabstop.ranges.iter().all(|range| range.is_empty()) {
+            return false;
+        }
+        self.active_tabstops.clear();
+        self.active_tabstops.insert(self.current_tabstop);
+        let mut parent = self[self.current_tabstop].parent;
+        while let Some(tabstop) = parent {
+            self.active_tabstops.insert(tabstop);
+            parent = self[tabstop].parent;
+        }
+        true
+        // TODO: if the user removes the selection(s) in one snippet (but
+        // there are still other cursors in other snippets) and jumps to the
+        // next tabstop the selection in that tabstop is restored (at the
+        // next tabstop). This could be annoying since its not possible to
+        // remove a snippet cursor until the snippet is complete. On the other
+        // hand it may be useful since the user may just have meant to edit
+        // a subselection (like with s) of the tabstops and so the selection
+        // removal was just temporary. Potentially this could have some sort of
+        // separate keymap
+    }
+
+    pub fn tabstop_selection(&self, primary_idx: usize, direction: Direction) -> Selection {
+        let tabstop = &self[self.current_tabstop];
+        tabstop.selection(direction, primary_idx, self.ranges.len())
+    }
+
+    pub fn insert_subsnippet(mut self, snippet: RenderedSnippet) -> Option<Self> {
+        if snippet.ranges.len() % self.ranges.len() != 0
+            || !is_exact_subset(self.ranges.iter().copied(), snippet.ranges.iter().copied())
+        {
+            log::warn!("number of subsnippets did not match, discarding outer snippet");
+            return ActiveSnippet::new(snippet);
+        }
+        let mut cnt = 0;
+        let parent = self[self.current_tabstop].parent;
+        let tabstops = snippet.tabstops.into_iter().map(|mut tabstop| {
+            cnt += 1;
+            if let Some(parent) = &mut tabstop.parent {
+                parent.0 += self.current_tabstop.0;
+            } else {
+                tabstop.parent = parent;
+            }
+            tabstop
+        });
+        self.tabstops
+            .splice(self.current_tabstop.0..=self.current_tabstop.0, tabstops);
+        self.activate_tabstop();
+        Some(self)
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use std::iter::{self};
+
+    use ropey::Rope;
+
+    use crate::snippets::{ActiveSnippet, Snippet, SnippetRenderCtx};
+    use crate::{Selection, Transaction};
+
+    #[test]
+    fn fully_remove() {
+        let snippet = Snippet::parse("foo(${1:bar})$0").unwrap();
+        let mut doc = Rope::from("bar.\n");
+        let (transaction, _, snippet) = snippet.render(
+            &doc,
+            &Selection::point(4),
+            |_| (4, 4),
+            &mut SnippetRenderCtx::test_ctx(),
+        );
+        assert!(transaction.apply(&mut doc));
+        assert_eq!(doc, "bar.foo(bar)\n");
+        let mut snippet = ActiveSnippet::new(snippet).unwrap();
+        let edit = Transaction::change(&doc, iter::once((4, 12, None)));
+        assert!(edit.apply(&mut doc));
+        snippet.map(edit.changes());
+        assert!(!snippet.is_valid(&Selection::point(4)))
+    }
+
+    #[test]
+    fn tabstop_zero_with_placeholder() {
+        // The `$0` tabstop should not have placeholder text. When we receive a snippet like this
+        // (from older versions of clangd for example) we should discard the placeholder text.
+        let snippet = Snippet::parse("sizeof(${0:expression-or-type})").unwrap();
+        let mut doc = Rope::from("\n");
+        let (transaction, _, snippet) = snippet.render(
+            &doc,
+            &Selection::point(0),
+            |_| (0, 0),
+            &mut SnippetRenderCtx::test_ctx(),
+        );
+        assert!(transaction.apply(&mut doc));
+        assert_eq!(doc, "sizeof()\n");
+        assert!(ActiveSnippet::new(snippet).is_none());
+    }
+}

helix-core/src/snippets/elaborate.rs

@@ -0,0 +1,385 @@
+use std::mem::swap;
+use std::ops::Index;
+use std::sync::Arc;
+
+use anyhow::{anyhow, Result};
+use helix_stdx::rope::RopeSliceExt;
+use helix_stdx::Range;
+use regex_cursor::engines::meta::Builder as RegexBuilder;
+use regex_cursor::engines::meta::Regex;
+use regex_cursor::regex_automata::util::syntax::Config as RegexConfig;
+use ropey::RopeSlice;
+
+use crate::case_conversion::to_lower_case_with;
+use crate::case_conversion::to_upper_case_with;
+use crate::case_conversion::{to_camel_case_with, to_pascal_case_with};
+use crate::snippets::parser::{self, CaseChange, FormatItem};
+use crate::snippets::{TabstopIdx, LAST_TABSTOP_IDX};
+use crate::Tendril;
+
+#[derive(Debug)]
+pub struct Snippet {
+    elements: Vec<SnippetElement>,
+    tabstops: Vec<Tabstop>,
+}
+
+impl Snippet {
+    pub fn parse(snippet: &str) -> Result<Self> {
+        let parsed_snippet = parser::parse(snippet)
+            .map_err(|rest| anyhow!("Failed to parse snippet. Remaining input: {}", rest))?;
+        Ok(Snippet::new(parsed_snippet))
+    }
+
+    pub fn new(elements: Vec<parser::SnippetElement>) -> Snippet {
+        let mut res = Snippet {
+            elements: Vec::new(),
+            tabstops: Vec::new(),
+        };
+        res.elements = res.elaborate(elements, None).into();
+        res.fixup_tabstops();
+        res.ensure_last_tabstop();
+        res.renumber_tabstops();
+        res
+    }
+
+    pub fn elements(&self) -> &[SnippetElement] {
+        &self.elements
+    }
+
+    pub fn tabstops(&self) -> impl Iterator<Item = &Tabstop> {
+        self.tabstops.iter()
+    }
+
+    fn renumber_tabstops(&mut self) {
+        Self::renumber_tabstops_in(&self.tabstops, &mut self.elements);
+        for i in 0..self.tabstops.len() {
+            if let Some(parent) = self.tabstops[i].parent {
+                let parent = self
+                    .tabstops
+                    .binary_search_by_key(&parent, |tabstop| tabstop.idx)
+                    .expect("all tabstops have been resolved");
+                self.tabstops[i].parent = Some(TabstopIdx(parent));
+            }
+            let tabstop = &mut self.tabstops[i];
+            if let TabstopKind::Placeholder { default } = &tabstop.kind {
+                let mut default = default.clone();
+                tabstop.kind = TabstopKind::Empty;
+                Self::renumber_tabstops_in(&self.tabstops, Arc::get_mut(&mut default).unwrap());
+                self.tabstops[i].kind = TabstopKind::Placeholder { default };
+            }
+        }
+    }
+
+    fn renumber_tabstops_in(tabstops: &[Tabstop], elements: &mut [SnippetElement]) {
+        for elem in elements {
+            match elem {
+                SnippetElement::Tabstop { idx } => {
+                    idx.0 = tabstops
+                        .binary_search_by_key(&*idx, |tabstop| tabstop.idx)
+                        .expect("all tabstops have been resolved")
+                }
+                SnippetElement::Variable { default, .. } => {
+                    if let Some(default) = default {
+                        Self::renumber_tabstops_in(tabstops, default);
+                    }
+                }
+                SnippetElement::Text(_) => (),
+            }
+        }
+    }
+
+    fn fixup_tabstops(&mut self) {
+        self.tabstops.sort_by_key(|tabstop| tabstop.idx);
+        self.tabstops.dedup_by(|tabstop1, tabstop2| {
+            if tabstop1.idx != tabstop2.idx {
+                return false;
+            }
+            // use the first non empty tabstop for all multicursor tabstops
+            if tabstop2.kind.is_empty() {
+                swap(tabstop2, tabstop1)
+            }
+            true
+        })
+    }
+
+    fn ensure_last_tabstop(&mut self) {
+        if matches!(self.tabstops.last(), Some(tabstop) if tabstop.idx == LAST_TABSTOP_IDX) {
+            return;
+        }
+        self.tabstops.push(Tabstop {
+            idx: LAST_TABSTOP_IDX,
+            parent: None,
+            kind: TabstopKind::Empty,
+        });
+        self.elements.push(SnippetElement::Tabstop {
+            idx: LAST_TABSTOP_IDX,
+        })
+    }
+
+    fn elaborate(
+        &mut self,
+        default: Vec<parser::SnippetElement>,
+        parent: Option<TabstopIdx>,
+    ) -> Box<[SnippetElement]> {
+        default
+            .into_iter()
+            .map(|val| match val {
+                parser::SnippetElement::Tabstop {
+                    tabstop,
+                    transform: None,
+                } => SnippetElement::Tabstop {
+                    idx: self.elaborate_placeholder(tabstop, parent, Vec::new()),
+                },
+                parser::SnippetElement::Tabstop {
+                    tabstop,
+                    transform: Some(transform),
+                } => SnippetElement::Tabstop {
+                    idx: self.elaborate_transform(tabstop, parent, transform),
+                },
+                parser::SnippetElement::Placeholder { tabstop, value } => SnippetElement::Tabstop {
+                    idx: self.elaborate_placeholder(tabstop, parent, value),
+                },
+                parser::SnippetElement::Choice { tabstop, choices } => SnippetElement::Tabstop {
+                    idx: self.elaborate_choice(tabstop, parent, choices),
+                },
+                parser::SnippetElement::Variable {
+                    name,
+                    default,
+                    transform,
+                } => SnippetElement::Variable {
+                    name,
+                    default: default.map(|default| self.elaborate(default, parent)),
+                    // TODO: error for invalid transforms
+                    transform: transform.and_then(Transform::new).map(Box::new),
+                },
+                parser::SnippetElement::Text(text) => SnippetElement::Text(text),
+            })
+            .collect()
+    }
+
+    fn elaborate_choice(
+        &mut self,
+        idx: usize,
+        parent: Option<TabstopIdx>,
+        choices: Vec<Tendril>,
+    ) -> TabstopIdx {
+        let idx = TabstopIdx::elaborate(idx);
+        self.tabstops.push(Tabstop {
+            idx,
+            parent,
+            kind: TabstopKind::Choice {
+                choices: choices.into(),
+            },
+        });
+        idx
+    }
+
+    fn elaborate_placeholder(
+        &mut self,
+        idx: usize,
+        parent: Option<TabstopIdx>,
+        mut default: Vec<parser::SnippetElement>,
+    ) -> TabstopIdx {
+        let idx = TabstopIdx::elaborate(idx);
+        if idx == LAST_TABSTOP_IDX && !default.is_empty() {
+            // Older versions of clangd for example may send a snippet like `${0:placeholder}`
+            // which is considered by VSCode to be a misuse of the `$0` tabstop.
+            log::warn!("Discarding placeholder text for the `$0` tabstop ({default:?}). \
+                The `$0` tabstop signifies the final cursor position and should not include placeholder text.");
+            default.clear();
+        }
+        let default = self.elaborate(default, Some(idx));
+        self.tabstops.push(Tabstop {
+            idx,
+            parent,
+            kind: TabstopKind::Placeholder {
+                default: default.into(),
+            },
+        });
+        idx
+    }
+
+    fn elaborate_transform(
+        &mut self,
+        idx: usize,
+        parent: Option<TabstopIdx>,
+        transform: parser::Transform,
+    ) -> TabstopIdx {
+        let idx = TabstopIdx::elaborate(idx);
+        if let Some(transform) = Transform::new(transform) {
+            self.tabstops.push(Tabstop {
+                idx,
+                parent,
+                kind: TabstopKind::Transform(Arc::new(transform)),
+            })
+        } else {
+            // TODO: proper error
+            self.tabstops.push(Tabstop {
+                idx,
+                parent,
+                kind: TabstopKind::Empty,
+            })
+        }
+        idx
+    }
+}
+
+impl Index<TabstopIdx> for Snippet {
+    type Output = Tabstop;
+    fn index(&self, index: TabstopIdx) -> &Tabstop {
+        &self.tabstops[index.0]
+    }
+}
+
+#[derive(Debug)]
+pub enum SnippetElement {
+    Tabstop {
+        idx: TabstopIdx,
+    },
+    Variable {
+        name: Tendril,
+        default: Option<Box<[SnippetElement]>>,
+        transform: Option<Box<Transform>>,
+    },
+    Text(Tendril),
+}
+
+#[derive(Debug)]
+pub struct Tabstop {
+    idx: TabstopIdx,
+    pub parent: Option<TabstopIdx>,
+    pub kind: TabstopKind,
+}
+
+#[derive(Debug)]
+pub enum TabstopKind {
+    Choice { choices: Arc<[Tendril]> },
+    Placeholder { default: Arc<[SnippetElement]> },
+    Empty,
+    Transform(Arc<Transform>),
+}
+
+impl TabstopKind {
+    pub fn is_empty(&self) -> bool {
+        matches!(self, TabstopKind::Empty)
+    }
+}
+
+#[derive(Debug)]
+pub struct Transform {
+    regex: Regex,
+    regex_str: Box<str>,
+    global: bool,
+    replacement: Box<[FormatItem]>,
+}
+
+impl PartialEq for Transform {
+    fn eq(&self, other: &Self) -> bool {
+        self.replacement == other.replacement
+            && self.global == other.global
+            // doens't compare m and i setting but close enough
+            && self.regex_str == other.regex_str
+    }
+}
+
+impl Transform {
+    fn new(transform: parser::Transform) -> Option<Transform> {
+        let mut config = RegexConfig::new();
+        let mut global = false;
+        let mut invalid_config = false;
+        for c in transform.options.chars() {
+            match c {
+                'i' => {
+                    config = config.case_insensitive(true);
+                }
+                'm' => {
+                    config = config.multi_line(true);
+                }
+                'g' => {
+                    global = true;
+                }
+                // we ignore 'u' since we always want to
+                // do unicode aware matching
+                _ => invalid_config = true,
+            }
+        }
+        if invalid_config {
+            log::error!("invalid transform configuration characters {transform:?}");
+        }
+        let regex = match RegexBuilder::new().syntax(config).build(&transform.regex) {
+            Ok(regex) => regex,
+            Err(err) => {
+                log::error!("invalid transform {err} {transform:?}");
+                return None;
+            }
+        };
+        Some(Transform {
+            regex,
+            regex_str: transform.regex.as_str().into(),
+            global,
+            replacement: transform.replacement.into(),
+        })
+    }
+
+    pub fn apply(&self, mut doc: RopeSlice<'_>, range: Range) -> Tendril {
+        let mut buf = Tendril::new();
+        let it = self
+            .regex
+            .captures_iter(doc.regex_input_at(range))
+            .enumerate();
+        doc = doc.slice(range);
+        let mut last_match = 0;
+        for (_, cap) in it {
+            // unwrap on 0 is OK because captures only reports matches
+            let m = cap.get_group(0).unwrap();
+            buf.extend(doc.byte_slice(last_match..m.start).chunks());
+            last_match = m.end;
+            for fmt in &*self.replacement {
+                match *fmt {
+                    FormatItem::Text(ref text) => {
+                        buf.push_str(text);
+                    }
+                    FormatItem::Capture(i) => {
+                        if let Some(cap) = cap.get_group(i) {
+                            buf.extend(doc.byte_slice(cap.range()).chunks());
+                        }
+                    }
+                    FormatItem::CaseChange(i, change) => {
+                        if let Some(cap) = cap.get_group(i).filter(|i| !i.is_empty()) {
+                            let mut chars = doc.byte_slice(cap.range()).chars();
+                            match change {
+                                CaseChange::Upcase => to_upper_case_with(chars, &mut buf),
+                                CaseChange::Downcase => to_lower_case_with(chars, &mut buf),
+                                CaseChange::Capitalize => {
+                                    let first_char = chars.next().unwrap();
+                                    buf.extend(first_char.to_uppercase());
+                                    buf.extend(chars);
+                                }
+                                CaseChange::PascalCase => to_pascal_case_with(chars, &mut buf),
+                                CaseChange::CamelCase => to_camel_case_with(chars, &mut buf),
+                            }
+                        }
+                    }
+                    FormatItem::Conditional(i, ref if_, ref else_) => {
+                        if cap.get_group(i).map_or(true, |mat| mat.is_empty()) {
+                            buf.push_str(else_)
+                        } else {
+                            buf.push_str(if_)
+                        }
+                    }
+                }
+            }
+            if !self.global {
+                break;
+            }
+        }
+        buf.extend(doc.byte_slice(last_match..).chunks());
+        buf
+    }
+}
+
+impl TabstopIdx {
+    fn elaborate(idx: usize) -> Self {
+        TabstopIdx(idx.wrapping_sub(1))
+    }
+}

