(crawler) Correct content type probing to only run on URLs that are suspected to be binary

For now this is just the ctime for the feeds db.  We may want to store this per-record in the future.

ROADMAP.md

@@ -21,7 +21,7 @@ word n-grams known beforehand.  This limits the ability to interpret longer quer
 The positions mask should be supplemented or replaced with a more accurate (e.g.) gamma coded positions
 list, as is the civilized way of doing this.
 
-Completed with PR https://github.com/MarginaliaSearch/MarginaliaSearch/pull/99
+Completed with PR [#99](https://github.com/MarginaliaSearch/MarginaliaSearch/pull/99)
 
 ## Hybridize crawler w/ Common Crawl data
 
@@ -41,6 +41,12 @@ The search engine has a bit of a problem showing spicy content mixed in with the
 to have a way to filter this out.  It's likely something like a URL blacklist (e.g. [UT1](https://dsi.ut-capitole.fr/blacklists/index_en.php) )
 combined with naive bayesian filter would go a long way, or something more sophisticated...?
 
+## Web Design Overhaul
+
+The design is kinda clunky and hard to maintain, and needlessly outdated-looking.  
+
+In progress: PR [#127](https://github.com/MarginaliaSearch/MarginaliaSearch/pull/127)  -- demo available at https://test.marginalia.nu/
+
 ## Additional Language Support
 
 It would be desirable if the search engine supported more languages than English.  This is partially about
@@ -56,7 +62,7 @@ it should be extended to all domains.  It would also be interesting to offer sea
 RSS data itself, or use the RSS set to feed a special live index that updates faster than the
 main dataset. 
 
-Completed with PR [#122](https://github.com/MarginaliaSearch/MarginaliaSearch/pull/122) 
+Completed with PR [#122](https://github.com/MarginaliaSearch/MarginaliaSearch/pull/122) and PR [#125](https://github.com/MarginaliaSearch/MarginaliaSearch/pull/125)
 
 ## Support for binary formats like PDF
 

code/functions/live-capture/api/java/nu/marginalia/api/feeds/FeedsClient.java

@@ -59,12 +59,6 @@ public class FeedsClient {
                 .forEachRemaining(rsp -> consumer.accept(rsp.getDomain(), new ArrayList<>(rsp.getUrlList())));
     }
 
-    public record UpdatedDomain(String domain, List<String> urls) {
-        public UpdatedDomain(RpcUpdatedLinksResponse rsp) {
-            this(rsp.getDomain(), new ArrayList<>(rsp.getUrlList()));
-        }
-    }
-
     /** Get the hash of the feed data, for identifying when the data has been updated */
     public String getFeedDataHash() {
         return channelPool.call(FeedApiGrpc.FeedApiBlockingStub::getFeedDataHash)

code/functions/live-capture/api/src/main/protobuf/feeds.proto

@@ -46,6 +46,7 @@ message RpcFeed {
   string feedUrl = 3;
   string updated = 4;
   repeated RpcFeedItem items = 5;
+  int64 fetchTimestamp = 6;
 }
 
 message RpcFeedItem {

code/functions/live-capture/java/nu/marginalia/rss/db/FeedDb.java

@@ -12,9 +12,11 @@ import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
 import java.io.BufferedInputStream;
+import java.io.IOException;
 import java.nio.file.Files;
 import java.nio.file.Path;
 import java.nio.file.StandardCopyOption;
+import java.nio.file.attribute.PosixFileAttributes;
 import java.security.MessageDigest;
 import java.time.Instant;
 import java.util.Base64;
@@ -209,4 +211,20 @@ public class FeedDb {
 
         reader.getLinksUpdatedSince(since, consumer);
     }
+
+    public Instant getFetchTime() {
+        if (!Files.exists(readerDbPath)) {
+            return Instant.ofEpochMilli(0);
+        }
+
+        try {
+            return Files.readAttributes(readerDbPath, PosixFileAttributes.class)
+                    .creationTime()
+                    .toInstant();
+        }
+        catch (IOException ex) {
+            logger.error("Failed to read the creatiom time of {}", readerDbPath);
+            return Instant.ofEpochMilli(0);
+        }
+    }
 }

