mundanevision20/plocate
1
1
Fork 0
mirror of http://git.sesse.net/plocate synced 2025-10-06 03:12:40 +02:00
Code Activity

Format everything with clang-format.

Browse Source 
clang-format isn't ideal, but it's better than manual formatting
in the long run.
This commit is contained in:
Steinar H. Gunderson
2020-09-29 00:13:18 +02:00
parent ce2c78a6e4
commit 00639a652f
3 changed files with 156 additions and 77 deletions
Download Patch File Download Diff File Expand all files Collapse all files

67
.clang-format Normal file
View File

@@ -0,0 +1,67 @@
AlignAfterOpenBracket: Align
AlignConsecutiveAssignments: false
AlignConsecutiveDeclarations: false
AlignEscapedNewlines: DontAlign
AlignOperands: false
AlignTrailingComments: false
AllowAllParametersOfDeclarationOnNextLine: true
AllowShortBlocksOnASingleLine: false
AllowShortCaseLabelsOnASingleLine: false
AllowShortFunctionsOnASingleLine: Inline
AllowShortIfStatementsOnASingleLine: false
AllowShortLoopsOnASingleLine: false
AlwaysBreakAfterReturnType: None
AlwaysBreakBeforeMultilineStrings: false
AlwaysBreakTemplateDeclarations: true
BinPackArguments: true
BinPackParameters: true
BreakBeforeBinaryOperators: None
BreakBeforeBraces: WebKit
BreakBeforeTernaryOperators: true
BreakConstructorInitializers: BeforeColon
#BreakInheritanceList: BeforeColon
BreakStringLiterals: false
CompactNamespaces: false
ConstructorInitializerAllOnOneLineOrOnePerLine: false
Cpp11BracedListStyle: false
DerivePointerAlignment: false
FixNamespaceComments: false
IncludeBlocks: Regroup
IndentCaseLabels: false
IndentPPDirectives: None
IndentWrappedFunctionNames: false
KeepEmptyLinesAtTheStartOfBlocks: false
Language: Cpp
MaxEmptyLinesToKeep: 1
NamespaceIndentation: None
PointerAlignment: Right
ReflowComments: true
SortIncludes: true
SortUsingDeclarations: true
SpaceAfterCStyleCast: false
SpaceAfterTemplateKeyword: false
SpaceBeforeAssignmentOperators: true
#SpaceBeforeCpp11BracedList: false
SpaceBeforeCtorInitializerColon: true
SpaceBeforeInheritanceColon: true
SpaceBeforeParens: ControlStatements
SpaceBeforeRangeBasedForLoopColon: true
SpaceInEmptyParentheses: false
SpacesBeforeTrailingComments: 2
SpacesInAngles: false
SpacesInCStyleCastParentheses: false
SpacesInContainerLiterals: false
SpacesInParentheses: false
SpacesInSquareBrackets: false
Standard: Cpp11
# HACK: TabWidth 200 makes sure we use tabs for single-tab indentation only,
# never for alignment. This makes the code readable at any tab width.
# It works only because we have ColumnLimit 0.
ColumnLimit: 0
UseTab: ForContinuationAndIndentation
AccessModifierOffset: -200
ConstructorInitializerIndentWidth: 200
ContinuationIndentWidth: 200
IndentWidth: 200
TabWidth: 200

