(crawler) Add custom cookie store implementation

Apache HttpClient's cookie implementation builds an enormous concurrent hashmap with every cookie for every domain ever crawled.  This is a big waste of resources.

Replacing it with a fairly crude domain-isolated instance, as we are primarily interested in answering whether a cookie is set, and we will never retain cookies long term.

code/processes/converting-process/test/nu/marginalia/converting/CrawlingThenConvertingIntegrationTest.java

@@ -20,7 +20,6 @@ import nu.marginalia.model.crawldata.CrawledDocument;
 import nu.marginalia.model.crawldata.CrawledDomain;
 import nu.marginalia.model.crawldata.SerializableCrawlData;
 import nu.marginalia.parquet.crawldata.CrawledDocumentParquetRecordFileWriter;
-import org.apache.hc.client5.http.cookie.BasicCookieStore;
 import org.junit.jupiter.api.*;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
@@ -247,7 +246,7 @@ public class CrawlingThenConvertingIntegrationTest {
     private CrawledDomain crawl(CrawlerMain.CrawlSpecRecord specs, Predicate<EdgeDomain> domainBlacklist) throws Exception {
         List<SerializableCrawlData> data = new ArrayList<>();
 
-        try (var recorder = new WarcRecorder(fileName, new BasicCookieStore());
+        try (var recorder = new WarcRecorder(fileName);
              var db = new DomainStateDb(dbTempFile))
         {
             new CrawlerRetreiver(httpFetcher, new DomainProber(domainBlacklist), specs, db, recorder).crawlDomain();

code/processes/crawling-process/java/nu/marginalia/crawl/CrawlerMain.java

@@ -438,7 +438,7 @@ public class CrawlerMain extends ProcessMainClass {
                 Files.deleteIfExists(tempFile);
             }
 
-            try (var warcRecorder = new WarcRecorder(newWarcFile, fetcher); // write to a temp file for now
+            try (var warcRecorder = new WarcRecorder(newWarcFile); // write to a temp file for now
                  var retriever = new CrawlerRetreiver(fetcher, domainProber, specification, domainStateDb, warcRecorder);
                  CrawlDataReference reference = getReference()
             )

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

@@ -1,6 +1,6 @@
 package nu.marginalia.crawl.fetcher;
 
-import org.apache.hc.core5.http.io.support.ClassicRequestBuilder;
+import org.apache.hc.client5.http.classic.methods.HttpGet;
 
 /** Encapsulates request modifiers; the ETag and Last-Modified tags for a resource */
 public record ContentTags(String etag, String lastMod) {
@@ -17,14 +17,14 @@ public record ContentTags(String etag, String lastMod) {
     }
 
     /** Paints the tags onto the request builder. */
-    public void paint(ClassicRequestBuilder getBuilder) {
+    public void paint(HttpGet request) {
 
         if (etag != null) {
-            getBuilder.addHeader("If-None-Match", etag);
+            request.addHeader("If-None-Match", etag);
         }
 
         if (lastMod != null) {
-            getBuilder.addHeader("If-Modified-Since", lastMod);
+            request.addHeader("If-Modified-Since", lastMod);
         }
     }
 }

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

@@ -1,34 +0,0 @@
-package nu.marginalia.crawl.fetcher;
-
-import java.io.IOException;
-import java.net.CookieHandler;
-import java.net.URI;
-import java.util.List;
-import java.util.Map;
-import java.util.concurrent.ConcurrentHashMap;
-
-public class Cookies extends CookieHandler {
-    final ThreadLocal<ConcurrentHashMap<String, List<String>>> cookieJar = ThreadLocal.withInitial(ConcurrentHashMap::new);
-
-    public void clear() {
-        cookieJar.get().clear();
-    }
-
-    public boolean hasCookies() {
-        return !cookieJar.get().isEmpty();
-    }
-
-    public List<String> getCookies() {
-        return cookieJar.get().values().stream().flatMap(List::stream).toList();
-    }
-
-    @Override
-    public Map<String, List<String>> get(URI uri, Map<String, List<String>> requestHeaders) throws IOException {
-        return cookieJar.get();
-    }
-
-    @Override
-    public void put(URI uri, Map<String, List<String>> responseHeaders) throws IOException {
-        cookieJar.get().putAll(responseHeaders);
-    }
-}

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

@@ -0,0 +1,56 @@
+package nu.marginalia.crawl.fetcher;
+
+import org.apache.hc.client5.http.classic.methods.HttpUriRequestBase;
+import org.apache.hc.core5.http.ClassicHttpRequest;
+import org.apache.hc.core5.http.HttpResponse;
+
+import java.util.HashMap;
+import java.util.Map;
+import java.util.StringJoiner;
+
+public class DomainCookies {
+    private final Map<String, String> cookies = new HashMap<>();
+
+    public boolean hasCookies() {
+        return !cookies.isEmpty();
+    }
+
+    public void updateCookieStore(HttpResponse response) {
+        for (var header : response.getHeaders()) {
+            if (header.getName().equalsIgnoreCase("Set-Cookie")) {
+                parseCookieHeader(header.getValue());
+            }
+        }
+    }
+
+    private void parseCookieHeader(String value) {
+        // Parse the Set-Cookie header value and extract the cookies
+
+        String[] parts = value.split(";");
+        String cookie = parts[0].trim();
+
+        if (cookie.contains("=")) {
+            String[] cookieParts = cookie.split("=");
+            String name = cookieParts[0].trim();
+            String val = cookieParts[1].trim();
+            cookies.put(name, val);
+        }
+    }
+
+    public void paintRequest(HttpUriRequestBase request) {
+        request.addHeader("Cookie", createCookieHeader());
+    }
+
+    public void paintRequest(ClassicHttpRequest request) {
+        request.addHeader("Cookie", createCookieHeader());
+    }
+
+    private String createCookieHeader() {
+        StringJoiner sj = new StringJoiner("; ");
+        for (var cookie : cookies.entrySet()) {
+            sj.add(cookie.getKey() + "=" + cookie.getValue());
+        }
+        return sj.toString();
+    }
+
+}