helix-core/src/snippets/parser.rs

@@ -0,0 +1,935 @@
+/*!
+A parser for LSP/VSCode style snippet syntax
+See <https://microsoft.github.io/language-server-protocol/specifications/lsp/3.17/specification/#snippet_syntax>.
+
+``` text
+any         ::= tabstop | placeholder | choice | variable | text
+tabstop     ::= '$' int | '${' int '}'
+placeholder ::= '${' int ':' any '}'
+choice      ::= '${' int '|' text (',' text)* '|}'
+variable    ::= '$' var | '${' var }'
+                | '${' var ':' any '}'
+                | '${' var '/' regex '/' (format | text)+ '/' options '}'
+format      ::= '$' int | '${' int '}'
+                | '${' int ':' '/upcase' | '/downcase' | '/capitalize' '}'
+                | '${' int ':+' if '}'
+                | '${' int ':?' if ':' else '}'
+                | '${' int ':-' else '}' | '${' int ':' else '}'
+regex       ::= Regular Expression value (ctor-string)
+options     ::= Regular Expression option (ctor-options)
+var         ::= [_a-zA-Z] [_a-zA-Z0-9]*
+int         ::= [0-9]+
+text        ::= .*
+if          ::= text
+else        ::= text
+```
+*/
+
+use crate::Tendril;
+use helix_parsec::*;
+
+#[derive(Debug, PartialEq, Eq, Clone, Copy)]
+pub enum CaseChange {
+    Upcase,
+    Downcase,
+    Capitalize,
+    PascalCase,
+    CamelCase,
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub enum FormatItem {
+    Text(Tendril),
+    Capture(usize),
+    CaseChange(usize, CaseChange),
+    Conditional(usize, Tendril, Tendril),
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub struct Transform {
+    pub regex: Tendril,
+    pub replacement: Vec<FormatItem>,
+    pub options: Tendril,
+}
+
+#[derive(Debug, PartialEq, Eq)]
+pub enum SnippetElement {
+    Tabstop {
+        tabstop: usize,
+        transform: Option<Transform>,
+    },
+    Placeholder {
+        tabstop: usize,
+        value: Vec<SnippetElement>,
+    },
+    Choice {
+        tabstop: usize,
+        choices: Vec<Tendril>,
+    },
+    Variable {
+        name: Tendril,
+        default: Option<Vec<SnippetElement>>,
+        transform: Option<Transform>,
+    },
+    Text(Tendril),
+}
+
+pub fn parse(s: &str) -> Result<Vec<SnippetElement>, &str> {
+    snippet().parse(s).and_then(|(remainder, snippet)| {
+        if remainder.is_empty() {
+            Ok(snippet)
+        } else {
+            Err(remainder)
+        }
+    })
+}
+
+fn var<'a>() -> impl Parser<'a, Output = &'a str> {
+    // var = [_a-zA-Z][_a-zA-Z0-9]*
+    move |input: &'a str| {
+        input
+            .char_indices()
+            .take_while(|(p, c)| {
+                *c == '_'
+                    || if *p == 0 {
+                        c.is_ascii_alphabetic()
+                    } else {
+                        c.is_ascii_alphanumeric()
+                    }
+            })
+            .last()
+            .map(|(index, c)| {
+                let index = index + c.len_utf8();
+                (&input[index..], &input[0..index])
+            })
+            .ok_or(input)
+    }
+}
+
+const TEXT_ESCAPE_CHARS: &[char] = &['\\', '}', '$'];
+const CHOICE_TEXT_ESCAPE_CHARS: &[char] = &['\\', '|', ','];
+
+fn text<'a>(
+    escape_chars: &'static [char],
+    term_chars: &'static [char],
+) -> impl Parser<'a, Output = Tendril> {
+    move |input: &'a str| {
+        let mut chars = input.char_indices().peekable();
+        let mut res = Tendril::new();
+        while let Some((i, c)) = chars.next() {
+            match c {
+                '\\' => {
+                    if let Some(&(_, c)) = chars.peek() {
+                        if escape_chars.contains(&c) {
+                            chars.next();
+                            res.push(c);
+                            continue;
+                        }
+                    }
+                    res.push('\\');
+                }
+                c if term_chars.contains(&c) => return Ok((&input[i..], res)),
+                c => res.push(c),
+            }
+        }
+
+        Ok(("", res))
+    }
+}
+
+fn digit<'a>() -> impl Parser<'a, Output = usize> {
+    filter_map(take_while(|c| c.is_ascii_digit()), |s| s.parse().ok())
+}
+
+fn case_change<'a>() -> impl Parser<'a, Output = CaseChange> {
+    use CaseChange::*;
+
+    choice!(
+        map("upcase", |_| Upcase),
+        map("downcase", |_| Downcase),
+        map("capitalize", |_| Capitalize),
+        map("pascalcase", |_| PascalCase),
+        map("camelcase", |_| CamelCase),
+    )
+}
+
+fn format<'a>() -> impl Parser<'a, Output = FormatItem> {
+    use FormatItem::*;
+
+    choice!(
+        // '$' int
+        map(right("$", digit()), Capture),
+        // '${' int '}'
+        map(seq!("${", digit(), "}"), |seq| Capture(seq.1)),
+        // '${' int ':' '/upcase' | '/downcase' | '/capitalize' '}'
+        map(seq!("${", digit(), ":/", case_change(), "}"), |seq| {
+            CaseChange(seq.1, seq.3)
+        }),
+        // '${' int ':+' if '}'
+        map(
+            seq!("${", digit(), ":+", text(TEXT_ESCAPE_CHARS, &['}']), "}"),
+            |seq| { Conditional(seq.1, seq.3, Tendril::new()) }
+        ),
+        // '${' int ':?' if ':' else '}'
+        map(
+            seq!(
+                "${",
+                digit(),
+                ":?",
+                text(TEXT_ESCAPE_CHARS, &[':']),
+                ":",
+                text(TEXT_ESCAPE_CHARS, &['}']),
+                "}"
+            ),
+            |seq| { Conditional(seq.1, seq.3, seq.5) }
+        ),
+        // '${' int ':-' else '}' | '${' int ':' else '}'
+        map(
+            seq!(
+                "${",
+                digit(),
+                ":",
+                optional("-"),
+                text(TEXT_ESCAPE_CHARS, &['}']),
+                "}"
+            ),
+            |seq| { Conditional(seq.1, Tendril::new(), seq.4) }
+        ),
+    )
+}
+
+fn regex<'a>() -> impl Parser<'a, Output = Transform> {
+    map(
+        seq!(
+            "/",
+            // TODO parse as ECMAScript and convert to rust regex
+            text(&['/'], &['/']),
+            "/",
+            zero_or_more(choice!(
+                format(),
+                // text doesn't parse $, if format fails we just accept the $ as text
+                map("$", |_| FormatItem::Text("$".into())),
+                map(text(&['\\', '/'], &['/', '$']), FormatItem::Text),
+            )),
+            "/",
+            // vscode really doesn't allow escaping } here
+            // so it's impossible to write a regex escape containing a }
+            // we can consider deviating here and allowing the escape
+            text(&[], &['}']),
+        ),
+        |(_, value, _, replacement, _, options)| Transform {
+            regex: value,
+            replacement,
+            options,
+        },
+    )
+}
+
+fn tabstop<'a>() -> impl Parser<'a, Output = SnippetElement> {
+    map(
+        or(
+            map(right("$", digit()), |i| (i, None)),
+            map(
+                seq!("${", digit(), optional(regex()), "}"),
+                |(_, i, transform, _)| (i, transform),
+            ),
+        ),
+        |(tabstop, transform)| SnippetElement::Tabstop { tabstop, transform },
+    )
+}
+
+fn placeholder<'a>() -> impl Parser<'a, Output = SnippetElement> {
+    map(
+        seq!(
+            "${",
+            digit(),
+            ":",
+            // according to the grammar there is just a single anything here.
+            // However in the prose it is explained that placeholders can be nested.
+            // The example there contains both a placeholder text and a nested placeholder
+            // which indicates a list. Looking at the VSCode sourcecode, the placeholder
+            // is indeed parsed as zero_or_more so the grammar is simply incorrect here
+            zero_or_more(anything(TEXT_ESCAPE_CHARS, true)),
+            "}"
+        ),
+        |seq| SnippetElement::Placeholder {
+            tabstop: seq.1,
+            value: seq.3,
+        },
+    )
+}
+
+fn choice<'a>() -> impl Parser<'a, Output = SnippetElement> {
+    map(
+        seq!(
+            "${",
+            digit(),
+            "|",
+            sep(text(CHOICE_TEXT_ESCAPE_CHARS, &['|', ',']), ","),
+            "|}",
+        ),
+        |seq| SnippetElement::Choice {
+            tabstop: seq.1,
+            choices: seq.3,
+        },
+    )
+}
+
+fn variable<'a>() -> impl Parser<'a, Output = SnippetElement> {
+    choice!(
+        // $var
+        map(right("$", var()), |name| SnippetElement::Variable {
+            name: name.into(),
+            default: None,
+            transform: None,
+        }),
+        // ${var}
+        map(seq!("${", var(), "}",), |values| SnippetElement::Variable {
+            name: values.1.into(),
+            default: None,
+            transform: None,
+        }),
+        // ${var:default}
+        map(
+            seq!(
+                "${",
+                var(),
+                ":",
+                zero_or_more(anything(TEXT_ESCAPE_CHARS, true)),
+                "}",
+            ),
+            |values| SnippetElement::Variable {
+                name: values.1.into(),
+                default: Some(values.3),
+                transform: None,
+            }
+        ),
+        // ${var/value/format/options}
+        map(seq!("${", var(), regex(), "}"), |values| {
+            SnippetElement::Variable {
+                name: values.1.into(),
+                default: None,
+                transform: Some(values.2),
+            }
+        }),
+    )
+}
+
+fn anything<'a>(
+    escape_chars: &'static [char],
+    end_at_brace: bool,
+) -> impl Parser<'a, Output = SnippetElement> {
+    let term_chars: &[_] = if end_at_brace { &['$', '}'] } else { &['$'] };
+    move |input: &'a str| {
+        let parser = choice!(
+            tabstop(),
+            placeholder(),
+            choice(),
+            variable(),
+            map("$", |_| SnippetElement::Text("$".into())),
+            map(text(escape_chars, term_chars), SnippetElement::Text),
+        );
+        parser.parse(input)
+    }
+}
+
+fn snippet<'a>() -> impl Parser<'a, Output = Vec<SnippetElement>> {
+    one_or_more(anything(TEXT_ESCAPE_CHARS, false))
+}
+
+#[cfg(test)]
+mod test {
+    use crate::snippets::{Snippet, SnippetRenderCtx};
+
+    use super::SnippetElement::*;
+    use super::*;
+
+    #[test]
+    fn empty_string_is_error() {
+        assert_eq!(Err(""), parse(""));
+    }
+
+    #[test]
+    fn parse_placeholders_in_function_call() {
+        assert_eq!(
+            Ok(vec![
+                Text("match(".into()),
+                Placeholder {
+                    tabstop: 1,
+                    value: vec![Text("Arg1".into())],
+                },
+                Text(")".into()),
+            ]),
+            parse("match(${1:Arg1})")
+        );
+        // The `$0` tabstop should not have placeholder text. The parser should handle this case
+        // normally and then the placeholder text should be discarded during elaboration.
+        assert_eq!(
+            Ok(vec![
+                Text("sizeof(".into()),
+                Placeholder {
+                    tabstop: 0,
+                    value: vec![Text("expression-or-type".into())],
+                },
+                Text(")".into()),
+            ]),
+            parse("sizeof(${0:expression-or-type})")
+        );
+    }
+
+    #[test]
+    fn unterminated_placeholder() {
+        assert_eq!(
+            Ok(vec![
+                Text("match(".into()),
+                Text("$".into()),
+                Text("{1:)".into())
+            ]),
+            parse("match(${1:)")
+        )
+    }
+
+    #[test]
+    fn parse_empty_placeholder() {
+        assert_eq!(
+            Ok(vec![
+                Text("match(".into()),
+                Placeholder {
+                    tabstop: 1,
+                    value: vec![],
+                },
+                Text(")".into()),
+            ]),
+            parse("match(${1:})")
+        )
+    }
+
+    #[test]
+    fn parse_placeholders_in_statement() {
+        assert_eq!(
+            Ok(vec![
+                Text("local ".into()),
+                Placeholder {
+                    tabstop: 1,
+                    value: vec![Text("var".into())],
+                },
+                Text(" = ".into()),
+                Placeholder {
+                    tabstop: 1,
+                    value: vec![Text("value".into())],
+                },
+            ]),
+            parse("local ${1:var} = ${1:value}")
+        )
+    }
+
+    #[test]
+    fn parse_tabstop_nested_in_placeholder() {
+        assert_eq!(
+            Ok(vec![Placeholder {
+                tabstop: 1,
+                value: vec![
+                    Text("var, ".into()),
+                    Tabstop {
+                        tabstop: 2,
+                        transform: None
+                    }
+                ],
+            }]),
+            parse("${1:var, $2}")
+        )
+    }
+
+    #[test]
+    fn parse_placeholder_nested_in_placeholder() {
+        assert_eq!(
+            Ok({
+                vec![Placeholder {
+                    tabstop: 1,
+                    value: vec![
+                        Text("foo ".into()),
+                        Placeholder {
+                            tabstop: 2,
+                            value: vec![Text("bar".into())],
+                        },
+                    ],
+                }]
+            }),
+            parse("${1:foo ${2:bar}}")
+        )
+    }
+
+    #[test]
+    fn parse_all() {
+        assert_eq!(
+            Ok(vec![
+                Text("hello ".into()),
+                Tabstop {
+                    tabstop: 1,
+                    transform: None
+                },
+                Tabstop {
+                    tabstop: 2,
+                    transform: None
+                },
+                Text(" ".into()),
+                Choice {
+                    tabstop: 1,
+                    choices: vec!["one".into(), "two".into(), "three".into()],
+                },
+                Text(" ".into()),
+                Variable {
+                    name: "name".into(),
+                    default: Some(vec![Text("foo".into())]),
+                    transform: None,
+                },
+                Text(" ".into()),
+                Variable {
+                    name: "var".into(),
+                    default: None,
+                    transform: None,
+                },
+                Text(" ".into()),
+                Variable {
+                    name: "TM".into(),
+                    default: None,
+                    transform: None,
+                },
+            ]),
+            parse("hello $1${2} ${1|one,two,three|} ${name:foo} $var $TM")
+        );
+    }
+
+    #[test]
+    fn regex_capture_replace() {
+        assert_eq!(
+            Ok({
+                vec![Variable {
+                    name: "TM_FILENAME".into(),
+                    default: None,
+                    transform: Some(Transform {
+                        regex: "(.*).+$".into(),
+                        replacement: vec![FormatItem::Capture(1), FormatItem::Text("$".into())],
+                        options: Tendril::new(),
+                    }),
+                }]
+            }),
+            parse("${TM_FILENAME/(.*).+$/$1$/}")
+        );
+    }
+
+    #[test]
+    fn rust_macro() {
+        assert_eq!(
+            Ok({
+                vec![
+                    Text("macro_rules! ".into()),
+                    Tabstop {
+                        tabstop: 1,
+                        transform: None,
+                    },
+                    Text(" {\n    (".into()),
+                    Tabstop {
+                        tabstop: 2,
+                        transform: None,
+                    },
+                    Text(") => {\n        ".into()),
+                    Tabstop {
+                        tabstop: 0,
+                        transform: None,
+                    },
+                    Text("\n    };\n}".into()),
+                ]
+            }),
+            parse("macro_rules! $1 {\n    ($2) => {\n        $0\n    };\n}")
+        );
+    }
+
+    fn assert_text(snippet: &str, parsed_text: &str) {
+        let snippet = Snippet::parse(snippet).unwrap();
+        let mut rendered_snippet = snippet.prepare_render();
+        let rendered_text = snippet
+            .render_at(
+                &mut rendered_snippet,
+                "".into(),
+                false,
+                &mut SnippetRenderCtx::test_ctx(),
+                0,
+            )
+            .0;
+        assert_eq!(rendered_text, parsed_text)
+    }
+
+    #[test]
+    fn robust_parsing() {
+        assert_text("$", "$");
+        assert_text("\\\\$", "\\$");
+        assert_text("{", "{");
+        assert_text("\\}", "}");
+        assert_text("\\abc", "\\abc");
+        assert_text("foo${f:\\}}bar", "foo}bar");
+        assert_text("\\{", "\\{");
+        assert_text("I need \\\\\\$", "I need \\$");
+        assert_text("\\", "\\");
+        assert_text("\\{{", "\\{{");
+        assert_text("{{", "{{");
+        assert_text("{{dd", "{{dd");
+        assert_text("}}", "}}");
+        assert_text("ff}}", "ff}}");
+        assert_text("farboo", "farboo");
+        assert_text("far{{}}boo", "far{{}}boo");
+        assert_text("far{{123}}boo", "far{{123}}boo");
+        assert_text("far\\{{123}}boo", "far\\{{123}}boo");
+        assert_text("far{{id:bern}}boo", "far{{id:bern}}boo");
+        assert_text("far{{id:bern {{basel}}}}boo", "far{{id:bern {{basel}}}}boo");
+        assert_text(
+            "far{{id:bern {{id:basel}}}}boo",
+            "far{{id:bern {{id:basel}}}}boo",
+        );
+        assert_text(
+            "far{{id:bern {{id2:basel}}}}boo",
+            "far{{id:bern {{id2:basel}}}}boo",
+        );
+        assert_text("${}$\\a\\$\\}\\\\", "${}$\\a$}\\");
+        assert_text("farboo", "farboo");
+        assert_text("far{{}}boo", "far{{}}boo");
+        assert_text("far{{123}}boo", "far{{123}}boo");
+        assert_text("far\\{{123}}boo", "far\\{{123}}boo");
+        assert_text("far`123`boo", "far`123`boo");
+        assert_text("far\\`123\\`boo", "far\\`123\\`boo");
+        assert_text("\\$far-boo", "$far-boo");
+    }
+
+    fn assert_snippet(snippet: &str, expect: &[SnippetElement]) {
+        let elements = parse(snippet).unwrap();
+        assert_eq!(elements, expect.to_owned())
+    }
+
+    #[test]
+    fn parse_variable() {
+        use SnippetElement::*;
+        assert_snippet(
+            "$far-boo",
+            &[
+                Variable {
+                    name: "far".into(),
+                    default: None,
+                    transform: None,
+                },
+                Text("-boo".into()),
+            ],
+        );
+        assert_snippet(
+            "far$farboo",
+            &[
+                Text("far".into()),
+                Variable {
+                    name: "farboo".into(),
+                    transform: None,
+                    default: None,
+                },
+            ],
+        );
+        assert_snippet(
+            "far${farboo}",
+            &[
+                Text("far".into()),
+                Variable {
+                    name: "farboo".into(),
+                    transform: None,
+                    default: None,
+                },
+            ],
+        );
+        assert_snippet(
+            "$123",
+            &[Tabstop {
+                tabstop: 123,
+                transform: None,
+            }],
+        );
+        assert_snippet(
+            "$farboo",
+            &[Variable {
+                name: "farboo".into(),
+                transform: None,
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "$far12boo",
+            &[Variable {
+                name: "far12boo".into(),
+                transform: None,
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "000_${far}_000",
+            &[
+                Text("000_".into()),
+                Variable {
+                    name: "far".into(),
+                    transform: None,
+                    default: None,
+                },
+                Text("_000".into()),
+            ],
+        );
+    }
+
+    #[test]
+    fn parse_variable_transform() {
+        assert_snippet(
+            "${foo///}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: Tendril::new(),
+                    replacement: Vec::new(),
+                    options: Tendril::new(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/regex/format/gmi}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: "regex".into(),
+                    replacement: vec![FormatItem::Text("format".into())],
+                    options: "gmi".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/([A-Z][a-z])/format/}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: "([A-Z][a-z])".into(),
+                    replacement: vec![FormatItem::Text("format".into())],
+                    options: Tendril::new(),
+                }),
+                default: None,
+            }],
+        );
+
+        // invalid regex TODO: reneable tests once we actually parse this regex flavor
+        // assert_text(
+        //     "${foo/([A-Z][a-z])/format/GMI}",
+        //     "${foo/([A-Z][a-z])/format/GMI}",
+        // );
+        // assert_text(
+        //     "${foo/([A-Z][a-z])/format/funky}",
+        //     "${foo/([A-Z][a-z])/format/funky}",
+        // );
+        // assert_text("${foo/([A-Z][a-z]/format/}", "${foo/([A-Z][a-z]/format/}");
+        assert_text(
+            "${foo/regex\\/format/options}",
+            "${foo/regex\\/format/options}",
+        );
+
+        // tricky regex
+        assert_snippet(
+            "${foo/m\\/atch/$1/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: "m/atch".into(),
+                    replacement: vec![FormatItem::Capture(1)],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+
+        // incomplete
+        assert_text("${foo///", "${foo///");
+        assert_text("${foo/regex/format/options", "${foo/regex/format/options");
+
+        // format string
+        assert_snippet(
+            "${foo/.*/${0:fooo}/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![FormatItem::Conditional(0, Tendril::new(), "fooo".into())],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/.*/${1}/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![FormatItem::Capture(1)],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/.*/$1/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![FormatItem::Capture(1)],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/.*/This-$1-encloses/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![
+                        FormatItem::Text("This-".into()),
+                        FormatItem::Capture(1),
+                        FormatItem::Text("-encloses".into()),
+                    ],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/.*/complex${1:else}/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![
+                        FormatItem::Text("complex".into()),
+                        FormatItem::Conditional(1, Tendril::new(), "else".into()),
+                    ],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/.*/complex${1:-else}/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![
+                        FormatItem::Text("complex".into()),
+                        FormatItem::Conditional(1, Tendril::new(), "else".into()),
+                    ],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/.*/complex${1:+if}/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![
+                        FormatItem::Text("complex".into()),
+                        FormatItem::Conditional(1, "if".into(), Tendril::new()),
+                    ],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/.*/complex${1:?if:else}/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![
+                        FormatItem::Text("complex".into()),
+                        FormatItem::Conditional(1, "if".into(), "else".into()),
+                    ],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${foo/.*/complex${1:/upcase}/i}",
+            &[Variable {
+                name: "foo".into(),
+                transform: Some(Transform {
+                    regex: ".*".into(),
+                    replacement: vec![
+                        FormatItem::Text("complex".into()),
+                        FormatItem::CaseChange(1, CaseChange::Upcase),
+                    ],
+                    options: "i".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${TM_DIRECTORY/src\\//$1/}",
+            &[Variable {
+                name: "TM_DIRECTORY".into(),
+                transform: Some(Transform {
+                    regex: "src/".into(),
+                    replacement: vec![FormatItem::Capture(1)],
+                    options: Tendril::new(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${TM_SELECTED_TEXT/a/\\/$1/g}",
+            &[Variable {
+                name: "TM_SELECTED_TEXT".into(),
+                transform: Some(Transform {
+                    regex: "a".into(),
+                    replacement: vec![FormatItem::Text("/".into()), FormatItem::Capture(1)],
+                    options: "g".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${TM_SELECTED_TEXT/a/in\\/$1ner/g}",
+            &[Variable {
+                name: "TM_SELECTED_TEXT".into(),
+                transform: Some(Transform {
+                    regex: "a".into(),
+                    replacement: vec![
+                        FormatItem::Text("in/".into()),
+                        FormatItem::Capture(1),
+                        FormatItem::Text("ner".into()),
+                    ],
+                    options: "g".into(),
+                }),
+                default: None,
+            }],
+        );
+        assert_snippet(
+            "${TM_SELECTED_TEXT/a/end\\//g}",
+            &[Variable {
+                name: "TM_SELECTED_TEXT".into(),
+                transform: Some(Transform {
+                    regex: "a".into(),
+                    replacement: vec![FormatItem::Text("end/".into())],
+                    options: "g".into(),
+                }),
+                default: None,
+            }],
+        );
+    }
+    // TODO port more tests from https://github.com/microsoft/vscode/blob/dce493cb6e36346ef2714e82c42ce14fc461b15c/src/vs/editor/contrib/snippet/test/browser/snippetParser.test.ts
+}