67
plocate-build.cpp
View File

@@ -1,23 +1,23 @@
#include <assert.h>
#include <stdio.h>
#include <string.h>
#include <algorithm>
#include <unordered_map>
#include <string>
#include <vector>
#include <chrono>
#include <unistd.h>
#include <fcntl.h>
#include <sys/mman.h>
#include <arpa/inet.h>
#include <endian.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <zstd.h>
#include "vp4.h"
#define P4NENC_BOUND(n) ((n+127)/128+(n+32)*sizeof(uint32_t))
#include <algorithm>
#include <arpa/inet.h>
#include <assert.h>
#include <chrono>
#include <endian.h>
#include <fcntl.h>
#include <stdio.h>
#include <string.h>
#include <string>
#include <sys/mman.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <unistd.h>
#include <unordered_map>
#include <vector>
#include <zstd.h>
#define P4NENC_BOUND(n) ((n + 127) / 128 + (n + 32) * sizeof(uint32_t))
#define dprintf(...)
//#define dprintf(...) fprintf(stderr, __VA_ARGS__);
@@ -25,7 +25,7 @@ using namespace std;
using namespace std::chrono;
string zstd_compress(const string &src, string *tempbuf);
static inline uint32_t read_unigram(const string_view s, size_t idx)
{
if (idx < s.size()) {
@@ -42,16 +42,14 @@ static inline uint32_t read_trigram(const string_view s, size_t start)
(read_unigram(s, start + 2) << 16);
}
enum
{
DBE_NORMAL = 0, /* A non-directory file */
DBE_DIRECTORY = 1, /* A directory */
DBE_END = 2 /* End of directory contents; contains no name */
enum {
DBE_NORMAL = 0, /* A non-directory file */
DBE_DIRECTORY = 1, /* A directory */
DBE_END = 2 /* End of directory contents; contains no name */
};
// From mlocate.
struct db_header
{
struct db_header {
uint8_t magic[8];
uint32_t conf_size;
uint8_t version;
@@ -60,8 +58,7 @@ struct db_header
};
// From mlocate.
struct db_directory
{
struct db_directory {
uint64_t time_sec;
uint32_t time_nsec;
uint8_t pad[4];
@@ -143,7 +140,8 @@ void PostingListBuilder::write_header(uint32_t docid)
class Corpus {
public:
Corpus(size_t block_size) : block_size(block_size) {}
Corpus(size_t block_size)
: block_size(block_size) {}
void add_file(string filename);
void flush_block();
@@ -207,8 +205,8 @@ const char *handle_directory(const char *ptr, Corpus *corpus)
if (dir_path == "/") {
dir_path = "";
}
for ( ;; ) {
for (;;) {
uint8_t type = *ptr++;
if (type == DBE_NORMAL) {
string filename = ptr;
@@ -282,7 +280,8 @@ void do_build(const char *infile, const char *outfile, int block_size)
break;
}
string s(buf);
if (s.back() == '\n') s.pop_back();
if (s.back() == '\n')
s.pop_back();
corpus.add_file(move(s));
}
fclose(fp);
@@ -299,7 +298,7 @@ void do_build(const char *infile, const char *outfile, int block_size)
bytes_used += pl_builder.encoded.size();
}
dprintf("%zu files, %zu different trigrams, %zu entries, avg len %.2f, longest %zu\n",
corpus.num_files, corpus.invindex.size(), trigrams, double(trigrams) / corpus.invindex.size(), longest_posting_list);
corpus.num_files, corpus.invindex.size(), trigrams, double(trigrams) / corpus.invindex.size(), longest_posting_list);
dprintf("%zu bytes used for posting lists (%.2f bits/entry)\n", bytes_used, 8 * bytes_used / double(trigrams));
//steady_clock::time_point end = steady_clock::now();
@@ -356,7 +355,7 @@ void do_build(const char *infile, const char *outfile, int block_size)
bytes_for_filename_index += sizeof(offset);
bytes_for_filenames += filename.size();
}
// Write the actual filenames.
for (const string &filename : corpus.filename_blocks) {
fwrite(filename.data(), filename.size(), 1, outfp);