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

@@ -23,6 +23,7 @@ public interface HttpFetcher extends AutoCloseable {
 
     HttpFetchResult fetchContent(EdgeUrl url,
                                  WarcRecorder recorder,
+                                 DomainCookies cookies,
                                  CrawlDelayTimer timer,
                                  ContentTags tags,
                                  ProbeType probeType);

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

@@ -17,6 +17,7 @@ import nu.marginalia.model.crawldata.CrawlerDomainStatus;
 import org.apache.hc.client5.http.ConnectionKeepAliveStrategy;
 import org.apache.hc.client5.http.HttpRequestRetryStrategy;
 import org.apache.hc.client5.http.classic.HttpClient;
+import org.apache.hc.client5.http.classic.methods.HttpGet;
 import org.apache.hc.client5.http.config.ConnectionConfig;
 import org.apache.hc.client5.http.config.RequestConfig;
 import org.apache.hc.client5.http.cookie.BasicCookieStore;
@@ -34,6 +35,7 @@ import org.apache.hc.core5.http.io.entity.EntityUtils;
 import org.apache.hc.core5.http.io.support.ClassicRequestBuilder;
 import org.apache.hc.core5.http.message.MessageSupport;
 import org.apache.hc.core5.http.protocol.HttpContext;
+import org.apache.hc.core5.pool.PoolStats;
 import org.apache.hc.core5.util.TimeValue;
 import org.apache.hc.core5.util.Timeout;
 import org.jsoup.Jsoup;
@@ -76,14 +78,20 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
     }
 
     private final CloseableHttpClient client;
+    private PoolingHttpClientConnectionManager connectionManager;
+
+    public PoolStats getPoolStats() {
+        return connectionManager.getTotalStats();
+    }
 
     private CloseableHttpClient createClient() throws NoSuchAlgorithmException {
         final ConnectionConfig connectionConfig = ConnectionConfig.custom()
                 .setSocketTimeout(10, TimeUnit.SECONDS)
                 .setConnectTimeout(30, TimeUnit.SECONDS)
+                .setValidateAfterInactivity(TimeValue.ofSeconds(5))
                 .build();
 
-        final PoolingHttpClientConnectionManager connectionManager = PoolingHttpClientConnectionManagerBuilder.create()
+        connectionManager = PoolingHttpClientConnectionManagerBuilder.create()
                 .setMaxConnPerRoute(2)
                 .setMaxConnTotal(5000)
                 .setDefaultConnectionConfig(connectionConfig)
@@ -91,15 +99,27 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
                 .build();
 
         connectionManager.setDefaultSocketConfig(SocketConfig.custom()
-                .setSoLinger(TimeValue.ofSeconds(15))
+                .setSoLinger(TimeValue.ofSeconds(-1))
                 .setSoTimeout(Timeout.ofSeconds(10))
                 .build()
         );
 
+        Thread.ofPlatform().daemon(true).start(() -> {
+            try {
+                for (;;) {
+                    TimeUnit.SECONDS.sleep(15);
+                    logger.info("Connection pool stats: {}", connectionManager.getTotalStats());
+                }
+            }
+            catch (InterruptedException e) {
+                Thread.currentThread().interrupt();
+            }
+        });
+
         final RequestConfig defaultRequestConfig = RequestConfig.custom()
                 .setCookieSpec(StandardCookieSpec.RELAXED)
                 .setResponseTimeout(10, TimeUnit.SECONDS)
-                .setConnectionRequestTimeout(8, TimeUnit.SECONDS)
+                .setConnectionRequestTimeout(5, TimeUnit.MINUTES)
                 .build();
 
         return HttpClients.custom()
@@ -287,6 +307,7 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
      * recorded in the WARC file on failure.
      */
     public ContentTypeProbeResult probeContentType(EdgeUrl url,
+                                                   DomainCookies cookies,
                                                    CrawlDelayTimer timer,
                                                    ContentTags tags) {
         if (!tags.isEmpty() || !contentTypeLogic.isUrlLikeBinary(url)) {
@@ -299,9 +320,11 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
                     .addHeader("Accept-Encoding", "gzip")
                     .build();
 
-            var result = SendLock.wrapSend(client, head, (rsp) -> {
-                EntityUtils.consume(rsp.getEntity());
+            cookies.paintRequest(head);
 
+            return SendLock.wrapSend(client, head, (rsp) -> {
+                cookies.updateCookieStore(rsp);
+                EntityUtils.consume(rsp.getEntity());
                 int statusCode = rsp.getCode();
 
                 // Handle redirects
@@ -339,8 +362,6 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
                     return new ContentTypeProbeResult.BadContentType(contentType, statusCode);
                 }
             });
-
-            return result;
         }
         catch (SocketTimeoutException ex) {
 