helix-core/src/snippets/render.rs

@@ -0,0 +1,355 @@
+use std::borrow::Cow;
+use std::ops::{Index, IndexMut};
+use std::sync::Arc;
+
+use helix_stdx::Range;
+use ropey::{Rope, RopeSlice};
+use smallvec::SmallVec;
+
+use crate::indent::{normalize_indentation, IndentStyle};
+use crate::movement::Direction;
+use crate::snippets::elaborate;
+use crate::snippets::TabstopIdx;
+use crate::snippets::{Snippet, SnippetElement, Transform};
+use crate::{selection, Selection, Tendril, Transaction};
+
+#[derive(Debug, Clone, PartialEq)]
+pub enum TabstopKind {
+    Choice { choices: Arc<[Tendril]> },
+    Placeholder,
+    Empty,
+    Transform(Arc<Transform>),
+}
+
+#[derive(Debug, PartialEq)]
+pub struct Tabstop {
+    pub ranges: SmallVec<[Range; 1]>,
+    pub parent: Option<TabstopIdx>,
+    pub kind: TabstopKind,
+}
+
+impl Tabstop {
+    pub fn has_placeholder(&self) -> bool {
+        matches!(
+            self.kind,
+            TabstopKind::Choice { .. } | TabstopKind::Placeholder
+        )
+    }
+
+    pub fn selection(
+        &self,
+        direction: Direction,
+        primary_idx: usize,
+        snippet_ranges: usize,
+    ) -> Selection {
+        Selection::new(
+            self.ranges
+                .iter()
+                .map(|&range| {
+                    let mut range = selection::Range::new(range.start, range.end);
+                    if direction == Direction::Backward {
+                        range = range.flip()
+                    }
+                    range
+                })
+                .collect(),
+            primary_idx * (self.ranges.len() / snippet_ranges),
+        )
+    }
+}
+
+#[derive(Debug, Default, PartialEq)]
+pub struct RenderedSnippet {
+    pub tabstops: Vec<Tabstop>,
+    pub ranges: Vec<Range>,
+}
+
+impl RenderedSnippet {
+    pub fn first_selection(&self, direction: Direction, primary_idx: usize) -> Selection {
+        self.tabstops[0].selection(direction, primary_idx, self.ranges.len())
+    }
+}
+
+impl Index<TabstopIdx> for RenderedSnippet {
+    type Output = Tabstop;
+    fn index(&self, index: TabstopIdx) -> &Tabstop {
+        &self.tabstops[index.0]
+    }
+}
+
+impl IndexMut<TabstopIdx> for RenderedSnippet {
+    fn index_mut(&mut self, index: TabstopIdx) -> &mut Tabstop {
+        &mut self.tabstops[index.0]
+    }
+}
+
+impl Snippet {
+    pub fn prepare_render(&self) -> RenderedSnippet {
+        let tabstops =
+            self.tabstops()
+                .map(|tabstop| Tabstop {
+                    ranges: SmallVec::new(),
+                    parent: tabstop.parent,
+                    kind: match &tabstop.kind {
+                        elaborate::TabstopKind::Choice { choices } => TabstopKind::Choice {
+                            choices: choices.clone(),
+                        },
+                        // start out as empty: the first non-empty placeholder will change this to
+                        // a placeholder automatically
+                        elaborate::TabstopKind::Empty
+                        | elaborate::TabstopKind::Placeholder { .. } => TabstopKind::Empty,
+                        elaborate::TabstopKind::Transform(transform) => {
+                            TabstopKind::Transform(transform.clone())
+                        }
+                    },
+                })
+                .collect();
+        RenderedSnippet {
+            tabstops,
+            ranges: Vec::new(),
+        }
+    }
+
+    pub fn render_at(
+        &self,
+        snippet: &mut RenderedSnippet,
+        indent: RopeSlice<'_>,
+        at_newline: bool,
+        ctx: &mut SnippetRenderCtx,
+        pos: usize,
+    ) -> (Tendril, usize) {
+        let mut ctx = SnippetRender {
+            dst: snippet,
+            src: self,
+            indent,
+            text: Tendril::new(),
+            off: pos,
+            ctx,
+            at_newline,
+        };
+        ctx.render_elements(self.elements());
+        let end = ctx.off;
+        let text = ctx.text;
+        snippet.ranges.push(Range { start: pos, end });
+        (text, end - pos)
+    }
+
+    pub fn render(
+        &self,
+        doc: &Rope,
+        selection: &Selection,
+        change_range: impl FnMut(&selection::Range) -> (usize, usize),
+        ctx: &mut SnippetRenderCtx,
+    ) -> (Transaction, Selection, RenderedSnippet) {
+        let mut snippet = self.prepare_render();
+        let mut off = 0;
+        let (transaction, selection) = Transaction::change_by_selection_ignore_overlapping(
+            doc,
+            selection,
+            change_range,
+            |replacement_start, replacement_end| {
+                let line_idx = doc.char_to_line(replacement_start);
+                let line_start = doc.line_to_char(line_idx);
+                let prefix = doc.slice(line_start..replacement_start);
+                let indent_len = prefix.chars().take_while(|c| c.is_whitespace()).count();
+                let indent = prefix.slice(..indent_len);
+                let at_newline = indent_len == replacement_start - line_start;
+
+                let (replacement, replacement_len) = self.render_at(
+                    &mut snippet,
+                    indent,
+                    at_newline,
+                    ctx,
+                    (replacement_start as i128 + off) as usize,
+                );
+                off +=
+                    replacement_start as i128 - replacement_end as i128 + replacement_len as i128;
+
+                Some(replacement)
+            },
+        );
+        (transaction, selection, snippet)
+    }
+}
+
+pub type VariableResolver = dyn FnMut(&str) -> Option<Cow<str>>;
+pub struct SnippetRenderCtx {
+    pub resolve_var: Box<VariableResolver>,
+    pub tab_width: usize,
+    pub indent_style: IndentStyle,
+    pub line_ending: &'static str,
+}
+
+impl SnippetRenderCtx {
+    #[cfg(test)]
+    pub(super) fn test_ctx() -> SnippetRenderCtx {
+        SnippetRenderCtx {
+            resolve_var: Box::new(|_| None),
+            tab_width: 4,
+            indent_style: IndentStyle::Spaces(4),
+            line_ending: "\n",
+        }
+    }
+}
+
+struct SnippetRender<'a> {
+    ctx: &'a mut SnippetRenderCtx,
+    dst: &'a mut RenderedSnippet,
+    src: &'a Snippet,
+    indent: RopeSlice<'a>,
+    text: Tendril,
+    off: usize,
+    at_newline: bool,
+}
+
+impl SnippetRender<'_> {
+    fn render_elements(&mut self, elements: &[SnippetElement]) {
+        for element in elements {
+            self.render_element(element)
+        }
+    }
+
+    fn render_element(&mut self, element: &SnippetElement) {
+        match *element {
+            SnippetElement::Tabstop { idx } => self.render_tabstop(idx),
+            SnippetElement::Variable {
+                ref name,
+                ref default,
+                ref transform,
+            } => {
+                // TODO: allow resolve_var access to the doc and make it return rope slice
+                // so we can access selections and other document content without allocating
+                if let Some(val) = (self.ctx.resolve_var)(name) {
+                    if let Some(transform) = transform {
+                        self.push_multiline_str(&transform.apply(
+                            (&*val).into(),
+                            Range {
+                                start: 0,
+                                end: val.chars().count(),
+                            },
+                        ));
+                    } else {
+                        self.push_multiline_str(&val)
+                    }
+                } else if let Some(default) = default {
+                    self.render_elements(default)
+                }
+            }
+            SnippetElement::Text(ref text) => self.push_multiline_str(text),
+        }
+    }
+
+    fn push_multiline_str(&mut self, text: &str) {
+        let mut lines = text
+            .split('\n')
+            .map(|line| line.strip_suffix('\r').unwrap_or(line));
+        let first_line = lines.next().unwrap();
+        self.push_str(first_line, self.at_newline);
+        for line in lines {
+            self.push_newline();
+            self.push_str(line, true);
+        }
+    }
+
+    fn push_str(&mut self, mut text: &str, at_newline: bool) {
+        if at_newline {
+            let old_len = self.text.len();
+            let old_indent_len = normalize_indentation(
+                self.indent,
+                text.into(),
+                &mut self.text,
+                self.ctx.indent_style,
+                self.ctx.tab_width,
+            );
+            // this is ok because indentation can only be ascii chars (' ' and '\t')
+            self.off += self.text.len() - old_len;
+            text = &text[old_indent_len..];
+            if text.is_empty() {
+                self.at_newline = true;
+                return;
+            }
+        }
+        self.text.push_str(text);
+        self.off += text.chars().count();
+    }
+
+    fn push_newline(&mut self) {
+        self.off += self.ctx.line_ending.chars().count() + self.indent.len_chars();
+        self.text.push_str(self.ctx.line_ending);
+        self.text.extend(self.indent.chunks());
+    }
+
+    fn render_tabstop(&mut self, tabstop: TabstopIdx) {
+        let start = self.off;
+        let end = match &self.src[tabstop].kind {
+            elaborate::TabstopKind::Placeholder { default } if !default.is_empty() => {
+                self.render_elements(default);
+                self.dst[tabstop].kind = TabstopKind::Placeholder;
+                self.off
+            }
+            _ => start,
+        };
+        self.dst[tabstop].ranges.push(Range { start, end });
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use helix_stdx::Range;
+
+    use crate::snippets::render::Tabstop;
+    use crate::snippets::{Snippet, SnippetRenderCtx};
+
+    use super::TabstopKind;
+
+    fn assert_snippet(snippet: &str, expect: &str, tabstops: &[Tabstop]) {
+        let snippet = Snippet::parse(snippet).unwrap();
+        let mut rendered_snippet = snippet.prepare_render();
+        let rendered_text = snippet
+            .render_at(
+                &mut rendered_snippet,
+                "\t".into(),
+                false,
+                &mut SnippetRenderCtx::test_ctx(),
+                0,
+            )
+            .0;
+        assert_eq!(rendered_text, expect);
+        assert_eq!(&rendered_snippet.tabstops, tabstops);
+        assert_eq!(
+            rendered_snippet.ranges.last().unwrap().end,
+            rendered_text.chars().count()
+        );
+        assert_eq!(rendered_snippet.ranges.last().unwrap().start, 0)
+    }
+
+    #[test]
+    fn rust_macro() {
+        assert_snippet(
+            "macro_rules! ${1:name} {\n\t($3) => {\n\t\t$2\n\t};\n}",
+            "macro_rules! name {\n\t    () => {\n\t        \n\t    };\n\t}",
+            &[
+                Tabstop {
+                    ranges: vec![Range { start: 13, end: 17 }].into(),
+                    parent: None,
+                    kind: TabstopKind::Placeholder,
+                },
+                Tabstop {
+                    ranges: vec![Range { start: 42, end: 42 }].into(),
+                    parent: None,
+                    kind: TabstopKind::Empty,
+                },
+                Tabstop {
+                    ranges: vec![Range { start: 26, end: 26 }].into(),
+                    parent: None,
+                    kind: TabstopKind::Empty,
+                },
+                Tabstop {
+                    ranges: vec![Range { start: 53, end: 53 }].into(),
+                    parent: None,
+                    kind: TabstopKind::Empty,
+                },
+            ],
+        );
+    }
+}