99
plocate.cpp
View File

@@ -1,27 +1,27 @@
#include <stdio.h>
#include <string.h>
#include <algorithm>
#include <unordered_map>
#include <string>
#include <vector>
#include <chrono>
#include <unistd.h>
#include <fcntl.h>
#include <sys/mman.h>
#include <arpa/inet.h>
#include <endian.h>
#include <zstd.h>
#include "vp4.h"
#define P4NENC_BOUND(n) ((n+127)/128+(n+32)*sizeof(uint32_t))
#include <algorithm>
#include <arpa/inet.h>
#include <chrono>
#include <endian.h>
#include <fcntl.h>
#include <stdio.h>
#include <string.h>
#include <string>
#include <sys/mman.h>
#include <unistd.h>
#include <unordered_map>
#include <vector>
#include <zstd.h>
#define P4NENC_BOUND(n) ((n + 127) / 128 + (n + 32) * sizeof(uint32_t))
using namespace std;
using namespace std::chrono;
#define dprintf(...)
//#define dprintf(...) fprintf(stderr, __VA_ARGS__);
static inline uint32_t read_unigram(const string &s, size_t idx)
{
if (idx < s.size()) {
@@ -33,12 +33,12 @@ static inline uint32_t read_unigram(const string &s, size_t idx)
static inline uint32_t read_trigram(const string &s, size_t start)
{
return read_unigram(s, start) |
(read_unigram(s, start + 1) << 8) |
return read_unigram(s, start) | (read_unigram(s, start + 1) << 8) |
(read_unigram(s, start + 2) << 16);
}
bool has_access(const char *filename, unordered_map<string, bool> *access_rx_cache)
bool has_access(const char *filename,
unordered_map<string, bool> *access_rx_cache)
{
const char *end = strchr(filename + 1, '/');
while (end != nullptr) {
@@ -71,7 +71,8 @@ public:
Corpus(int fd);
~Corpus();
const Trigram *find_trigram(uint32_t trgm) const;
const unsigned char *get_compressed_posting_list(const Trigram *trigram) const;
const unsigned char *
get_compressed_posting_list(const Trigram *trigram) const;
string_view get_compressed_filename_block(uint32_t docid) const;
private:
@@ -112,16 +113,17 @@ Corpus::~Corpus()
const Trigram *Corpus::find_trigram(uint32_t trgm) const
{
const Trigram *trgmptr = lower_bound(trgm_begin, trgm_end, trgm, [](const Trigram &trgm, uint32_t t) {
return trgm.trgm < t;
});
const Trigram *trgmptr = lower_bound(
trgm_begin, trgm_end, trgm,
[](const Trigram &trgm, uint32_t t) { return trgm.trgm < t; });
if (trgmptr == trgm_end || trgmptr->trgm != trgm) {
return nullptr;
}
return trgmptr;
}
const unsigned char *Corpus::get_compressed_posting_list(const Trigram *trgmptr) const
const unsigned char *
Corpus::get_compressed_posting_list(const Trigram *trgmptr) const
{
return reinterpret_cast<const unsigned char *>(data + trgmptr->offset);
}
@@ -129,24 +131,29 @@ const unsigned char *Corpus::get_compressed_posting_list(const Trigram *trgmptr)
string_view Corpus::get_compressed_filename_block(uint32_t docid) const
{
const char *compressed = (const char *)(data + filename_offsets[docid]);
size_t compressed_size = filename_offsets[docid + 1] - filename_offsets[docid]; // Allowed we have a sentinel block at the end.
return {compressed, compressed_size};
size_t compressed_size =
filename_offsets[docid + 1] -
filename_offsets[docid]; // Allowed we have a sentinel block at the end.
return { compressed, compressed_size };
}
size_t scan_docid(const string &needle, uint32_t docid, const Corpus &corpus, unordered_map<string, bool> *access_rx_cache)
size_t scan_docid(const string &needle, uint32_t docid, const Corpus &corpus,
unordered_map<string, bool> *access_rx_cache)
{
string_view compressed = corpus.get_compressed_filename_block(docid);
size_t matched = 0;
string block;
block.resize(ZSTD_getFrameContentSize(compressed.data(), compressed.size()) + 1);