@@ -362,6 +383,7 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
     @Override
     public HttpFetchResult fetchContent(EdgeUrl url,
                                            WarcRecorder warcRecorder,
+                                           DomainCookies cookies,
                                            CrawlDelayTimer timer,
                                            ContentTags contentTags,
                                            ProbeType probeType)
@@ -369,7 +391,7 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
         try {
             if (probeType == HttpFetcher.ProbeType.FULL) {
                 try {
-                    var probeResult = probeContentType(url, timer, contentTags);
+                    var probeResult = probeContentType(url, cookies, timer, contentTags);
                     logger.info(crawlerAuditMarker, "Probe result {} for {}", probeResult.getClass().getSimpleName(), url);
                     switch (probeResult) {
                         case HttpFetcher.ContentTypeProbeResult.NoOp():
@@ -398,16 +420,16 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
 
             }
 
-            ClassicRequestBuilder getBuilder = ClassicRequestBuilder.get(url.asURI())
-                    .addHeader("User-Agent", userAgentString)
-                    .addHeader("Accept-Encoding", "gzip")
-                    .addHeader("Accept-Language", "en,*;q=0.5")
-                    .addHeader("Accept", "text/html, application/xhtml+xml, text/*;q=0.8");
+            HttpGet request = new HttpGet(url.asURI());
+            request.addHeader("User-Agent", userAgentString);
+            request.addHeader("Accept-Encoding", "gzip");
+            request.addHeader("Accept-Language", "en,*;q=0.5");
+            request.addHeader("Accept", "text/html, application/xhtml+xml, text/*;q=0.8");
 
-            contentTags.paint(getBuilder);
+            contentTags.paint(request);
 
             try (var sl = new SendLock()) {
-                HttpFetchResult result = warcRecorder.fetch(client, getBuilder.build());
+                HttpFetchResult result = warcRecorder.fetch(client, cookies, request);
 
                 if (result instanceof HttpFetchResult.ResultOk ok) {
                     if (ok.statusCode() == 304) {
@@ -494,56 +516,61 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
     }
 
 
-    private SitemapResult fetchSingleSitemap(EdgeUrl sitemapUrl) throws URISyntaxException, IOException, InterruptedException {
-        ClassicHttpRequest getRequest = ClassicRequestBuilder.get(sitemapUrl.asURI())
-                .addHeader("User-Agent", userAgentString)
-                .addHeader("Accept-Encoding", "gzip")
-                .addHeader("Accept", "text/*, */*;q=0.9")
-                .addHeader("User-Agent", userAgentString)
-                .build();
+    private SitemapResult fetchSingleSitemap(EdgeUrl sitemapUrl) throws URISyntaxException {
+        HttpGet getRequest = new HttpGet(sitemapUrl.asURI());
+
+        getRequest.addHeader("User-Agent", userAgentString);
+        getRequest.addHeader("Accept-Encoding", "gzip");
+        getRequest.addHeader("Accept", "text/*, */*;q=0.9");
+        getRequest.addHeader("User-Agent", userAgentString);
 
         try (var sl = new SendLock()) {
             return client.execute(getRequest, response -> {
-                if (response.getCode() != 200) {
-                    return new SitemapResult.SitemapError();
+                try {
+                    if (response.getCode() != 200) {
+                        return new SitemapResult.SitemapError();
+                    }
+
+                    Document parsedSitemap = Jsoup.parse(
+                            EntityUtils.toString(response.getEntity()),
+                            sitemapUrl.toString(),
+                            Parser.xmlParser()
+                    );
+
+                    if (parsedSitemap.childrenSize() == 0) {
+                        return new SitemapResult.SitemapError();
+                    }
+
+                    String rootTagName = parsedSitemap.child(0).tagName();
+
+                    return switch (rootTagName.toLowerCase()) {
+                        case "sitemapindex" -> {
+                            List<String> references = new ArrayList<>();
+                            for (var locTag : parsedSitemap.getElementsByTag("loc")) {
+                                references.add(locTag.text().trim());
+                            }
+                            yield new SitemapResult.SitemapReferences(Collections.unmodifiableList(references));
+                        }
+                        case "urlset" -> {
+                            List<String> urls = new ArrayList<>();
+                            for (var locTag : parsedSitemap.select("url > loc")) {
+                                urls.add(locTag.text().trim());
+                            }
+                            yield new SitemapResult.SitemapUrls(Collections.unmodifiableList(urls));
+                        }
+                        case "rss", "atom" -> {
+                            List<String> urls = new ArrayList<>();
+                            for (var locTag : parsedSitemap.select("link, url")) {
+                                urls.add(locTag.text().trim());
+                            }
+                            yield new SitemapResult.SitemapUrls(Collections.unmodifiableList(urls));
+                        }
+                        default -> new SitemapResult.SitemapError();
+                    };
                 }
-
-                Document parsedSitemap = Jsoup.parse(
-                        EntityUtils.toString(response.getEntity()),
-                        sitemapUrl.toString(),
-                        Parser.xmlParser()
-                );
-
-                if (parsedSitemap.childrenSize() == 0) {
-                    return new SitemapResult.SitemapError();
+                finally {
+                    EntityUtils.consume(response.getEntity());
                 }
-
-                String rootTagName = parsedSitemap.child(0).tagName();
-
-                return switch (rootTagName.toLowerCase()) {
-                    case "sitemapindex" -> {
-                        List<String> references = new ArrayList<>();
-                        for (var locTag : parsedSitemap.getElementsByTag("loc")) {
-                            references.add(locTag.text().trim());
-                        }
-                        yield new SitemapResult.SitemapReferences(Collections.unmodifiableList(references));
-                    }
-                    case "urlset" -> {
-                        List<String> urls = new ArrayList<>();
-                        for (var locTag : parsedSitemap.select("url > loc")) {
-                            urls.add(locTag.text().trim());
-                        }
-                        yield new SitemapResult.SitemapUrls(Collections.unmodifiableList(urls));
-                    }
-                    case "rss", "atom" -> {
-                        List<String> urls = new ArrayList<>();
-                        for (var locTag : parsedSitemap.select("link, url")) {
-                            urls.add(locTag.text().trim());
-                        }
-                        yield new SitemapResult.SitemapUrls(Collections.unmodifiableList(urls));
-                    }
-                    default -> new SitemapResult.SitemapError();
-                };
             });
         }
         catch (Exception ex) {
@@ -574,13 +601,12 @@ public class HttpFetcherImpl implements HttpFetcher, HttpRequestRetryStrategy {
     private Optional<SimpleRobotRules> fetchAndParseRobotsTxt(EdgeUrl url, WarcRecorder recorder) {
         try (var sl = new SendLock()) {
 