helix-core/src/syntax/config.rs

@@ -0,0 +1,629 @@
+use crate::{auto_pairs::AutoPairs, diagnostic::Severity, Language};
+
+use globset::GlobSet;
+use helix_stdx::rope;
+use serde::{ser::SerializeSeq as _, Deserialize, Serialize};
+
+use std::{
+    collections::{HashMap, HashSet},
+    fmt::{self, Display},
+    num::NonZeroU8,
+    path::PathBuf,
+    str::FromStr,
+};
+
+#[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "kebab-case")]
+pub struct Configuration {
+    pub language: Vec<LanguageConfiguration>,
+    #[serde(default)]
+    pub language_server: HashMap<String, LanguageServerConfiguration>,
+}
+
+#[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "kebab-case", deny_unknown_fields)]
+pub struct LanguageConfiguration {
+    #[serde(skip)]
+    pub(super) language: Option<Language>,
+
+    #[serde(rename = "name")]
+    pub language_id: String, // c-sharp, rust, tsx
+    #[serde(rename = "language-id")]
+    // see the table under https://microsoft.github.io/language-server-protocol/specifications/lsp/3.17/specification/#textDocumentItem
+    pub language_server_language_id: Option<String>, // csharp, rust, typescriptreact, for the language-server
+    pub scope: String,             // source.rust
+    pub file_types: Vec<FileType>, // filename extension or ends_with? <Gemfile, rb, etc>
+    #[serde(default)]
+    pub shebangs: Vec<String>, // interpreter(s) associated with language
+    #[serde(default)]
+    pub roots: Vec<String>, // these indicate project roots <.git, Cargo.toml>
+    #[serde(
+        default,
+        skip_serializing,
+        deserialize_with = "from_comment_tokens",
+        alias = "comment-token"
+    )]
+    pub comment_tokens: Option<Vec<String>>,
+    #[serde(
+        default,
+        skip_serializing,
+        deserialize_with = "from_block_comment_tokens"
+    )]
+    pub block_comment_tokens: Option<Vec<BlockCommentToken>>,
+    pub text_width: Option<usize>,
+    pub soft_wrap: Option<SoftWrap>,
+
+    #[serde(default)]
+    pub auto_format: bool,
+
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub formatter: Option<FormatterConfiguration>,
+
+    /// If set, overrides `editor.path-completion`.
+    pub path_completion: Option<bool>,
+    /// If set, overrides `editor.word-completion`.
+    pub word_completion: Option<WordCompletion>,
+
+    #[serde(default)]
+    pub diagnostic_severity: Severity,
+
+    pub grammar: Option<String>, // tree-sitter grammar name, defaults to language_id
+
+    // content_regex
+    #[serde(default, skip_serializing, deserialize_with = "deserialize_regex")]
+    pub injection_regex: Option<rope::Regex>,
+    // first_line_regex
+    //
+    #[serde(
+        default,
+        skip_serializing_if = "Vec::is_empty",
+        serialize_with = "serialize_lang_features",
+        deserialize_with = "deserialize_lang_features"
+    )]
+    pub language_servers: Vec<LanguageServerFeatures>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub indent: Option<IndentationConfiguration>,
+
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub debugger: Option<DebugAdapterConfig>,
+
+    /// Automatic insertion of pairs to parentheses, brackets,
+    /// etc. Defaults to true. Optionally, this can be a list of 2-tuples
+    /// to specify a list of characters to pair. This overrides the
+    /// global setting.
+    #[serde(default, skip_serializing, deserialize_with = "deserialize_auto_pairs")]
+    pub auto_pairs: Option<AutoPairs>,
+
+    pub rulers: Option<Vec<u16>>, // if set, override editor's rulers
+
+    /// Hardcoded LSP root directories relative to the workspace root, like `examples` or `tools/fuzz`.
+    /// Falling back to the current working directory if none are configured.
+    pub workspace_lsp_roots: Option<Vec<PathBuf>>,
+    #[serde(default)]
+    pub persistent_diagnostic_sources: Vec<String>,
+    /// Overrides the `editor.rainbow-brackets` config key for the language.
+    pub rainbow_brackets: Option<bool>,
+}
+
+impl LanguageConfiguration {
+    pub fn language(&self) -> Language {
+        // This value must be set by `super::Loader::new`.
+        self.language.unwrap()
+    }
+}
+
+#[derive(Debug, PartialEq, Eq, Hash)]
+pub enum FileType {
+    /// The extension of the file, either the `Path::extension` or the full
+    /// filename if the file does not have an extension.
+    Extension(String),
+    /// A Unix-style path glob. This is compared to the file's absolute path, so
+    /// it can be used to detect files based on their directories. If the glob
+    /// is not an absolute path and does not already start with a glob pattern,
+    /// a glob pattern will be prepended to it.
+    Glob(globset::Glob),
+}
+
+impl Serialize for FileType {
+    fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
+    where
+        S: serde::Serializer,
+    {
+        use serde::ser::SerializeMap;
+
+        match self {
+            FileType::Extension(extension) => serializer.serialize_str(extension),
+            FileType::Glob(glob) => {
+                let mut map = serializer.serialize_map(Some(1))?;
+                map.serialize_entry("glob", glob.glob())?;
+                map.end()
+            }
+        }
+    }
+}
+
+impl<'de> Deserialize<'de> for FileType {
+    fn deserialize<D>(deserializer: D) -> Result<Self, D::Error>
+    where
+        D: serde::de::Deserializer<'de>,
+    {
+        struct FileTypeVisitor;
+
+        impl<'de> serde::de::Visitor<'de> for FileTypeVisitor {
+            type Value = FileType;
+
+            fn expecting(&self, formatter: &mut std::fmt::Formatter) -> std::fmt::Result {
+                formatter.write_str("string or table")
+            }
+
+            fn visit_str<E>(self, value: &str) -> Result<Self::Value, E>
+            where
+                E: serde::de::Error,
+            {
+                Ok(FileType::Extension(value.to_string()))
+            }
+
+            fn visit_map<M>(self, mut map: M) -> Result<Self::Value, M::Error>
+            where
+                M: serde::de::MapAccess<'de>,
+            {
+                match map.next_entry::<String, String>()? {
+                    Some((key, mut glob)) if key == "glob" => {
+                        // If the glob isn't an absolute path or already starts
+                        // with a glob pattern, add a leading glob so we
+                        // properly match relative paths.
+                        if !glob.starts_with('/') && !glob.starts_with("*/") {
+                            glob.insert_str(0, "*/");
+                        }
+
+                        globset::Glob::new(glob.as_str())
+                            .map(FileType::Glob)
+                            .map_err(|err| {
+                                serde::de::Error::custom(format!("invalid `glob` pattern: {}", err))
+                            })
+                    }
+                    Some((key, _value)) => Err(serde::de::Error::custom(format!(
+                        "unknown key in `file-types` list: {}",
+                        key
+                    ))),
+                    None => Err(serde::de::Error::custom(
+                        "expected a `suffix` key in the `file-types` entry",
+                    )),
+                }
+            }
+        }
+
+        deserializer.deserialize_any(FileTypeVisitor)
+    }
+}
+
+fn from_comment_tokens<'de, D>(deserializer: D) -> Result<Option<Vec<String>>, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    #[derive(Deserialize)]
+    #[serde(untagged)]
+    enum CommentTokens {
+        Multiple(Vec<String>),
+        Single(String),
+    }
+    Ok(
+        Option::<CommentTokens>::deserialize(deserializer)?.map(|tokens| match tokens {
+            CommentTokens::Single(val) => vec![val],
+            CommentTokens::Multiple(vals) => vals,
+        }),
+    )
+}
+
+#[derive(Clone, Debug, Serialize, Deserialize)]
+pub struct BlockCommentToken {
+    pub start: String,
+    pub end: String,
+}
+
+impl Default for BlockCommentToken {
+    fn default() -> Self {
+        BlockCommentToken {
+            start: "/*".to_string(),
+            end: "*/".to_string(),
+        }
+    }
+}
+
+fn from_block_comment_tokens<'de, D>(
+    deserializer: D,
+) -> Result<Option<Vec<BlockCommentToken>>, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    #[derive(Deserialize)]
+    #[serde(untagged)]
+    enum BlockCommentTokens {
+        Multiple(Vec<BlockCommentToken>),
+        Single(BlockCommentToken),
+    }
+    Ok(
+        Option::<BlockCommentTokens>::deserialize(deserializer)?.map(|tokens| match tokens {
+            BlockCommentTokens::Single(val) => vec![val],
+            BlockCommentTokens::Multiple(vals) => vals,
+        }),
+    )
+}
+
+#[derive(Clone, Copy, Debug, Serialize, Deserialize, PartialEq, Eq, Hash)]
+#[serde(rename_all = "kebab-case")]
+pub enum LanguageServerFeature {
+    Format,
+    GotoDeclaration,
+    GotoDefinition,
+    GotoTypeDefinition,
+    GotoReference,
+    GotoImplementation,
+    // Goto, use bitflags, combining previous Goto members?
+    SignatureHelp,
+    Hover,
+    DocumentHighlight,
+    Completion,
+    CodeAction,
+    WorkspaceCommand,
+    DocumentSymbols,
+    WorkspaceSymbols,
+    // Symbols, use bitflags, see above?
+    Diagnostics,
+    RenameSymbol,
+    InlayHints,
+    DocumentColors,
+}
+
+impl Display for LanguageServerFeature {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        use LanguageServerFeature::*;
+        let feature = match self {
+            Format => "format",
+            GotoDeclaration => "goto-declaration",
+            GotoDefinition => "goto-definition",
+            GotoTypeDefinition => "goto-type-definition",
+            GotoReference => "goto-reference",
+            GotoImplementation => "goto-implementation",
+            SignatureHelp => "signature-help",
+            Hover => "hover",
+            DocumentHighlight => "document-highlight",
+            Completion => "completion",
+            CodeAction => "code-action",
+            WorkspaceCommand => "workspace-command",
+            DocumentSymbols => "document-symbols",
+            WorkspaceSymbols => "workspace-symbols",
+            Diagnostics => "diagnostics",
+            RenameSymbol => "rename-symbol",
+            InlayHints => "inlay-hints",
+            DocumentColors => "document-colors",
+        };
+        write!(f, "{feature}",)
+    }
+}
+
+#[derive(Debug, Serialize, Deserialize)]
+#[serde(untagged, rename_all = "kebab-case", deny_unknown_fields)]
+enum LanguageServerFeatureConfiguration {
+    #[serde(rename_all = "kebab-case")]
+    Features {
+        #[serde(default, skip_serializing_if = "HashSet::is_empty")]
+        only_features: HashSet<LanguageServerFeature>,
+        #[serde(default, skip_serializing_if = "HashSet::is_empty")]
+        except_features: HashSet<LanguageServerFeature>,
+        name: String,
+    },
+    Simple(String),
+}
+
+#[derive(Debug, Default)]
+pub struct LanguageServerFeatures {
+    pub name: String,
+    pub only: HashSet<LanguageServerFeature>,
+    pub excluded: HashSet<LanguageServerFeature>,
+}
+
+impl LanguageServerFeatures {
+    pub fn has_feature(&self, feature: LanguageServerFeature) -> bool {
+        (self.only.is_empty() || self.only.contains(&feature)) && !self.excluded.contains(&feature)
+    }
+}
+
+fn deserialize_lang_features<'de, D>(
+    deserializer: D,
+) -> Result<Vec<LanguageServerFeatures>, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    let raw: Vec<LanguageServerFeatureConfiguration> = Deserialize::deserialize(deserializer)?;
+    let res = raw
+        .into_iter()
+        .map(|config| match config {
+            LanguageServerFeatureConfiguration::Simple(name) => LanguageServerFeatures {
+                name,
+                ..Default::default()
+            },
+            LanguageServerFeatureConfiguration::Features {
+                only_features,
+                except_features,
+                name,
+            } => LanguageServerFeatures {
+                name,
+                only: only_features,
+                excluded: except_features,
+            },
+        })
+        .collect();
+    Ok(res)
+}
+fn serialize_lang_features<S>(
+    map: &Vec<LanguageServerFeatures>,
+    serializer: S,
+) -> Result<S::Ok, S::Error>
+where
+    S: serde::Serializer,
+{
+    let mut serializer = serializer.serialize_seq(Some(map.len()))?;
+    for features in map {
+        let features = if features.only.is_empty() && features.excluded.is_empty() {
+            LanguageServerFeatureConfiguration::Simple(features.name.to_owned())
+        } else {
+            LanguageServerFeatureConfiguration::Features {
+                only_features: features.only.clone(),
+                except_features: features.excluded.clone(),
+                name: features.name.to_owned(),
+            }
+        };
+        serializer.serialize_element(&features)?;
+    }
+    serializer.end()
+}
+
+fn deserialize_required_root_patterns<'de, D>(deserializer: D) -> Result<Option<GlobSet>, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    let patterns = Vec::<String>::deserialize(deserializer)?;
+    if patterns.is_empty() {
+        return Ok(None);
+    }
+    let mut builder = globset::GlobSetBuilder::new();
+    for pattern in patterns {
+        let glob = globset::Glob::new(&pattern).map_err(serde::de::Error::custom)?;
+        builder.add(glob);
+    }
+    builder.build().map(Some).map_err(serde::de::Error::custom)
+}
+
+#[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "kebab-case")]
+pub struct LanguageServerConfiguration {
+    pub command: String,
+    #[serde(default)]
+    #[serde(skip_serializing_if = "Vec::is_empty")]
+    pub args: Vec<String>,
+    #[serde(default, skip_serializing_if = "HashMap::is_empty")]
+    pub environment: HashMap<String, String>,
+    #[serde(default, skip_serializing, deserialize_with = "deserialize_lsp_config")]
+    pub config: Option<serde_json::Value>,
+    #[serde(default = "default_timeout")]
+    pub timeout: u64,
+    #[serde(
+        default,
+        skip_serializing,
+        deserialize_with = "deserialize_required_root_patterns"
+    )]
+    pub required_root_patterns: Option<GlobSet>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+#[serde(rename_all = "kebab-case")]
+pub struct FormatterConfiguration {
+    pub command: String,
+    #[serde(default)]
+    #[serde(skip_serializing_if = "Vec::is_empty")]
+    pub args: Vec<String>,
+}
+
+#[derive(Debug, PartialEq, Eq, Clone, Deserialize, Serialize)]
+#[serde(rename_all = "kebab-case")]
+pub struct AdvancedCompletion {
+    pub name: Option<String>,
+    pub completion: Option<String>,
+    pub default: Option<String>,
+}
+
+#[derive(Debug, PartialEq, Eq, Clone, Deserialize, Serialize)]
+#[serde(rename_all = "kebab-case", untagged)]
+pub enum DebugConfigCompletion {
+    Named(String),
+    Advanced(AdvancedCompletion),
+}
+
+#[derive(Debug, PartialEq, Eq, Clone, Deserialize, Serialize)]
+#[serde(untagged)]
+pub enum DebugArgumentValue {
+    String(String),
+    Array(Vec<String>),
+    Boolean(bool),
+}
+
+#[derive(Debug, PartialEq, Eq, Clone, Deserialize, Serialize)]
+#[serde(rename_all = "kebab-case")]
+pub struct DebugTemplate {
+    pub name: String,
+    pub request: String,
+    #[serde(default)]
+    pub completion: Vec<DebugConfigCompletion>,
+    pub args: HashMap<String, DebugArgumentValue>,
+}
+
+#[derive(Debug, PartialEq, Eq, Clone, Deserialize, Serialize)]
+#[serde(rename_all = "kebab-case")]
+pub struct DebugAdapterConfig {
+    pub name: String,
+    pub transport: String,
+    #[serde(default)]
+    pub command: String,
+    #[serde(default)]
+    pub args: Vec<String>,
+    pub port_arg: Option<String>,
+    pub templates: Vec<DebugTemplate>,
+    #[serde(default)]
+    pub quirks: DebuggerQuirks,
+}
+
+// Different workarounds for adapters' differences
+#[derive(Debug, Default, PartialEq, Eq, Clone, Serialize, Deserialize)]
+pub struct DebuggerQuirks {
+    #[serde(default)]
+    pub absolute_paths: bool,
+}
+
+#[derive(Debug, Serialize, Deserialize)]
+#[serde(rename_all = "kebab-case")]
+pub struct IndentationConfiguration {
+    #[serde(deserialize_with = "deserialize_tab_width")]
+    pub tab_width: usize,
+    pub unit: String,
+}
+
+/// How the indentation for a newly inserted line should be determined.
+/// If the selected heuristic is not available (e.g. because the current
+/// language has no tree-sitter indent queries), a simpler one will be used.
+#[derive(Debug, Default, Clone, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "kebab-case")]
+pub enum IndentationHeuristic {
+    /// Just copy the indentation of the line that the cursor is currently on.
+    Simple,
+    /// Use tree-sitter indent queries to compute the expected absolute indentation level of the new line.
+    TreeSitter,
+    /// Use tree-sitter indent queries to compute the expected difference in indentation between the new line
+    /// and the line before. Add this to the actual indentation level of the line before.
+    #[default]
+    Hybrid,
+}
+
+/// Configuration for auto pairs
+#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "kebab-case", deny_unknown_fields, untagged)]
+pub enum AutoPairConfig {
+    /// Enables or disables auto pairing. False means disabled. True means to use the default pairs.
+    Enable(bool),
+
+    /// The mappings of pairs.
+    Pairs(HashMap<char, char>),
+}
+
+impl Default for AutoPairConfig {
+    fn default() -> Self {
+        AutoPairConfig::Enable(true)
+    }
+}
+
+impl From<&AutoPairConfig> for Option<AutoPairs> {
+    fn from(auto_pair_config: &AutoPairConfig) -> Self {
+        match auto_pair_config {
+            AutoPairConfig::Enable(false) => None,
+            AutoPairConfig::Enable(true) => Some(AutoPairs::default()),
+            AutoPairConfig::Pairs(pairs) => Some(AutoPairs::new(pairs.iter())),
+        }
+    }
+}
+
+impl From<AutoPairConfig> for Option<AutoPairs> {
+    fn from(auto_pairs_config: AutoPairConfig) -> Self {
+        (&auto_pairs_config).into()
+    }
+}
+
+impl FromStr for AutoPairConfig {
+    type Err = std::str::ParseBoolError;
+
+    // only do bool parsing for runtime setting
+    fn from_str(s: &str) -> Result<Self, Self::Err> {
+        let enable: bool = s.parse()?;
+        Ok(AutoPairConfig::Enable(enable))
+    }
+}
+
+#[derive(Debug, Clone, Default, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(default, rename_all = "kebab-case", deny_unknown_fields)]
+pub struct SoftWrap {
+    /// Soft wrap lines that exceed viewport width. Default to off
+    // NOTE: Option on purpose because the struct is shared between language config and global config.
+    // By default the option is None so that the language config falls back to the global config unless explicitly set.
+    pub enable: Option<bool>,
+    /// Maximum space left free at the end of the line.
+    /// This space is used to wrap text at word boundaries. If that is not possible within this limit
+    /// the word is simply split at the end of the line.
+    ///
+    /// This is automatically hard-limited to a quarter of the viewport to ensure correct display on small views.
+    ///
+    /// Default to 20
+    pub max_wrap: Option<u16>,
+    /// Maximum number of indentation that can be carried over from the previous line when softwrapping.
+    /// If a line is indented further then this limit it is rendered at the start of the viewport instead.
+    ///
+    /// This is automatically hard-limited to a quarter of the viewport to ensure correct display on small views.
+    ///
+    /// Default to 40
+    pub max_indent_retain: Option<u16>,
+    /// Indicator placed at the beginning of softwrapped lines
+    ///
+    /// Defaults to ↪
+    pub wrap_indicator: Option<String>,
+    /// Softwrap at `text_width` instead of viewport width if it is shorter
+    pub wrap_at_text_width: Option<bool>,
+}
+
+#[derive(Debug, Clone, Copy, Default, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(default, rename_all = "kebab-case", deny_unknown_fields)]
+pub struct WordCompletion {
+    pub enable: Option<bool>,
+    pub trigger_length: Option<NonZeroU8>,
+}
+
+fn deserialize_regex<'de, D>(deserializer: D) -> Result<Option<rope::Regex>, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    Option::<String>::deserialize(deserializer)?
+        .map(|buf| rope::Regex::new(&buf).map_err(serde::de::Error::custom))
+        .transpose()
+}
+
+fn deserialize_lsp_config<'de, D>(deserializer: D) -> Result<Option<serde_json::Value>, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    Option::<toml::Value>::deserialize(deserializer)?
+        .map(|toml| toml.try_into().map_err(serde::de::Error::custom))
+        .transpose()
+}
+
+fn deserialize_tab_width<'de, D>(deserializer: D) -> Result<usize, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    usize::deserialize(deserializer).and_then(|n| {
+        if n > 0 && n <= 16 {
+            Ok(n)
+        } else {
+            Err(serde::de::Error::custom(
+                "tab width must be a value from 1 to 16 inclusive",
+            ))
+        }
+    })
+}
+
+pub fn deserialize_auto_pairs<'de, D>(deserializer: D) -> Result<Option<AutoPairs>, D::Error>
+where
+    D: serde::Deserializer<'de>,
+{
+    Ok(Option::<AutoPairConfig>::deserialize(deserializer)?.and_then(AutoPairConfig::into))
+}
+
+fn default_timeout() -> u64 {
+    20
+}