block.resize(ZSTD_getFrameContentSize(compressed.data(), compressed.size()) +
1);
ZSTD_decompress(&block[0], block.size(), compressed.data(), compressed.size());
ZSTD_decompress(&block[0], block.size(), compressed.data(),
compressed.size());
block[block.size() - 1] = '\0';
for (const char *filename = block.data();
filename != block.data() + block.size();
filename += strlen(filename) + 1) {
filename != block.data() + block.size();
filename += strlen(filename) + 1) {
if (strstr(filename, needle.c_str()) == nullptr) {
continue;
}
@@ -172,7 +179,7 @@ void do_search_file(const string &needle, const char *filename)
exit(EXIT_FAILURE);
}
//steady_clock::time_point start = steady_clock::now();
// steady_clock::time_point start = steady_clock::now();
if (access("/", R_OK | X_OK)) {
// We can't find anything, no need to bother...
return;
@@ -195,24 +202,27 @@ void do_search_file(const string &needle, const char *filename)
auto last = unique(trigrams.begin(), trigrams.end());
trigrams.erase(last, trigrams.end());
}
sort(trigrams.begin(), trigrams.end(), [&](const Trigram *a, const Trigram *b) {
return a->num_docids < b->num_docids;
});
sort(trigrams.begin(), trigrams.end(),
[&](const Trigram *a, const Trigram *b) {
return a->num_docids < b->num_docids;
});
vector<uint32_t> in1, in2, out;
for (const Trigram *trgmptr : trigrams) {
//uint32_t trgm = trgmptr->trgm;
// uint32_t trgm = trgmptr->trgm;
size_t num = trgmptr->num_docids;
const unsigned char *pldata = corpus.get_compressed_posting_list(trgmptr);
if (in1.empty()) {
in1.resize(num + 128);
p4nd1dec128v32(const_cast<unsigned char *>(pldata), num, &in1[0]);
in1.resize(num);
dprintf("trigram '%c%c%c' decoded to %zu entries\n", trgm & 0xff, (trgm >> 8) & 0xff, (trgm >> 16) & 0xff, num);
dprintf("trigram '%c%c%c' decoded to %zu entries\n", trgm & 0xff,
(trgm >> 8) & 0xff, (trgm >> 16) & 0xff, num);
} else {
if (num > in1.size() * 100) {
dprintf("trigram '%c%c%c' has %zu entries, ignoring the rest (will weed out false positives later)\n",
trgm & 0xff, (trgm >> 8) & 0xff, (trgm >> 16) & 0xff, num);
dprintf("trigram '%c%c%c' has %zu entries, ignoring the rest (will "
"weed out false positives later)\n",
trgm & 0xff, (trgm >> 8) & 0xff, (trgm >> 16) & 0xff, num);
break;
}
@@ -222,9 +232,12 @@ void do_search_file(const string &needle, const char *filename)
p4nd1dec128v32(const_cast<unsigned char *>(pldata), num, &in2[0]);
out.clear();
set_intersection(in1.begin(), in1.end(), in2.begin(), in2.begin() + num, back_inserter(out));
set_intersection(in1.begin(), in1.end(), in2.begin(), in2.begin() + num,
back_inserter(out));
swap(in1, out);
dprintf("trigram '%c%c%c' decoded to %zu entries, %zu left\n", trgm & 0xff, (trgm >> 8) & 0xff, (trgm >> 16) & 0xff, num, in1.size());
dprintf("trigram '%c%c%c' decoded to %zu entries, %zu left\n",
trgm & 0xff, (trgm >> 8) & 0xff, (trgm >> 16) & 0xff, num,
in1.size());
if (in1.empty()) {
dprintf("no matches (intersection list is empty)\n");
break;
@@ -234,7 +247,7 @@ void do_search_file(const string &needle, const char *filename)
steady_clock::time_point end = steady_clock::now();
dprintf("Intersection took %.1f ms. Doing final verification and printing:\n",
1e3 * duration<float>(end - start).count());
1e3 * duration<float>(end - start).count());
unordered_map<string, bool> access_rx_cache;
@@ -244,7 +257,7 @@ void do_search_file(const string &needle, const char *filename)
}
end = steady_clock::now();
dprintf("Done in %.1f ms, found %d matches.\n",
1e3 * duration<float>(end - start).count(), matched);
1e3 * duration<float>(end - start).count(), matched);
}
int main(int argc, char **argv)