-            ClassicHttpRequest request = ClassicRequestBuilder.get(url.asURI())
-                    .addHeader("User-Agent", userAgentString)
-                    .addHeader("Accept-Encoding", "gzip")
-                    .addHeader("Accept", "text/*, */*;q=0.9")
-                    .build();
+            HttpGet request = new HttpGet(url.asURI());
+            request.addHeader("User-Agent", userAgentString);
+            request.addHeader("Accept-Encoding", "gzip");
+            request.addHeader("Accept", "text/*, */*;q=0.9");
 
-            HttpFetchResult result = recorder.fetch(client, request);
+            HttpFetchResult result = recorder.fetch(client, new DomainCookies(), request);
 
             return DocumentBodyExtractor.asBytes(result).mapOpt((contentType, body) ->
                 robotsParser.parseContent(url.toString(),

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

@@ -2,6 +2,7 @@ package nu.marginalia.crawl.fetcher.warc;
 
 import org.apache.commons.io.IOUtils;
 import org.apache.commons.io.input.BOMInputStream;
+import org.apache.hc.client5.http.classic.methods.HttpGet;
 import org.apache.hc.core5.http.ClassicHttpResponse;
 import org.apache.hc.core5.http.Header;
 import org.netpreserve.jwarc.WarcTruncationReason;
@@ -43,7 +44,9 @@ public abstract class WarcInputBuffer implements AutoCloseable {
      *  and suppressed from the headers.
      *  If an error occurs, a buffer will be created with no content and an error status.
      */
-    static WarcInputBuffer forResponse(ClassicHttpResponse response, Duration timeLimit) throws IOException {
+    static WarcInputBuffer forResponse(ClassicHttpResponse response,
+                                       HttpGet request,
+                                       Duration timeLimit) throws IOException {
         if (response == null)
             return new ErrorBuffer();
 
@@ -54,16 +57,29 @@ public abstract class WarcInputBuffer implements AutoCloseable {
             return new ErrorBuffer();
         }
 
-        InputStream is = entity.getContent();
-        long length = entity.getContentLength();
+        InputStream is = null;
+        try {
+            is = entity.getContent();
+            long length = entity.getContentLength();
 
-        try (response) {
             if (length > 0 && length < 8192) {
                 // If the content is small and not compressed, we can just read it into memory
-                return new MemoryBuffer(response.getHeaders(), timeLimit, is, (int) length);
+                return new MemoryBuffer(response.getHeaders(), request, timeLimit, is, (int) length);
             } else {
                 // Otherwise, we unpack it into a file and read it from there
-                return new FileBuffer(response.getHeaders(), timeLimit, is);
+                return new FileBuffer(response.getHeaders(), request, timeLimit, is);
+            }
+        }
+        finally {
+            try {
+                is.skip(Long.MAX_VALUE);
+            }
+            catch (IOException e) {
+                // Ignore the exception
+            }
+            finally {
+                // Close the input stream
+                IOUtils.closeQuietly(is);
             }
         }
 
@@ -71,7 +87,7 @@ public abstract class WarcInputBuffer implements AutoCloseable {
     }
 
     /** Copy an input stream to an output stream, with a maximum size and time limit */
-    protected void copy(InputStream is, OutputStream os, Duration timeLimit) {
+    protected void copy(InputStream is, HttpGet request, OutputStream os, Duration timeLimit) {
         Instant start = Instant.now();
         Instant timeout = start.plus(timeLimit);
         long size = 0;
@@ -86,6 +102,11 @@ public abstract class WarcInputBuffer implements AutoCloseable {
                 Duration remaining = Duration.between(Instant.now(), timeout);
                 if (remaining.isNegative()) {
                     truncationReason = WarcTruncationReason.TIME;
+                    // Abort the request if the time limit is exceeded
+                    // so we don't keep the connection open forever or are forced to consume
+                    // the stream to the end
+
+                    request.abort();
                     break;
                 }
 