helix-core/src/syntax/tree_cursor.rs

@@ -1,264 +0,0 @@
-use std::{cmp::Reverse, ops::Range};
-
-use super::{LanguageLayer, LayerId};
-
-use slotmap::HopSlotMap;
-use tree_sitter::Node;
-
-/// The byte range of an injection layer.
-///
-/// Injection ranges may overlap, but all overlapping parts are subsets of their parent ranges.
-/// This allows us to sort the ranges ahead of time in order to efficiently find a range that
-/// contains a point with maximum depth.
-#[derive(Debug)]
-struct InjectionRange {
-    start: usize,
-    end: usize,
-    layer_id: LayerId,
-    depth: u32,
-}
-
-pub struct TreeCursor<'a> {
-    layers: &'a HopSlotMap<LayerId, LanguageLayer>,
-    root: LayerId,
-    current: LayerId,
-    injection_ranges: Vec<InjectionRange>,
-    // TODO: Ideally this would be a `tree_sitter::TreeCursor<'a>` but
-    // that returns very surprising results in testing.
-    cursor: Node<'a>,
-}
-
-impl<'a> TreeCursor<'a> {
-    pub(super) fn new(layers: &'a HopSlotMap<LayerId, LanguageLayer>, root: LayerId) -> Self {
-        let mut injection_ranges = Vec::new();
-
-        for (layer_id, layer) in layers.iter() {
-            // Skip the root layer
-            if layer.parent.is_none() {
-                continue;
-            }
-            for byte_range in layer.ranges.iter() {
-                let range = InjectionRange {
-                    start: byte_range.start_byte,
-                    end: byte_range.end_byte,
-                    layer_id,
-                    depth: layer.depth,
-                };
-                injection_ranges.push(range);
-            }
-        }
-
-        injection_ranges.sort_unstable_by_key(|range| (range.end, Reverse(range.depth)));
-
-        let cursor = layers[root].tree().root_node();
-
-        Self {
-            layers,
-            root,
-            current: root,
-            injection_ranges,
-            cursor,
-        }
-    }
-
-    pub fn node(&self) -> Node<'a> {
-        self.cursor
-    }
-
-    pub fn goto_parent(&mut self) -> bool {
-        if let Some(parent) = self.node().parent() {
-            self.cursor = parent;
-            return true;
-        }
-
-        // If we are already on the root layer, we cannot ascend.
-        if self.current == self.root {
-            return false;
-        }
-
-        // Ascend to the parent layer.
-        let range = self.node().byte_range();
-        let parent_id = self.layers[self.current]
-            .parent
-            .expect("non-root layers have a parent");
-        self.current = parent_id;
-        let root = self.layers[self.current].tree().root_node();
-        self.cursor = root
-            .descendant_for_byte_range(range.start, range.end)
-            .unwrap_or(root);
-
-        true
-    }
-
-    pub fn goto_parent_with<P>(&mut self, predicate: P) -> bool
-    where
-        P: Fn(&Node) -> bool,
-    {
-        while self.goto_parent() {
-            if predicate(&self.node()) {
-                return true;
-            }
-        }
-
-        false
-    }
-
-    /// Finds the injection layer that has exactly the same range as the given `range`.
-    fn layer_id_of_byte_range(&self, search_range: Range<usize>) -> Option<LayerId> {
-        let start_idx = self
-            .injection_ranges
-            .partition_point(|range| range.end < search_range.end);
-
-        self.injection_ranges[start_idx..]
-            .iter()
-            .take_while(|range| range.end == search_range.end)
-            .find_map(|range| (range.start == search_range.start).then_some(range.layer_id))
-    }
-
-    fn goto_first_child_impl(&mut self, named: bool) -> bool {
-        // Check if the current node's range is an exact injection layer range.
-        if let Some(layer_id) = self
-            .layer_id_of_byte_range(self.node().byte_range())
-            .filter(|&layer_id| layer_id != self.current)
-        {
-            // Switch to the child layer.
-            self.current = layer_id;
-            self.cursor = self.layers[self.current].tree().root_node();
-            return true;
-        }
-
-        let child = if named {
-            self.cursor.named_child(0)
-        } else {
-            self.cursor.child(0)
-        };
-
-        if let Some(child) = child {
-            // Otherwise descend in the current tree.
-            self.cursor = child;
-            true
-        } else {
-            false
-        }
-    }
-
-    pub fn goto_first_child(&mut self) -> bool {
-        self.goto_first_child_impl(false)
-    }
-
-    pub fn goto_first_named_child(&mut self) -> bool {
-        self.goto_first_child_impl(true)
-    }
-
-    fn goto_next_sibling_impl(&mut self, named: bool) -> bool {
-        let sibling = if named {
-            self.cursor.next_named_sibling()
-        } else {
-            self.cursor.next_sibling()
-        };
-
-        if let Some(sibling) = sibling {
-            self.cursor = sibling;
-            true
-        } else {
-            false
-        }
-    }
-
-    pub fn goto_next_sibling(&mut self) -> bool {
-        self.goto_next_sibling_impl(false)
-    }
-
-    pub fn goto_next_named_sibling(&mut self) -> bool {
-        self.goto_next_sibling_impl(true)
-    }
-
-    fn goto_prev_sibling_impl(&mut self, named: bool) -> bool {
-        let sibling = if named {
-            self.cursor.prev_named_sibling()
-        } else {
-            self.cursor.prev_sibling()
-        };
-
-        if let Some(sibling) = sibling {
-            self.cursor = sibling;
-            true
-        } else {
-            false
-        }
-    }
-
-    pub fn goto_prev_sibling(&mut self) -> bool {
-        self.goto_prev_sibling_impl(false)
-    }
-
-    pub fn goto_prev_named_sibling(&mut self) -> bool {
-        self.goto_prev_sibling_impl(true)
-    }
-
-    /// Finds the injection layer that contains the given start-end range.
-    fn layer_id_containing_byte_range(&self, start: usize, end: usize) -> LayerId {
-        let start_idx = self
-            .injection_ranges
-            .partition_point(|range| range.end < end);
-
-        self.injection_ranges[start_idx..]
-            .iter()
-            .take_while(|range| range.start < end)
-            .find_map(|range| (range.start <= start).then_some(range.layer_id))
-            .unwrap_or(self.root)
-    }
-
-    pub fn reset_to_byte_range(&mut self, start: usize, end: usize) {
-        self.current = self.layer_id_containing_byte_range(start, end);
-        let root = self.layers[self.current].tree().root_node();
-        self.cursor = root.descendant_for_byte_range(start, end).unwrap_or(root);
-    }
-
-    /// Returns an iterator over the children of the node the TreeCursor is on
-    /// at the time this is called.
-    pub fn children(&'a mut self) -> ChildIter {
-        let parent = self.node();
-
-        ChildIter {
-            cursor: self,
-            parent,
-            named: false,
-        }
-    }
-
-    /// Returns an iterator over the named children of the node the TreeCursor is on
-    /// at the time this is called.
-    pub fn named_children(&'a mut self) -> ChildIter {
-        let parent = self.node();
-
-        ChildIter {
-            cursor: self,
-            parent,
-            named: true,
-        }
-    }
-}
-
-pub struct ChildIter<'n> {
-    cursor: &'n mut TreeCursor<'n>,
-    parent: Node<'n>,
-    named: bool,
-}
-
-impl<'n> Iterator for ChildIter<'n> {
-    type Item = Node<'n>;
-
-    fn next(&mut self) -> Option<Self::Item> {
-        // first iteration, just visit the first child
-        if self.cursor.node() == self.parent {
-            self.cursor
-                .goto_first_child_impl(self.named)
-                .then(|| self.cursor.node())
-        } else {
-            self.cursor
-                .goto_next_sibling_impl(self.named)
-                .then(|| self.cursor.node())
-        }
-    }
-}

helix-core/src/test.rs

@@ -65,7 +65,7 @@ pub fn print(s: &str) -> (String, Selection) {
         let head_at_beg = iter.next_if_eq(&"|").is_some();
         let last_grapheme = |s: &str| {
             UnicodeSegmentation::graphemes(s, true)
-                .last()
+                .next_back()
                 .map(String::from)
         };
 

helix-core/src/text_annotations.rs

@@ -1,8 +1,12 @@
 use std::cell::Cell;
+use std::cmp::Ordering;
+use std::fmt::Debug;
 use std::ops::Range;
+use std::ptr::NonNull;
 