code/functions/live-capture/java/nu/marginalia/rss/svc/FeedFetcherService.java

@@ -316,6 +316,8 @@ public class FeedFetcherService {
 
     public FeedItems parseFeed(String feedData, FeedDefinition definition) {
         try {
+            feedData = sanitizeEntities(feedData);
+
             List<Item> rawItems = rssReader.read(
                     // Massage the data to maximize the possibility of the flaky XML parser consuming it
                     new BOMInputStream(new ByteArrayInputStream(feedData.trim().getBytes(StandardCharsets.UTF_8)), false)
@@ -342,6 +344,32 @@ public class FeedFetcherService {
         }
     }
 
+    private static final Map<String, String> HTML_ENTITIES = Map.of(
+            "&raquo;", "»",
+            "&laquo;", "«",
+            "&mdash;", "--",
+            "&ndash;", "-",
+            "&rsquo;", "'",
+            "&lsquo;", "'",
+            "&nbsp;", ""
+    );
+
+    /** The XML parser will blow up if you insert HTML entities in the feed XML,
+     * which is unfortunately relatively common.  Replace them as far as is possible
+     * with their corresponding characters
+     */
+    static String sanitizeEntities(String feedData) {
+        String result = feedData;
+        for (Map.Entry<String, String> entry : HTML_ENTITIES.entrySet()) {
+            result = result.replace(entry.getKey(), entry.getValue());
+        }
+
+        // Handle lone ampersands not part of a recognized XML entity
+        result = result.replaceAll("&(?!(amp|lt|gt|apos|quot);)", "&amp;");
+
+        return result;
+    }
+
     /** Decide whether to keep URI fragments in the feed items.
      * <p></p>
      * We keep fragments if there are multiple different fragments in the items.

code/functions/live-capture/java/nu/marginalia/rss/svc/FeedsGrpcService.java

@@ -107,8 +107,7 @@ public class FeedsGrpcService extends FeedApiGrpc.FeedApiImplBase implements Dis
 
     @Override
     public void getFeed(RpcDomainId request,
-                        StreamObserver<RpcFeed> responseObserver)
-    {
+                        StreamObserver<RpcFeed> responseObserver) {
         if (!feedDb.isEnabled()) {
             responseObserver.onError(new IllegalStateException("Feed database is disabled on this node"));
             return;
@@ -126,7 +125,8 @@ public class FeedsGrpcService extends FeedApiGrpc.FeedApiImplBase implements Dis
                 .setDomainId(request.getDomainId())
                 .setDomain(domainName.get().toString())
                 .setFeedUrl(feedItems.feedUrl())
-                .setUpdated(feedItems.updated());
+                .setUpdated(feedItems.updated())
+                .setFetchTimestamp(feedDb.getFetchTime().toEpochMilli());
 
         for (var item : feedItems.items()) {
             retB.addItemsBuilder()

code/functions/live-capture/test/nu/marginalia/rss/svc/FeedFetcherServiceTest.java

@@ -99,7 +99,9 @@ class FeedFetcherServiceTest extends AbstractModule {
         feedFetcherService.setDeterministic();
         feedFetcherService.updateFeeds(FeedFetcherService.UpdateMode.REFRESH);
 
-        Assertions.assertFalse(feedDb.getFeed(new EdgeDomain("www.marginalia.nu")).isEmpty());
+        var result = feedDb.getFeed(new EdgeDomain("www.marginalia.nu"));
+        System.out.println(result);
+        Assertions.assertFalse(result.isEmpty());
     }
 