@@ -104,6 +125,7 @@ public abstract class WarcInputBuffer implements AutoCloseable {
                 }
                 else if (truncationReason != WarcTruncationReason.LENGTH) {
                     truncationReason = WarcTruncationReason.LENGTH;
+                    break;
                 }
 
             } catch (IOException e) {
@@ -111,13 +133,6 @@ public abstract class WarcInputBuffer implements AutoCloseable {
             }
         }
 
-        // Try to close the connection as long as we haven't timed out.
-        // As per Apache HttpClient's semantics, this will reset the connection
-        // and close the stream if we have timed out.
-
-        if (truncationReason != WarcTruncationReason.TIME) {
-            IOUtils.closeQuietly(is);
-        }
     }
 
     /** Takes a Content-Range header and checks if it is complete.
@@ -218,7 +233,7 @@ class ErrorBuffer extends WarcInputBuffer {
 /** Buffer for when we have the response in memory */
 class MemoryBuffer extends WarcInputBuffer {
     byte[] data;
-    public MemoryBuffer(Header[] headers, Duration timeLimit, InputStream responseStream, int size) {
+    public MemoryBuffer(Header[] headers, HttpGet request, Duration timeLimit, InputStream responseStream, int size) {
         super(suppressContentEncoding(headers));
 
         if (!isRangeComplete(headers)) {
@@ -229,7 +244,7 @@ class MemoryBuffer extends WarcInputBuffer {
 
         var outputStream = new ByteArrayOutputStream(size);
 
-        copy(responseStream, outputStream, timeLimit);
+        copy(responseStream, request, outputStream, timeLimit);
 
         data = outputStream.toByteArray();
     }
@@ -253,7 +268,7 @@ class MemoryBuffer extends WarcInputBuffer {
 class FileBuffer extends WarcInputBuffer {
     private final Path tempFile;
 
-    public FileBuffer(Header[] headers, Duration timeLimit, InputStream responseStream) throws IOException {
+    public FileBuffer(Header[] headers, HttpGet request, Duration timeLimit, InputStream responseStream) throws IOException {
         super(suppressContentEncoding(headers));
 
         if (!isRangeComplete(headers)) {
@@ -265,7 +280,7 @@ class FileBuffer extends WarcInputBuffer {
         this.tempFile = Files.createTempFile("rsp", ".html");
 
         try (var out = Files.newOutputStream(tempFile)) {
-            copy(responseStream, out, timeLimit);
+            copy(responseStream, request, out, timeLimit);
         }
         catch (Exception ex) {
             truncationReason = WarcTruncationReason.UNSPECIFIED;

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

@@ -1,6 +1,7 @@
 package nu.marginalia.crawl.fetcher.warc;
 
 import nu.marginalia.crawl.fetcher.ContentTags;
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.HttpFetcher;
 import nu.marginalia.crawl.fetcher.HttpFetcherImpl;
 import nu.marginalia.link_parser.LinkParser;
@@ -8,9 +9,7 @@ import nu.marginalia.model.EdgeDomain;
 import nu.marginalia.model.EdgeUrl;
 import nu.marginalia.model.body.HttpFetchResult;
 import org.apache.hc.client5.http.classic.HttpClient;
-import org.apache.hc.client5.http.cookie.BasicCookieStore;
-import org.apache.hc.client5.http.cookie.CookieStore;
-import org.apache.hc.core5.http.ClassicHttpRequest;
+import org.apache.hc.client5.http.classic.methods.HttpGet;
 import org.apache.hc.core5.http.NameValuePair;
 import org.jetbrains.annotations.Nullable;
 import org.netpreserve.jwarc.*;
@@ -53,23 +52,15 @@ public class WarcRecorder implements AutoCloseable {
     // Affix a version string in case we need to change the format in the future
     // in some way
     private final String warcRecorderVersion = "1.0";
-    private final CookieStore cookies;
     private final LinkParser linkParser = new LinkParser();
     /**
      * Create a new WarcRecorder that will write to the given file
      *
      * @param warcFile The file to write to
      */
-    public WarcRecorder(Path warcFile, HttpFetcherImpl fetcher) throws IOException {
+    public WarcRecorder(Path warcFile) throws IOException {
         this.warcFile = warcFile;
         this.writer = new WarcWriter(warcFile);
-        this.cookies = fetcher.getCookies();
-    }
-
-    public WarcRecorder(Path warcFile, CookieStore cookies) throws IOException {
-        this.warcFile = warcFile;
-        this.writer = new WarcWriter(warcFile);
-        this.cookies = cookies;
     }
 
     /**
@@ -79,24 +70,21 @@ public class WarcRecorder implements AutoCloseable {
     public WarcRecorder() throws IOException {
         this.warcFile = Files.createTempFile("warc", ".warc.gz");
         this.writer = new WarcWriter(this.warcFile);
-        this.cookies = new BasicCookieStore();
 
         temporaryFile = true;
     }
 
-    private boolean hasCookies() {
-        return !cookies.getCookies().isEmpty();
-    }
-
     public HttpFetchResult fetch(HttpClient client,
-                                 ClassicHttpRequest request)
+                                 DomainCookies cookies,
+                                 HttpGet request)
             throws NoSuchAlgorithmException, IOException, URISyntaxException, InterruptedException
     {
-        return fetch(client, request, Duration.ofMillis(MAX_TIME));
+        return fetch(client, cookies, request, Duration.ofMillis(MAX_TIME));
     }
 
     public HttpFetchResult fetch(HttpClient client,
-                                 ClassicHttpRequest request,
+                                 DomainCookies cookies,
+                                 HttpGet request,
                                  Duration timeout)
             throws NoSuchAlgorithmException, IOException, URISyntaxException, InterruptedException
     {
@@ -113,13 +101,15 @@ public class WarcRecorder implements AutoCloseable {
         // Inject a range header to attempt to limit the size of the response
         // to the maximum size we want to store, if the server supports it.
         request.addHeader("Range", "bytes=0-"+MAX_SIZE);
-
+        cookies.paintRequest(request);
         try {
-            return client.execute(request, response -> {
+            return client.execute(request,response -> {
 