-use crate::syntax::Highlight;
-use crate::Tendril;
+use crate::doc_formatter::FormattedGrapheme;
+use crate::syntax::{Highlight, OverlayHighlights};
+use crate::{Position, Tendril};
 
 /// An inline annotation is continuous text shown
 /// on the screen before the grapheme that starts at
@@ -75,19 +79,98 @@ impl Overlay {
     }
 }
 
-/// Line annotations allow for virtual text between normal
-/// text lines. They cause `height` empty lines to be inserted
-/// below the document line that contains `anchor_char_idx`.
+/// Line annotations allow inserting virtual text lines between normal text
+/// lines.  These lines can be filled with text in the rendering code as their
+/// contents have no effect beyond visual appearance.
 ///
-/// These lines can be filled with text in the rendering code
-/// as their contents have no effect beyond visual appearance.
+/// The height of virtual text is usually not known ahead of time as virtual
+/// text often requires softwrapping. Furthermore the height of some virtual
+/// text like side-by-side diffs depends on the height of the text (again
+/// influenced by softwrap) and other virtual text. Therefore line annotations
+/// are computed on the fly instead of ahead of time like other annotations.
 ///
-/// To insert a line after a document line simply set
-/// `anchor_char_idx` to `doc.line_to_char(line_idx)`
-#[derive(Debug, Clone)]
-pub struct LineAnnotation {
-    pub anchor_char_idx: usize,
-    pub height: usize,
+/// The core of this trait `insert_virtual_lines` function. It is called at the
+/// end of every  visual line and allows the `LineAnnotation` to insert empty
+/// virtual lines. Apart from that the `LineAnnotation` trait has multiple
+/// methods that allow it to track anchors in the document.
+///
+/// When a new traversal of a document starts `reset_pos` is called. Afterwards
+/// the other functions are called with indices that are larger then the
+/// one passed to `reset_pos`. This allows performing a binary search (use
+/// `partition_point`) in `reset_pos` once and then to only look at the next
+/// anchor during each method call.
+///
+/// The `reset_pos`, `skip_conceal` and `process_anchor` functions all return a
+/// `char_idx` anchor. This anchor is stored when transversing the document and
+/// when the grapheme at the anchor is traversed the `process_anchor` function
+/// is called.
+///
+/// # Note
+///
+/// All functions only receive immutable references to `self`.
+/// `LineAnnotation`s that want to store an internal position or
+/// state of some kind should use `Cell`. Using interior mutability for
+/// caches is preferable as otherwise a lot of lifetimes become invariant
+/// which complicates APIs a lot.
+pub trait LineAnnotation {
+    /// Resets the internal position to `char_idx`. This function is called
+    /// when a new traversal of a document starts.
+    ///
+    /// All `char_idx` passed to `insert_virtual_lines` are strictly monotonically increasing
+    /// with the first `char_idx` greater or equal to the `char_idx`
+    /// passed to this function.
+    ///
+    /// # Returns
+    ///
+    /// The `char_idx` of the next anchor this `LineAnnotation` is interested in,
+    /// replaces the currently registered anchor. Return `usize::MAX` to ignore
+    fn reset_pos(&mut self, _char_idx: usize) -> usize {
+        usize::MAX
+    }
+
+    /// Called when a text is concealed that contains an anchor registered by this `LineAnnotation`.
+    /// In this case the line decorations  **must** ensure that virtual text anchored within that
+    /// char range is skipped.
+    ///
+    /// # Returns
+    ///
+    /// The `char_idx` of the next anchor this `LineAnnotation` is interested in,
+    /// **after the end of conceal_end_char_idx**
+    /// replaces the currently registered anchor. Return `usize::MAX` to ignore
+    fn skip_concealed_anchors(&mut self, conceal_end_char_idx: usize) -> usize {
+        self.reset_pos(conceal_end_char_idx)
+    }
+
+    /// Process an anchor (horizontal position is provided) and returns the next anchor.
+    ///
+    /// # Returns
+    ///
+    /// The `char_idx` of the next anchor this `LineAnnotation` is interested in,
+    /// replaces the currently registered anchor. Return `usize::MAX` to ignore
+    fn process_anchor(&mut self, _grapheme: &FormattedGrapheme) -> usize {
+        usize::MAX
+    }
+
+    /// This function is called at the end of a visual line to insert virtual text
+    ///
+    /// # Returns
+    ///
+    /// The number of additional virtual lines to reserve
+    ///
+    /// # Note
+    ///
+    /// The `line_end_visual_pos` parameter indicates the visual vertical distance
+    /// from the start of block where the traversal starts.  This includes the offset
+    /// from other `LineAnnotations`. This allows inline annotations to consider
+    /// the height of the text and "align" two different documents (like for side
+    /// by side diffs).  These annotations that want to "align" two documents should
+    /// therefore be added last so that other virtual text is also considered while aligning
+    fn insert_virtual_lines(
+        &mut self,
+        line_end_char_idx: usize,
+        line_end_visual_pos: Position,
+        doc_line: usize,
+    ) -> Position;
 }
 
 #[derive(Debug)]
@@ -128,7 +211,7 @@ impl<A, M> Layer<'_, A, M> {
 }
 
 impl<'a, A, M> From<(&'a [A], M)> for Layer<'a, A, M> {
-    fn from((annotations, metadata): (&'a [A], M)) -> Layer<A, M> {
+    fn from((annotations, metadata): (&'a [A], M)) -> Layer<'a, A, M> {
         Layer {
             annotations,
             current_index: Cell::new(0),
@@ -143,13 +226,68 @@ fn reset_pos<A, M>(layers: &[Layer<A, M>], pos: usize, get_pos: impl Fn(&A) -> u
     }
 }
 
+/// Safety: We store LineAnnotation in a NonNull pointer. This is necessary to work
+/// around an unfortunate inconsistency in rusts variance system that unnnecesarily
+/// makes the lifetime invariant if implemented with safe code. This makes the
+/// DocFormatter API very cumbersome/basically impossible to work with.
+///
+/// Normally object types `dyn Foo + 'a` are covariant so if we used `Box<dyn LineAnnotation + 'a>` below
+/// everything would be alright. However we want to use `Cell<Box<dyn LineAnnotation + 'a>>`
+/// to be able to call the mutable function on `LineAnnotation`. The problem is that
+/// some types like `Cell` make all their arguments invariant. This is important for soundness
+/// normally for the same reasons that `&'a mut T` is invariant over `T`
+/// (see <https://doc.rust-lang.org/nomicon/subtyping.html>). However for `&'a mut` (`dyn Foo + 'b`)
+/// there is a specical rule in the language to make `'b` covariant (otherwise trait objects would be
+/// super annoying to use). See  <https://users.rust-lang.org/t/solved-variance-of-dyn-trait-a> for
+/// why this is sound. Sadly that rule doesn't apply to `Cell<Box<(dyn Foo + 'a)>`
+/// (or other invariant types like `UnsafeCell` or `*mut (dyn Foo + 'a)`).
+///
+/// We sidestep the problem by using `NonNull` which is covariant. In the
+/// special case of trait objects this is sound (easily checked by adding a
+/// `PhantomData<&'a mut Foo + 'a)>` field). We don't need an explicit `Cell`
+/// type here because we never hand out any refereces to the trait objects. That
+/// means any reference to the pointer can create a valid multable reference
+/// that is covariant over `'a` (or in other words it's a raw pointer, as long as
+/// we don't hand out references we are free to do whatever we want).
+struct RawBox<T: ?Sized>(NonNull<T>);
+
+impl<T: ?Sized> RawBox<T> {
+    /// Safety: Only a single mutable reference
+    /// created by this function may exist at a given time.
+    #[allow(clippy::mut_from_ref)]
+    unsafe fn get(&self) -> &mut T {
+        &mut *self.0.as_ptr()
+    }
+}
+impl<T: ?Sized> From<Box<T>> for RawBox<T> {
+    fn from(box_: Box<T>) -> Self {
+        // obviously safe because Box::into_raw never returns null
+        unsafe { Self(NonNull::new_unchecked(Box::into_raw(box_))) }
+    }
+}
+
+impl<T: ?Sized> Drop for RawBox<T> {
+    fn drop(&mut self) {
+        unsafe { drop(Box::from_raw(self.0.as_ptr())) }
+    }
+}
+
 /// Annotations that change that is displayed when the document is render.
 /// Also commonly called virtual text.
-#[derive(Default, Debug, Clone)]
+#[derive(Default)]
 pub struct TextAnnotations<'a> {
     inline_annotations: Vec<Layer<'a, InlineAnnotation, Option<Highlight>>>,
     overlays: Vec<Layer<'a, Overlay, Option<Highlight>>>,
-    line_annotations: Vec<Layer<'a, LineAnnotation, ()>>,
+    line_annotations: Vec<(Cell<usize>, RawBox<dyn LineAnnotation + 'a>)>,
+}
+
+impl Debug for TextAnnotations<'_> {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        f.debug_struct("TextAnnotations")
+            .field("inline_annotations", &self.inline_annotations)
+            .field("overlays", &self.overlays)
+            .finish_non_exhaustive()
+    }
 }
 
 impl<'a> TextAnnotations<'a> {
@@ -157,15 +295,12 @@ impl<'a> TextAnnotations<'a> {
     pub fn reset_pos(&self, char_idx: usize) {
         reset_pos(&self.inline_annotations, char_idx, |annot| annot.char_idx);
         reset_pos(&self.overlays, char_idx, |annot| annot.char_idx);
-        reset_pos(&self.line_annotations, char_idx, |annot| {
-            annot.anchor_char_idx
-        });
+        for (next_anchor, layer) in &self.line_annotations {
+            next_anchor.set(unsafe { layer.get().reset_pos(char_idx) });
+        }
     }
 
-    pub fn collect_overlay_highlights(
-        &self,
-        char_range: Range<usize>,
-    ) -> Vec<(usize, Range<usize>)> {
+    pub fn collect_overlay_highlights(&self, char_range: Range<usize>) -> OverlayHighlights {
         let mut highlights = Vec::new();
         self.reset_pos(char_range.start);
         for char_idx in char_range {
@@ -173,11 +308,11 @@ impl<'a> TextAnnotations<'a> {
                 // we don't know the number of chars the original grapheme takes
                 // however it doesn't matter as highlight boundaries are automatically
                 // aligned to grapheme boundaries in the rendering code
-                highlights.push((highlight.0, char_idx..char_idx + 1))
+                highlights.push((highlight, char_idx..char_idx + 1));
             }
         }
 
-        highlights
+        OverlayHighlights::Heterogenous { highlights }
     }
 
     /// Add new inline annotations.
@@ -196,7 +331,9 @@ impl<'a> TextAnnotations<'a> {
         layer: &'a [InlineAnnotation],
         highlight: Option<Highlight>,
     ) -> &mut Self {
-        self.inline_annotations.push((layer, highlight).into());
+        if !layer.is_empty() {
+            self.inline_annotations.push((layer, highlight).into());
+        }
         self
     }
 
@@ -211,7 +348,9 @@ impl<'a> TextAnnotations<'a> {
     /// If multiple layers contain overlay at the same position
     /// the overlay from the layer added last will be show.
     pub fn add_overlay(&mut self, layer: &'a [Overlay], highlight: Option<Highlight>) -> &mut Self {
-        self.overlays.push((layer, highlight).into());
+        if !layer.is_empty() {
+            self.overlays.push((layer, highlight).into());
+        }
         self
     }
 
@@ -219,8 +358,9 @@ impl<'a> TextAnnotations<'a> {
     ///
     /// The line annotations **must be sorted** by their `char_idx`.
     /// Multiple line annotations with the same `char_idx` **are not allowed**.
-    pub fn add_line_annotation(&mut self, layer: &'a [LineAnnotation]) -> &mut Self {
-        self.line_annotations.push((layer, ()).into());
+    pub fn add_line_annotation(&mut self, layer: Box<dyn LineAnnotation + 'a>) -> &mut Self {
+        self.line_annotations
+            .push((Cell::new(usize::MAX), layer.into()));
         self
     }
 
@@ -250,21 +390,35 @@ impl<'a> TextAnnotations<'a> {
         overlay
     }
 
-    pub(crate) fn annotation_lines_at(&self, char_idx: usize) -> usize {
-        self.line_annotations
-            .iter()
-            .map(|layer| {
-                let mut lines = 0;
-                while let Some(annot) = layer.annotations.get(layer.current_index.get()) {
-                    if annot.anchor_char_idx == char_idx {
-                        layer.current_index.set(layer.current_index.get() + 1);
-                        lines += annot.height
-                    } else {
-                        break;
+    pub(crate) fn process_virtual_text_anchors(&self, grapheme: &FormattedGrapheme) {
+        for (next_anchor, layer) in &self.line_annotations {
+            loop {
+                match next_anchor.get().cmp(&grapheme.char_idx) {
+                    Ordering::Less => next_anchor
+                        .set(unsafe { layer.get().skip_concealed_anchors(grapheme.char_idx) }),
+                    Ordering::Equal => {
+                        next_anchor.set(unsafe { layer.get().process_anchor(grapheme) })
                     }
-                }
-                lines
-            })
-            .sum()
+                    Ordering::Greater => break,
+                };
+            }
+        }
+    }
+
+    pub(crate) fn virtual_lines_at(
+        &self,
+        char_idx: usize,
+        line_end_visual_pos: Position,
+        doc_line: usize,
+    ) -> usize {
+        let mut virt_off = Position::new(0, 0);
+        for (_, layer) in &self.line_annotations {
+            virt_off += unsafe {
+                layer
+                    .get()
+                    .insert_virtual_lines(char_idx, line_end_visual_pos + virt_off, doc_line)
+            };
+        }
+        virt_off.row
     }
 }

helix-core/src/textobject.rs

@@ -1,13 +1,12 @@
 use std::fmt::Display;
 
 use ropey::RopeSlice;
-use tree_sitter::{Node, QueryCursor};
 
 use crate::chars::{categorize_char, char_is_whitespace, CharCategory};
 use crate::graphemes::{next_grapheme_boundary, prev_grapheme_boundary};
 use crate::line_ending::rope_is_line_ending;
 use crate::movement::Direction;
-use crate::syntax::LanguageConfiguration;
+use crate::syntax;
 use crate::Range;
 use crate::{surround, Syntax};
 