     @Tag("flaky")

code/functions/live-capture/test/nu/marginalia/rss/svc/TestXmlSanitization.java

@@ -0,0 +1,26 @@
+package nu.marginalia.rss.svc;
+
+import org.junit.jupiter.api.Assertions;
+import org.junit.jupiter.api.Test;
+
+public class TestXmlSanitization {
+
+    @Test
+    public void testPreservedEntities() {
+        Assertions.assertEquals("&", FeedFetcherService.sanitizeEntities("&"));
+        Assertions.assertEquals("<", FeedFetcherService.sanitizeEntities("<"));
+        Assertions.assertEquals(">", FeedFetcherService.sanitizeEntities(">"));
+        Assertions.assertEquals(""", FeedFetcherService.sanitizeEntities("""));
+        Assertions.assertEquals("'", FeedFetcherService.sanitizeEntities("'"));
+    }
+
+    @Test
+    public void testStrayAmpersand() {
+        Assertions.assertEquals("Bed & Breakfast", FeedFetcherService.sanitizeEntities("Bed & Breakfast"));
+    }
+
+    @Test
+    public void testTranslatedHtmlEntity() {
+        Assertions.assertEquals("Foo -- Bar", FeedFetcherService.sanitizeEntities("Foo — Bar"));
+    }
+}

code/processes/crawling-process/java/nu/marginalia/crawl/fetcher/HttpFetcherImpl.java

@@ -139,7 +139,7 @@ public class HttpFetcherImpl implements HttpFetcher {
     public ContentTypeProbeResult probeContentType(EdgeUrl url,
                                                    WarcRecorder warcRecorder,
                                                    ContentTags tags) throws RateLimitException {
-        if (tags.isEmpty()) {
+        if (tags.isEmpty() && contentTypeLogic.isUrlLikeBinary(url)) {
             var headBuilder = new Request.Builder().head()
                     .addHeader("User-agent", userAgentString)
                     .addHeader("Accept-Encoding", "gzip")

code/processes/crawling-process/test/nu/marginalia/crawl/retreival/fetcher/ContentTypeProberTest.java

@@ -42,24 +42,24 @@ class ContentTypeProberTest {
         port = r.nextInt(10000) + 8000;
         server = HttpServer.create(new InetSocketAddress("127.0.0.1", port), 10);
 
-        server.createContext("/html", exchange -> {
+        server.createContext("/html.gz", exchange -> {
             exchange.getResponseHeaders().add("Content-Type", "text/html");
             exchange.sendResponseHeaders(200, -1);
             exchange.close();
         });
-        server.createContext("/redir", exchange -> {
-            exchange.getResponseHeaders().add("Location", "/html");
+        server.createContext("/redir.gz", exchange -> {
+            exchange.getResponseHeaders().add("Location", "/html.gz");
             exchange.sendResponseHeaders(301, -1);
             exchange.close();
         });
 
-        server.createContext("/bin", exchange -> {
+        server.createContext("/bin.gz", exchange -> {
             exchange.getResponseHeaders().add("Content-Type", "application/binary");
             exchange.sendResponseHeaders(200, -1);
             exchange.close();
         });
 
-        server.createContext("/timeout", exchange -> {
+        server.createContext("/timeout.gz", exchange -> {
             try {
                 Thread.sleep(15_000);
             } catch (InterruptedException e) {
@@ -73,10 +73,10 @@ class ContentTypeProberTest {
 
         server.start();
 
-        htmlEndpoint = EdgeUrl.parse("http://localhost:" + port + "/html").get();
-        binaryEndpoint = EdgeUrl.parse("http://localhost:" + port + "/bin").get();
-        timeoutEndpoint = EdgeUrl.parse("http://localhost:" + port + "/timeout").get();
-        htmlRedirEndpoint = EdgeUrl.parse("http://localhost:" + port + "/redir").get();
+        htmlEndpoint = EdgeUrl.parse("http://localhost:" + port + "/html.gz").get();
+        binaryEndpoint = EdgeUrl.parse("http://localhost:" + port + "/bin.gz").get();
+        timeoutEndpoint = EdgeUrl.parse("http://localhost:" + port + "/timeout.gz").get();
+        htmlRedirEndpoint = EdgeUrl.parse("http://localhost:" + port + "/redir.gz").get();
 
         fetcher = new HttpFetcherImpl("test");
         recorder = new WarcRecorder(warcFile);