-                try (WarcInputBuffer inputBuffer = WarcInputBuffer.forResponse(response, timeout);
+                try (WarcInputBuffer inputBuffer = WarcInputBuffer.forResponse(response, request, timeout);
                      InputStream inputStream = inputBuffer.read()) {
 
+                    cookies.updateCookieStore(response);
+
                     // Build and write the request
 
                     WarcDigestBuilder requestDigestBuilder = new WarcDigestBuilder();
@@ -143,8 +133,9 @@ public class WarcRecorder implements AutoCloseable {
                     warcRequest.http(); // force HTTP header to be parsed before body is consumed so that caller can use it
                     writer.write(warcRequest);
 
-                    if (hasCookies()) {
-                        extraHeaders.put("X-Has-Cookies", List.of("1"));
+
+                    if (cookies.hasCookies()) {
+                        response.addHeader("X-Has-Cookies", 1);
                     }
 
                     byte[] responseHeaders = WarcProtocolReconstructor.getResponseHeader(response, inputBuffer.size()).getBytes(StandardCharsets.UTF_8);
@@ -259,7 +250,7 @@ public class WarcRecorder implements AutoCloseable {
         writer.write(item);
     }
 
-    private void saveOldResponse(EdgeUrl url, String contentType, int statusCode, byte[] documentBody, @Nullable String headers, ContentTags contentTags) {
+    private void saveOldResponse(EdgeUrl url, DomainCookies domainCookies, String contentType, int statusCode, byte[] documentBody, @Nullable String headers, ContentTags contentTags) {
         try {
             WarcDigestBuilder responseDigestBuilder = new WarcDigestBuilder();
             WarcDigestBuilder payloadDigestBuilder = new WarcDigestBuilder();
@@ -320,7 +311,7 @@ public class WarcRecorder implements AutoCloseable {
                     .date(Instant.now())
                     .body(MediaType.HTTP_RESPONSE, responseDataBuffer.copyBytes());
 
-            if (hasCookies()) {
+            if (domainCookies.hasCookies() || (headers != null && headers.contains("Set-Cookie:"))) {
                 builder.addHeader("X-Has-Cookies", "1");
             }
 
@@ -340,8 +331,8 @@ public class WarcRecorder implements AutoCloseable {
      * an E-Tag or Last-Modified header, and the server responds with a 304 Not Modified.  In this
      * scenario we want to record the data as it was in the previous crawl, but not re-fetch it.
      */
-    public void writeReferenceCopy(EdgeUrl url, String contentType, int statusCode, byte[] documentBody, @Nullable String headers, ContentTags ctags) {
-        saveOldResponse(url, contentType, statusCode, documentBody, headers, ctags);
+    public void writeReferenceCopy(EdgeUrl url, DomainCookies cookies, String contentType, int statusCode, byte[] documentBody, @Nullable String headers, ContentTags ctags) {
+        saveOldResponse(url, cookies, contentType, statusCode, documentBody, headers, ctags);
     }
 
     public void writeWarcinfoHeader(String ip, EdgeDomain domain, HttpFetcherImpl.DomainProbeResult result) throws IOException {

code/processes/crawling-process/java/nu/marginalia/crawl/retreival/CrawlerRetreiver.java

@@ -6,6 +6,7 @@ import nu.marginalia.contenttype.ContentType;
 import nu.marginalia.crawl.CrawlerMain;
 import nu.marginalia.crawl.DomainStateDb;
 import nu.marginalia.crawl.fetcher.ContentTags;
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.HttpFetcher;
 import nu.marginalia.crawl.fetcher.warc.WarcRecorder;
 import nu.marginalia.crawl.logic.LinkFilterSelector;
@@ -51,9 +52,10 @@ public class CrawlerRetreiver implements AutoCloseable {
     private final DomainStateDb domainStateDb;
     private final WarcRecorder warcRecorder;
     private final CrawlerRevisitor crawlerRevisitor;
+    private final DomainCookies cookies = new DomainCookies();
 
     private static final CrawlerConnectionThrottle connectionThrottle = new CrawlerConnectionThrottle(
-            Duration.ofMillis(50) // pace the connections to avoid network congestion at startup
+            Duration.ofSeconds(1) // pace the connections to avoid network congestion at startup
     );
 
     int errorCount = 0;
@@ -124,7 +126,7 @@ public class CrawlerRetreiver implements AutoCloseable {
                     }
 