@@ -260,18 +259,18 @@ pub fn textobject_treesitter(
     range: Range,
     textobject: TextObject,
     object_name: &str,
-    slice_tree: Node,
-    lang_config: &LanguageConfiguration,
+    syntax: &Syntax,
+    loader: &syntax::Loader,
     _count: usize,
 ) -> Range {
+    let root = syntax.tree().root_node();
+    let textobject_query = loader.textobject_query(syntax.root_language());
     let get_range = move || -> Option<Range> {
         let byte_pos = slice.char_to_byte(range.cursor(slice));
 
         let capture_name = format!("{}.{}", object_name, textobject); // eg. function.inner
-        let mut cursor = QueryCursor::new();
-        let node = lang_config
-            .textobject_query()?
-            .capture_nodes(&capture_name, slice_tree, slice, &mut cursor)?
+        let node = textobject_query?
+            .capture_nodes(&capture_name, &root, slice)?
             .filter(|node| node.byte_range().contains(&byte_pos))
             .min_by_key(|node| node.byte_range().len())?;
 

helix-core/src/transaction.rs

@@ -19,6 +19,16 @@ pub enum Operation {
     Insert(Tendril),
 }
 
+impl Operation {
+    /// The number of characters affected by the operation.
+    pub fn len_chars(&self) -> usize {
+        match self {
+            Self::Retain(n) | Self::Delete(n) => *n,
+            Self::Insert(s) => s.chars().count(),
+        }
+    }
+}
+
 #[derive(Debug, Copy, Clone, PartialEq, Eq)]
 pub enum Assoc {
     Before,
@@ -29,6 +39,12 @@ pub enum Assoc {
     /// Acts like `Before` if a word character is inserted
     /// before the position, otherwise acts like `After`
     BeforeWord,
+    /// Acts like `Before` but if the position is within an exact replacement
+    /// (exact size) the offset to the start of the replacement is kept
+    BeforeSticky,
+    /// Acts like `After` but if the position is within an exact replacement
+    /// (exact size) the offset to the start of the replacement is kept
+    AfterSticky,
 }
 
 impl Assoc {
@@ -40,13 +56,17 @@ impl Assoc {
     fn insert_offset(self, s: &str) -> usize {
         let chars = s.chars().count();
         match self {
-            Assoc::After => chars,
+            Assoc::After | Assoc::AfterSticky => chars,
             Assoc::AfterWord => s.chars().take_while(|&c| char_is_word(c)).count(),
             // return position before inserted text
-            Assoc::Before => 0,
+            Assoc::Before | Assoc::BeforeSticky => 0,
             Assoc::BeforeWord => chars - s.chars().rev().take_while(|&c| char_is_word(c)).count(),
         }
     }
+
+    pub fn sticky(self) -> bool {
+        matches!(self, Assoc::BeforeSticky | Assoc::AfterSticky)
+    }
 }
 
 #[derive(Debug, Default, Clone, PartialEq, Eq)]
@@ -456,8 +476,14 @@ impl ChangeSet {
                                 if pos == old_pos && assoc.stay_at_gaps() {
                                     new_pos
                                 } else {
-                                    // place to end of insert
-                                    new_pos + assoc.insert_offset(s)
+                                    let ins = assoc.insert_offset(s);
+                                    // if the deleted and inserted text have the exact same size
+                                    // keep the relative offset into the new text
+                                    if *len == ins && assoc.sticky() {
+                                        new_pos + (pos - old_pos)
+                                    } else {
+                                        new_pos + assoc.insert_offset(s)
+                                    }
                                 }
                             }),
                             i
@@ -753,7 +779,7 @@ impl<'a> ChangeIterator<'a> {
     }
 }
 
-impl<'a> Iterator for ChangeIterator<'a> {
+impl Iterator for ChangeIterator<'_> {
     type Item = Change;
 
     fn next(&mut self) -> Option<Self::Item> {

helix-core/src/uri.rs

@@ -1,12 +1,18 @@
-use std::path::{Path, PathBuf};
+use std::{
+    fmt,
+    path::{Path, PathBuf},
+    sync::Arc,
+};
 
 /// A generic pointer to a file location.
 ///
 /// Currently this type only supports paths to local files.
+///
+/// Cloning this type is cheap: the internal representation uses an Arc.
 #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord)]
 #[non_exhaustive]
 pub enum Uri {
-    File(PathBuf),
+    File(Arc<Path>),
 }
 
 impl Uri {
@@ -23,26 +29,18 @@ impl Uri {
             Self::File(path) => Some(path),
         }
     }
-
-    pub fn as_path_buf(self) -> Option<PathBuf> {
-        match self {
-            Self::File(path) => Some(path),
-        }
-    }
 }
 
 impl From<PathBuf> for Uri {
     fn from(path: PathBuf) -> Self {
-        Self::File(path)
+        Self::File(path.into())
     }
 }
 
-impl TryFrom<Uri> for PathBuf {
-    type Error = ();
-
-    fn try_from(uri: Uri) -> Result<Self, Self::Error> {
-        match uri {
-            Uri::File(path) => Ok(path),
+impl fmt::Display for Uri {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            Self::File(path) => write!(f, "{}", path.display()),
         }
     }
 }
@@ -59,11 +57,16 @@ pub enum UrlConversionErrorKind {
     UnableToConvert,
 }
 
-impl std::fmt::Display for UrlConversionError {
-    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+impl fmt::Display for UrlConversionError {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         match self.kind {
             UrlConversionErrorKind::UnsupportedScheme => {
-                write!(f, "unsupported scheme in URL: {}", self.source.scheme())
+                write!(
+                    f,
+                    "unsupported scheme '{}' in URL {}",
+                    self.source.scheme(),
+                    self.source
+                )
             }
             UrlConversionErrorKind::UnableToConvert => {
                 write!(f, "unable to convert URL to file path: {}", self.source)
@@ -77,7 +80,7 @@ impl std::error::Error for UrlConversionError {}
 fn convert_url_to_uri(url: &url::Url) -> Result<Uri, UrlConversionErrorKind> {
     if url.scheme() == "file" {
         url.to_file_path()
-            .map(|path| Uri::File(helix_stdx::path::normalize(path)))
+            .map(|path| Uri::File(helix_stdx::path::normalize(path).into()))
             .map_err(|_| UrlConversionErrorKind::UnableToConvert)
     } else {
         Err(UrlConversionErrorKind::UnsupportedScheme)

helix-core/src/wrap.rs

@@ -4,6 +4,8 @@ use textwrap::{Options, WordSplitter::NoHyphenation};
 /// Given a slice of text, return the text re-wrapped to fit it
 /// within the given width.
 pub fn reflow_hard_wrap(text: &str, text_width: usize) -> SmartString<LazyCompact> {
-    let options = Options::new(text_width).word_splitter(NoHyphenation);
+    let options = Options::new(text_width)
+        .word_splitter(NoHyphenation)
+        .word_separator(textwrap::WordSeparator::AsciiSpace);
     textwrap::refill(text, options).into()
 }

helix-core/tests/indent.rs

@@ -1,12 +1,11 @@
-use arc_swap::ArcSwap;
 use helix_core::{
     indent::{indent_level_for_line, treesitter_indent_for_pos, IndentStyle},
-    syntax::{Configuration, Loader},
+    syntax::{config::Configuration, Loader},
     Syntax,
 };
 use helix_stdx::rope::RopeSliceExt;
 use ropey::Rope;
-use std::{ops::Range, path::PathBuf, process::Command, sync::Arc};
+use std::{ops::Range, path::PathBuf, process::Command};
 
 #[test]
 fn test_treesitter_indent_rust() {
@@ -196,17 +195,12 @@ fn test_treesitter_indent(
     runtime.push("../runtime");
     std::env::set_var("HELIX_RUNTIME", runtime.to_str().unwrap());
 
-    let language_config = loader.language_config_for_scope(lang_scope).unwrap();
+    let language = loader.language_for_scope(lang_scope).unwrap();
+    let language_config = loader.language(language).config();
     let indent_style = IndentStyle::from_str(&language_config.indent.as_ref().unwrap().unit);
-    let highlight_config = language_config.highlight_config(&[]).unwrap();
     let text = doc.slice(..);
-    let syntax = Syntax::new(
-        text,
-        highlight_config,
-        Arc::new(ArcSwap::from_pointee(loader)),
-    )
-    .unwrap();
-    let indent_query = language_config.indent_query().unwrap();
+    let syntax = Syntax::new(text, language, &loader).unwrap();
+    let indent_query = loader.indent_query(language).unwrap();
 
     for i in 0..doc.len_lines() {
         let line = text.line(i);

helix-dap/Cargo.toml

@@ -22,6 +22,11 @@ serde = { version = "1.0", features = ["derive"] }
 serde_json = "1.0"
 tokio = { version = "1", features = ["rt", "rt-multi-thread", "io-util", "io-std", "time", "process", "macros", "fs", "parking_lot", "net", "sync"] }
 thiserror.workspace = true
+slotmap.workspace = true
+futures-executor.workspace = true
+futures-util.workspace = true
+tokio-stream.workspace = true
+
 
 [dev-dependencies]
-fern = "0.6"
+fern = "0.7"

helix-dap/src/client.rs

@@ -1,10 +1,11 @@
 use crate::{
-    requests::DisconnectArguments,
+    registry::DebugAdapterId,
+    requests::{DisconnectArguments, TerminateArguments},
     transport::{Payload, Request, Response, Transport},
     types::*,
     Error, Result,
 };
-use helix_core::syntax::DebuggerQuirks;
+use helix_core::syntax::config::{DebugAdapterConfig, DebuggerQuirks};
 
 use serde_json::Value;
 
@@ -27,12 +28,14 @@ use tokio::{
 
 #[derive(Debug)]
 pub struct Client {
-    id: usize,
+    id: DebugAdapterId,
     _process: Option<Child>,
     server_tx: UnboundedSender<Payload>,
     request_counter: AtomicU64,
     connection_type: Option<ConnectionType>,
     starting_request_args: Option<Value>,
+    /// The socket address of the debugger, if using TCP transport.
+    pub socket: Option<SocketAddr>,
     pub caps: Option<DebuggerCapabilities>,
     // thread_id -> frames
     pub stack_frames: HashMap<ThreadId, Vec<StackFrame>>,
@@ -41,23 +44,20 @@ pub struct Client {
     /// Currently active frame for the current thread.
     pub active_frame: Option<usize>,
     pub quirks: DebuggerQuirks,
-}
-
-#[derive(Clone, Copy, Debug)]
-pub enum ConnectionType {
-    Launch,
-    Attach,
+    /// The config which was used to start this debugger.
+    pub config: Option<DebugAdapterConfig>,
 }
 
 impl Client {
     // Spawn a process and communicate with it by either TCP or stdio
+    // The returned stream includes the Client ID so consumers can differentiate between multiple clients
     pub async fn process(
         transport: &str,
         command: &str,
         args: Vec<&str>,
         port_arg: Option<&str>,
-        id: usize,
-    ) -> Result<(Self, UnboundedReceiver<Payload>)> {
+        id: DebugAdapterId,
+    ) -> Result<(Self, UnboundedReceiver<(DebugAdapterId, Payload)>)> {
         if command.is_empty() {
             return Result::Err(Error::Other(anyhow!("Command not provided")));
         }
@@ -72,9 +72,9 @@ impl Client {
         rx: Box<dyn AsyncBufRead + Unpin + Send>,
         tx: Box<dyn AsyncWrite + Unpin + Send>,
         err: Option<Box<dyn AsyncBufRead + Unpin + Send>>,
-        id: usize,
+        id: DebugAdapterId,
         process: Option<Child>,
-    ) -> Result<(Self, UnboundedReceiver<Payload>)> {
+    ) -> Result<(Self, UnboundedReceiver<(DebugAdapterId, Payload)>)> {
         let (server_rx, server_tx) = Transport::start(rx, tx, err, id);
         let (client_tx, client_rx) = unbounded_channel();
 
@@ -86,22 +86,24 @@ impl Client {
             caps: None,
             connection_type: None,
             starting_request_args: None,
+            socket: None,
             stack_frames: HashMap::new(),
             thread_states: HashMap::new(),
             thread_id: None,
             active_frame: None,
             quirks: DebuggerQuirks::default(),
+            config: None,
         };
 
-        tokio::spawn(Self::recv(server_rx, client_tx));
+        tokio::spawn(Self::recv(id, server_rx, client_tx));
 
         Ok((client, client_rx))
     }
 
     pub async fn tcp(
         addr: std::net::SocketAddr,
-        id: usize,
-    ) -> Result<(Self, UnboundedReceiver<Payload>)> {
+        id: DebugAdapterId,
+    ) -> Result<(Self, UnboundedReceiver<(DebugAdapterId, Payload)>)> {
         let stream = TcpStream::connect(addr).await?;
         let (rx, tx) = stream.into_split();
         Self::streams(Box::new(BufReader::new(rx)), Box::new(tx), None, id, None)
@@ -110,8 +112,8 @@ impl Client {
     pub fn stdio(
         cmd: &str,
         args: Vec<&str>,
-        id: usize,
-    ) -> Result<(Self, UnboundedReceiver<Payload>)> {
+        id: DebugAdapterId,
+    ) -> Result<(Self, UnboundedReceiver<(DebugAdapterId, Payload)>)> {
         // Resolve path to the binary
         let cmd = helix_stdx::env::which(cmd)?;
 
@@ -119,6 +121,7 @@ impl Client {
             .args(args)
             .stdin(Stdio::piped())
             .stdout(Stdio::piped())
+            .stderr(Stdio::piped())
             // make sure the process is reaped on drop
             .kill_on_drop(true)
             .spawn();
@@ -128,16 +131,12 @@ impl Client {
         // TODO: do we need bufreader/writer here? or do we use async wrappers on unblock?
         let writer = BufWriter::new(process.stdin.take().expect("Failed to open stdin"));
         let reader = BufReader::new(process.stdout.take().expect("Failed to open stdout"));
-        let errors = process.stderr.take().map(BufReader::new);
+        let stderr = BufReader::new(process.stderr.take().expect("Failed to open stderr"));
 
         Self::streams(
-            Box::new(BufReader::new(reader)),
+            Box::new(reader),
             Box::new(writer),
-            // errors.map(|errors| Box::new(BufReader::new(errors))),
-            match errors {
-                Some(errors) => Some(Box::new(BufReader::new(errors))),
-                None => None,
-            },
+            Some(Box::new(stderr)),
             id,
             Some(process),
         )
@@ -165,8 +164,8 @@ impl Client {
         cmd: &str,
         args: Vec<&str>,
         port_format: &str,
-        id: usize,
-    ) -> Result<(Self, UnboundedReceiver<Payload>)> {
+        id: DebugAdapterId,
+    ) -> Result<(Self, UnboundedReceiver<(DebugAdapterId, Payload)>)> {
         let port = Self::get_port().await.unwrap();
 
         let process = Command::new(cmd)
@@ -181,40 +180,49 @@ impl Client {
 
         // Wait for adapter to become ready for connection
         time::sleep(time::Duration::from_millis(500)).await;
-
-        let stream = TcpStream::connect(SocketAddr::new(
-            IpAddr::V4(Ipv4Addr::new(127, 0, 0, 1)),
-            port,
-        ))
-        .await?;
+        let socket = SocketAddr::new(IpAddr::V4(Ipv4Addr::new(127, 0, 0, 1)), port);
+        let stream = TcpStream::connect(socket).await?;
 
         let (rx, tx) = stream.into_split();
-        Self::streams(
+        let mut result = Self::streams(
             Box::new(BufReader::new(rx)),
             Box::new(tx),
             None,
             id,
             Some(process),
-        )
+        );
+
+        // Set the socket address for the client
+        if let Ok((client, _)) = &mut result {
+            client.socket = Some(socket);
+        }
+
+        result
     }
 
-    async fn recv(mut server_rx: UnboundedReceiver<Payload>, client_tx: UnboundedSender<Payload>) {
+    async fn recv(
+        id: DebugAdapterId,
+        mut server_rx: UnboundedReceiver<Payload>,
+        client_tx: UnboundedSender<(DebugAdapterId, Payload)>,
+    ) {
         while let Some(msg) = server_rx.recv().await {
             match msg {
                 Payload::Event(ev) => {
-                    client_tx.send(Payload::Event(ev)).expect("Failed to send");
+                    client_tx
+                        .send((id, Payload::Event(ev)))
+                        .expect("Failed to send");
                 }
                 Payload::Response(_) => unreachable!(),
                 Payload::Request(req) => {
                     client_tx
-                        .send(Payload::Request(req))
+                        .send((id, Payload::Request(req)))
                         .expect("Failed to send");
                 }
             }
         }
     }
 