                     Instant recrawlStart = Instant.now();
-                    CrawlerRevisitor.RecrawlMetadata recrawlMetadata = crawlerRevisitor.recrawl(oldCrawlData, robotsRules, delayTimer);
+                    CrawlerRevisitor.RecrawlMetadata recrawlMetadata = crawlerRevisitor.recrawl(oldCrawlData, cookies, robotsRules, delayTimer);
                     Duration recrawlTime = Duration.between(recrawlStart, Instant.now());
 
                     // Play back the old crawl data (if present) and fetch the documents comparing etags and last-modified
@@ -274,7 +276,7 @@ public class CrawlerRetreiver implements AutoCloseable {
         try {
             var url = rootUrl.withPathAndParam("/", null);
 
-            HttpFetchResult result = fetcher.fetchContent(url, warcRecorder, timer, ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+            HttpFetchResult result = fetcher.fetchContent(url, warcRecorder, cookies, timer, ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
             timer.waitFetchDelay(0);
 
             if (result instanceof HttpFetchResult.ResultRedirect(EdgeUrl location)) {
@@ -337,7 +339,7 @@ public class CrawlerRetreiver implements AutoCloseable {
 
             // Grab the favicon if it exists
 
-            if (fetcher.fetchContent(faviconUrl, warcRecorder, timer, ContentTags.empty(), HttpFetcher.ProbeType.DISABLED) instanceof HttpFetchResult.ResultOk iconResult) {
+            if (fetcher.fetchContent(faviconUrl, warcRecorder, cookies, timer, ContentTags.empty(), HttpFetcher.ProbeType.DISABLED) instanceof HttpFetchResult.ResultOk iconResult) {
                 String contentType = iconResult.header("Content-Type");
                 byte[] iconData = iconResult.getBodyBytes();
 
@@ -407,7 +409,7 @@ public class CrawlerRetreiver implements AutoCloseable {
         if (parsedOpt.isEmpty())
             return false;
 
-        HttpFetchResult result = fetcher.fetchContent(parsedOpt.get(), warcRecorder, timer, ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+        HttpFetchResult result = fetcher.fetchContent(parsedOpt.get(), warcRecorder, cookies, timer, ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
         timer.waitFetchDelay(0);
 
         if (!(result instanceof HttpFetchResult.ResultOk ok)) {
@@ -435,7 +437,7 @@ public class CrawlerRetreiver implements AutoCloseable {
     {
         var contentTags = reference.getContentTags();
 
-        HttpFetchResult fetchedDoc = fetcher.fetchContent(top, warcRecorder, timer, contentTags, HttpFetcher.ProbeType.FULL);
+        HttpFetchResult fetchedDoc = fetcher.fetchContent(top, warcRecorder, cookies, timer, contentTags, HttpFetcher.ProbeType.FULL);
         timer.waitFetchDelay();
 
         if (Thread.interrupted()) {
@@ -461,7 +463,7 @@ public class CrawlerRetreiver implements AutoCloseable {
                 {
                     var doc = reference.doc();
 
-                    warcRecorder.writeReferenceCopy(top, doc.contentType, doc.httpStatus, doc.documentBodyBytes, doc.headers, contentTags);
+                    warcRecorder.writeReferenceCopy(top, cookies, doc.contentType, doc.httpStatus, doc.documentBodyBytes, doc.headers, contentTags);
 
                     fetchedDoc = new HttpFetchResult.Result304ReplacedWithReference(doc.url,
                             new ContentType(doc.contentType, "UTF-8"),

code/processes/crawling-process/java/nu/marginalia/crawl/retreival/revisit/CrawlerRevisitor.java

@@ -2,6 +2,7 @@ package nu.marginalia.crawl.retreival.revisit;
 
 import crawlercommons.robots.SimpleRobotRules;
 import nu.marginalia.crawl.fetcher.ContentTags;
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.warc.WarcRecorder;
 import nu.marginalia.crawl.retreival.CrawlDataReference;
 import nu.marginalia.crawl.retreival.CrawlDelayTimer;
@@ -37,6 +38,7 @@ public class CrawlerRevisitor {
 
     /** Performs a re-crawl of old documents, comparing etags and last-modified */
     public RecrawlMetadata recrawl(CrawlDataReference oldCrawlData,
+                       DomainCookies cookies,
                        SimpleRobotRules robotsRules,
                        CrawlDelayTimer delayTimer)
     throws InterruptedException {
@@ -132,6 +134,7 @@ public class CrawlerRevisitor {
                 }
                 // Add a WARC record so we don't repeat this
                 warcRecorder.writeReferenceCopy(url,
+                        cookies,
                         doc.contentType,
                         doc.httpStatus,
                         doc.documentBodyBytes,

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

@@ -11,6 +11,8 @@ import org.junit.jupiter.api.*;
 import java.io.IOException;
 import java.net.URISyntaxException;
 
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
 @Tag("slow")
 class HttpFetcherImplContentTypeProbeTest {
 
@@ -85,55 +87,59 @@ class HttpFetcherImplContentTypeProbeTest {
 
     @AfterEach
     public void tearDown() throws IOException {
+        var stats = fetcher.getPoolStats();
+        assertEquals(0, stats.getLeased());
+        assertEquals(0, stats.getPending());
+
         fetcher.close();
     }
 