-    pub fn id(&self) -> usize {
+    pub fn id(&self) -> DebugAdapterId {
         self.id
     }
 
@@ -357,6 +365,14 @@ impl Client {
         self.call::<requests::Disconnect>(args)
     }
 
+    pub fn terminate(
+        &mut self,
+        args: Option<TerminateArguments>,
+    ) -> impl Future<Output = Result<Value>> {
+        self.connection_type = None;
+        self.call::<requests::Terminate>(args)
+    }
+
     pub fn launch(&mut self, args: serde_json::Value) -> impl Future<Output = Result<Value>> {
         self.connection_type = Some(ConnectionType::Launch);
         self.starting_request_args = Some(args.clone());

helix-dap/src/lib.rs

@@ -1,12 +1,14 @@
 mod client;
+pub mod registry;
 mod transport;
 mod types;
 
-pub use client::{Client, ConnectionType};
-pub use events::Event;
+pub use client::Client;
 pub use transport::{Payload, Response, Transport};
 pub use types::*;
 
+use serde::de::DeserializeOwned;
+
 use thiserror::Error;
 #[derive(Error, Debug)]
 pub enum Error {
@@ -18,9 +20,86 @@ pub enum Error {
     Timeout(u64),
     #[error("server closed the stream")]
     StreamClosed,
+    #[error("Unhandled")]
+    Unhandled,
     #[error(transparent)]
     ExecutableNotFound(#[from] helix_stdx::env::ExecutableNotFoundError),
     #[error(transparent)]
     Other(#[from] anyhow::Error),
 }
 pub type Result<T> = core::result::Result<T, Error>;
+
+#[derive(Debug)]
+pub enum Request {
+    RunInTerminal(<requests::RunInTerminal as types::Request>::Arguments),
+    StartDebugging(<requests::StartDebugging as types::Request>::Arguments),
+}
+
+impl Request {
+    pub fn parse(command: &str, arguments: Option<serde_json::Value>) -> Result<Self> {
+        use crate::types::Request as _;
+
+        let arguments = arguments.unwrap_or_default();
+        let request = match command {
+            requests::RunInTerminal::COMMAND => Self::RunInTerminal(parse_value(arguments)?),
+            requests::StartDebugging::COMMAND => Self::StartDebugging(parse_value(arguments)?),
+            _ => return Err(Error::Unhandled),
+        };
+
+        Ok(request)
+    }
+}
+
+#[derive(Debug)]
+pub enum Event {
+    Initialized(<events::Initialized as events::Event>::Body),
+    Stopped(<events::Stopped as events::Event>::Body),
+    Continued(<events::Continued as events::Event>::Body),
+    Exited(<events::Exited as events::Event>::Body),
+    Terminated(<events::Terminated as events::Event>::Body),
+    Thread(<events::Thread as events::Event>::Body),
+    Output(<events::Output as events::Event>::Body),
+    Breakpoint(<events::Breakpoint as events::Event>::Body),
+    Module(<events::Module as events::Event>::Body),
+    LoadedSource(<events::LoadedSource as events::Event>::Body),
+    Process(<events::Process as events::Event>::Body),
+    Capabilities(<events::Capabilities as events::Event>::Body),
+    // ProgressStart(),
+    // ProgressUpdate(),
+    // ProgressEnd(),
+    // Invalidated(),
+    Memory(<events::Memory as events::Event>::Body),
+}
+
+impl Event {
+    pub fn parse(event: &str, body: Option<serde_json::Value>) -> Result<Self> {
+        use crate::events::Event as _;
+
+        let body = body.unwrap_or_default();
+        let event = match event {
+            events::Initialized::EVENT => Self::Initialized(parse_value(body)?),
+            events::Stopped::EVENT => Self::Stopped(parse_value(body)?),
+            events::Continued::EVENT => Self::Continued(parse_value(body)?),
+            events::Exited::EVENT => Self::Exited(parse_value(body)?),
+            events::Terminated::EVENT => Self::Terminated(parse_value(body)?),
+            events::Thread::EVENT => Self::Thread(parse_value(body)?),
+            events::Output::EVENT => Self::Output(parse_value(body)?),
+            events::Breakpoint::EVENT => Self::Breakpoint(parse_value(body)?),
+            events::Module::EVENT => Self::Module(parse_value(body)?),
+            events::LoadedSource::EVENT => Self::LoadedSource(parse_value(body)?),
+            events::Process::EVENT => Self::Process(parse_value(body)?),
+            events::Capabilities::EVENT => Self::Capabilities(parse_value(body)?),
+            events::Memory::EVENT => Self::Memory(parse_value(body)?),
+            _ => return Err(Error::Unhandled),
+        };
+
+        Ok(event)
+    }
+}
+
+fn parse_value<T>(value: serde_json::Value) -> Result<T>
+where
+    T: DeserializeOwned,
+{
+    serde_json::from_value(value).map_err(|err| err.into())
+}

helix-dap/src/registry.rs

@@ -0,0 +1,114 @@
+use crate::{Client, Payload, Result, StackFrame};
+use futures_executor::block_on;
+use futures_util::stream::SelectAll;
+use helix_core::syntax::config::DebugAdapterConfig;
+use slotmap::SlotMap;
+use std::fmt;
+use tokio_stream::wrappers::UnboundedReceiverStream;
+
+/// The resgistry is a struct that manages and owns multiple debugger clients
+/// This holds the responsibility of managing the lifecycle of each client
+/// plus showing the heirarcihical nature betweeen them
+pub struct Registry {
+    inner: SlotMap<DebugAdapterId, Client>,
+    /// The active debugger client
+    ///
+    /// TODO: You can have multiple active debuggers, so the concept of a single active debugger
+    /// may need to be changed
+    current_client_id: Option<DebugAdapterId>,
+    /// A stream of incoming messages from all debuggers
+    pub incoming: SelectAll<UnboundedReceiverStream<(DebugAdapterId, Payload)>>,
+}
+
+impl Registry {
+    /// Creates a new DebuggerService instance
+    pub fn new() -> Self {
+        Self {
+            inner: SlotMap::with_key(),
+            current_client_id: None,
+            incoming: SelectAll::new(),
+        }
+    }
+
+    pub fn start_client(
+        &mut self,
+        socket: Option<std::net::SocketAddr>,
+        config: &DebugAdapterConfig,
+    ) -> Result<DebugAdapterId> {
+        self.inner.try_insert_with_key(|id| {
+            let result = match socket {
+                Some(socket) => block_on(Client::tcp(socket, id)),
+                None => block_on(Client::process(
+                    &config.transport,
+                    &config.command,
+                    config.args.iter().map(|arg| arg.as_str()).collect(),
+                    config.port_arg.as_deref(),
+                    id,
+                )),
+            };
+
+            let (mut client, receiver) = result?;
+            self.incoming.push(UnboundedReceiverStream::new(receiver));
+
+            client.config = Some(config.clone());
+            block_on(client.initialize(config.name.clone()))?;
+            client.quirks = config.quirks.clone();
+
+            Ok(client)
+        })
+    }
+
+    pub fn remove_client(&mut self, id: DebugAdapterId) {
+        self.inner.remove(id);
+    }
+
+    pub fn get_client(&self, id: DebugAdapterId) -> Option<&Client> {
+        self.inner.get(id)
+    }
+
+    pub fn get_client_mut(&mut self, id: DebugAdapterId) -> Option<&mut Client> {
+        self.inner.get_mut(id)
+    }
+
+    pub fn get_active_client(&self) -> Option<&Client> {
+        self.current_client_id.and_then(|id| self.get_client(id))
+    }
+
+    pub fn get_active_client_mut(&mut self) -> Option<&mut Client> {
+        self.current_client_id
+            .and_then(|id| self.get_client_mut(id))
+    }
+
+    pub fn set_active_client(&mut self, id: DebugAdapterId) {
+        if self.get_client(id).is_some() {
+            self.current_client_id = Some(id);
+        } else {
+            self.current_client_id = None;
+        }
+    }
+
+    pub fn unset_active_client(&mut self) {
+        self.current_client_id = None;
+    }
+
+    pub fn current_stack_frame(&self) -> Option<&StackFrame> {
+        self.get_active_client()
+            .and_then(|debugger| debugger.current_stack_frame())
+    }
+}
+
+impl Default for Registry {
+    fn default() -> Self {
+        Self::new()
+    }
+}
+
+slotmap::new_key_type! {
+    pub struct DebugAdapterId;
+}
+
+impl fmt::Display for DebugAdapterId {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        write!(f, "{:?}", self.0)
+    }
+}

helix-dap/src/transport.rs

@@ -1,10 +1,10 @@
-use crate::{Error, Event, Result};
+use crate::{registry::DebugAdapterId, Error, Result};
 use anyhow::Context;
 use log::{error, info, warn};
 use serde::{Deserialize, Serialize};
 use serde_json::Value;
-use std::collections::HashMap;
 use std::sync::Arc;
+use std::{collections::HashMap, fmt::Debug};
 use tokio::{
     io::{AsyncBufRead, AsyncBufReadExt, AsyncReadExt, AsyncWrite, AsyncWriteExt},
     sync::{
@@ -32,11 +32,17 @@ pub struct Response {
     pub body: Option<Value>,
 }
 
+#[derive(Debug, PartialEq, Eq, Clone, Deserialize, Serialize)]
+pub struct Event {
+    pub event: String,
+    pub body: Option<Value>,
+}
+
 #[derive(Debug, Clone, Deserialize, Serialize)]
 #[serde(tag = "type", rename_all = "camelCase")]
 pub enum Payload {
     // type = "event"
-    Event(Box<Event>),
+    Event(Event),
     // type = "response"
     Response(Response),
     // type = "request"
@@ -46,7 +52,7 @@ pub enum Payload {
 #[derive(Debug)]
 pub struct Transport {
     #[allow(unused)]
-    id: usize,
+    id: DebugAdapterId,
     pending_requests: Mutex<HashMap<u64, Sender<Result<Response>>>>,
 }
 
@@ -55,7 +61,7 @@ impl Transport {
         server_stdout: Box<dyn AsyncBufRead + Unpin + Send>,
         server_stdin: Box<dyn AsyncWrite + Unpin + Send>,
         server_stderr: Option<Box<dyn AsyncBufRead + Unpin + Send>>,
-        id: usize,
+        id: DebugAdapterId,
     ) -> (UnboundedReceiver<Payload>, UnboundedSender<Payload>) {
         let (client_tx, rx) = unbounded_channel();
         let (tx, client_rx) = unbounded_channel();
@@ -67,7 +73,7 @@ impl Transport {
 
         let transport = Arc::new(transport);
 
-        tokio::spawn(Self::recv(transport.clone(), server_stdout, client_tx));
+        tokio::spawn(Self::recv(id, transport.clone(), server_stdout, client_tx));
         tokio::spawn(Self::send(transport, server_stdin, client_rx));
         if let Some(stderr) = server_stderr {
             tokio::spawn(Self::err(stderr));
@@ -77,12 +83,14 @@ impl Transport {
     }
 
     async fn recv_server_message(
+        id: DebugAdapterId,
         reader: &mut Box<dyn AsyncBufRead + Unpin + Send>,
         buffer: &mut String,
+        content: &mut Vec<u8>,
     ) -> Result<Payload> {
         let mut content_length = None;
         loop {
-            buffer.truncate(0);
+            buffer.clear();
             if reader.read_line(buffer).await? == 0 {
                 return Err(Error::StreamClosed);
             };
@@ -111,17 +119,17 @@ impl Transport {
         }
 
         let content_length = content_length.context("missing content length")?;
+        content.resize(content_length, 0);
+        reader.read_exact(content).await?;
+        let msg = std::str::from_utf8(content).context("invalid utf8 from server")?;
 
-        //TODO: reuse vector
-        let mut content = vec![0; content_length];
-        reader.read_exact(&mut content).await?;
-        let msg = std::str::from_utf8(&content).context("invalid utf8 from server")?;
-
-        info!("<- DAP {}", msg);
+        info!("[{}] <- DAP {}", id, msg);
 
         // try parsing as output (server response) or call (server request)
         let output: serde_json::Result<Payload> = serde_json::from_str(msg);
 
+        content.clear();
+
         Ok(output?)
     }
 
@@ -157,7 +165,7 @@ impl Transport {
         server_stdin: &mut Box<dyn AsyncWrite + Unpin + Send>,
         request: String,
     ) -> Result<()> {
-        info!("-> DAP {}", request);
+        info!("[{}] -> DAP {}", self.id, request);
 
         // send the headers
         server_stdin
@@ -172,15 +180,18 @@ impl Transport {
         Ok(())
     }
 
-    fn process_response(res: Response) -> Result<Response> {
+    fn process_response(&self, res: Response) -> Result<Response> {
         if res.success {
-            info!("<- DAP success in response to {}", res.request_seq);
+            info!(
+                "[{}] <- DAP success in response to {}",
+                self.id, res.request_seq
+            );
 
             Ok(res)
         } else {
             error!(
-                "<- DAP error {:?} ({:?}) for command #{} {}",
-                res.message, res.body, res.request_seq, res.command
+                "[{}] <- DAP error {:?} ({:?}) for command #{} {}",
+                self.id, res.message, res.body, res.request_seq, res.command
             );
 
             Err(Error::Other(anyhow::format_err!("{:?}", res.body)))
@@ -198,7 +209,7 @@ impl Transport {
                 let tx = self.pending_requests.lock().await.remove(&request_seq);
 
                 match tx {
-                    Some(tx) => match tx.send(Self::process_response(res)).await {
+                    Some(tx) => match tx.send(self.process_response(res)).await {
                         Ok(_) => (),
                         Err(_) => error!(
                             "Tried sending response into a closed channel (id={:?}), original request likely timed out",
@@ -218,37 +229,58 @@ impl Transport {
                 ref seq,
                 ..
             }) => {
-                info!("<- DAP request {} #{}", command, seq);
+                info!("[{}] <- DAP request {} #{}", self.id, command, seq);
                 client_tx.send(msg).expect("Failed to send");
                 Ok(())
             }
             Payload::Event(ref event) => {
-                info!("<- DAP event {:?}", event);
+                info!("[{}] <- DAP event {:?}", self.id, event);
                 client_tx.send(msg).expect("Failed to send");
                 Ok(())
             }
         }
     }
 
-    async fn recv_inner(
+    async fn recv(
+        id: DebugAdapterId,
         transport: Arc<Self>,
         mut server_stdout: Box<dyn AsyncBufRead + Unpin + Send>,
         client_tx: UnboundedSender<Payload>,
-    ) -> Result<()> {
-        let mut recv_buffer = String::new();
-        loop {
-            let msg = Self::recv_server_message(&mut server_stdout, &mut recv_buffer).await?;
-            transport.process_server_message(&client_tx, msg).await?;
-        }
-    }
-
-    async fn recv(
-        transport: Arc<Self>,
-        server_stdout: Box<dyn AsyncBufRead + Unpin + Send>,
-        client_tx: UnboundedSender<Payload>,
     ) {
-        if let Err(err) = Self::recv_inner(transport, server_stdout, client_tx).await {
-            error!("err: <- {:?}", err);
+        let mut recv_buffer = String::new();
+        let mut content_buffer = Vec::new();
+        loop {
+            match Self::recv_server_message(
+                id,
+                &mut server_stdout,
+                &mut recv_buffer,
+                &mut content_buffer,
+            )
+            .await
+            {
+                Ok(msg) => match transport.process_server_message(&client_tx, msg).await {
+                    Ok(_) => (),
+                    Err(err) => {
+                        error!(" [{id}] err: <- {err:?}");
+                        break;
+                    }
+                },
+                Err(err) => {
+                    if !matches!(err, Error::StreamClosed) {
+                        error!("Exiting after unexpected error: {err:?}");
+                    }
+
+                    // Close any outstanding requests.
+                    for (id, tx) in transport.pending_requests.lock().await.drain() {
+                        match tx.send(Err(Error::StreamClosed)).await {
+                            Ok(_) => (),
+                            Err(_) => {
+                                error!("Could not close request on a closed channel (id={id})");
+                            }
+                        }
+                    }
+                }
+            }
         }
     }