     @Test
     public void testProbeContentTypeHtmlShortcircuitPath() throws URISyntaxException {
-        var result = fetcher.probeContentType(new EdgeUrl("https://localhost/test.html"), new CrawlDelayTimer(50), ContentTags.empty());
-        Assertions.assertInstanceOf(HttpFetcher.ContentTypeProbeResult.Ok.class, result);
+        var result = fetcher.probeContentType(new EdgeUrl("https://localhost/test.html"), new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
+        Assertions.assertInstanceOf(HttpFetcher.ContentTypeProbeResult.NoOp.class, result);
     }
 
 
     @Test
     public void testProbeContentTypeHtmlShortcircuitTags() {
-        var result = fetcher.probeContentType(contentTypeBinaryUrl, new CrawlDelayTimer(50), new ContentTags("a", "b"));
-        Assertions.assertInstanceOf(HttpFetcher.ContentTypeProbeResult.Ok.class, result);
+        var result = fetcher.probeContentType(contentTypeBinaryUrl, new DomainCookies(), new CrawlDelayTimer(50), new ContentTags("a", "b"));
+        Assertions.assertInstanceOf(HttpFetcher.ContentTypeProbeResult.NoOp.class, result);
     }
 
     @Test
     public void testProbeContentTypeHtml() {
-        var result = fetcher.probeContentType(contentTypeHtmlUrl, new CrawlDelayTimer(50), ContentTags.empty());
+        var result = fetcher.probeContentType(contentTypeHtmlUrl, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
         Assertions.assertEquals(new HttpFetcher.ContentTypeProbeResult.Ok(contentTypeHtmlUrl), result);
     }
 
     @Test
     public void testProbeContentTypeBinary() {
-        var result = fetcher.probeContentType(contentTypeBinaryUrl, new CrawlDelayTimer(50), ContentTags.empty());
+        var result = fetcher.probeContentType(contentTypeBinaryUrl, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
         Assertions.assertEquals(new HttpFetcher.ContentTypeProbeResult.BadContentType("application/octet-stream", 200), result);
     }
 
     @Test
     public void testProbeContentTypeRedirect() {
-        var result = fetcher.probeContentType(redirectUrl, new CrawlDelayTimer(50), ContentTags.empty());
+        var result = fetcher.probeContentType(redirectUrl, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
         Assertions.assertEquals(new HttpFetcher.ContentTypeProbeResult.Redirect(contentTypeHtmlUrl), result);
     }
 
     @Test
     public void testProbeContentTypeBadHttpStatus() {
-        var result = fetcher.probeContentType(badHttpStatusUrl, new CrawlDelayTimer(50), ContentTags.empty());
+        var result = fetcher.probeContentType(badHttpStatusUrl, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
         Assertions.assertEquals(new HttpFetcher.ContentTypeProbeResult.HttpError(500, "Bad status code"), result);
     }
 
     @Test
     public void testOnlyGetAllowed() {
-        var result = fetcher.probeContentType(onlyGetAllowedUrl, new CrawlDelayTimer(50), ContentTags.empty());
+        var result = fetcher.probeContentType(onlyGetAllowedUrl, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
         Assertions.assertEquals(new HttpFetcher.ContentTypeProbeResult.Ok(onlyGetAllowedUrl), result);
     }
 
     @Test
     public void testTimeout() {
-        var result = fetcher.probeContentType(timeoutUrl, new CrawlDelayTimer(50), ContentTags.empty());
+        var result = fetcher.probeContentType(timeoutUrl, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
         Assertions.assertInstanceOf(HttpFetcher.ContentTypeProbeResult.Timeout.class, result);
     }
 

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

@@ -12,6 +12,8 @@ import org.junit.jupiter.api.*;
 import java.io.IOException;
 import java.net.URISyntaxException;
 
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
 @Tag("slow")
 class HttpFetcherImplDomainProbeTest {
 
@@ -47,6 +49,10 @@ class HttpFetcherImplDomainProbeTest {
 
     @AfterEach
     public void tearDown() throws IOException {
+        var stats = fetcher.getPoolStats();
+        assertEquals(0, stats.getLeased());
+        assertEquals(0, stats.getPending());
+
         fetcher.close();
     }
 

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

@@ -31,6 +31,7 @@ class HttpFetcherImplFetchTest {
     private static String lastModified = "Wed, 21 Oct 2024 07:28:00 GMT";
 
     private static EdgeUrl okUrl;
+    private static EdgeUrl okUrlSetsCookie;
     private static EdgeUrl okRangeResponseUrl;
     private static EdgeUrl okUrlWith304;
 
@@ -88,6 +89,19 @@ class HttpFetcherImplFetchTest {
                         .withStatus(200)
                         .withBody("Hello World")));
 
+        okUrlSetsCookie = new EdgeUrl("http://localhost:18089/okSetCookie.bin");
+        wireMockServer.stubFor(WireMock.head(WireMock.urlEqualTo(okUrlSetsCookie.path))
+                .willReturn(WireMock.aResponse()
+                        .withHeader("Content-Type", "text/html")
+                        .withHeader("Set-Cookie", "test=1")
+                        .withStatus(200)));
+        wireMockServer.stubFor(WireMock.get(WireMock.urlEqualTo(okUrlSetsCookie.path))
+                .willReturn(WireMock.aResponse()
+                        .withHeader("Content-Type", "text/html")
+                        .withHeader("Set-Cookie", "test=1")
+                        .withStatus(200)
+                        .withBody("Hello World")));
+
         okUrlWith304 = new EdgeUrl("http://localhost:18089/ok304.bin");
         wireMockServer.stubFor(WireMock.head(WireMock.urlEqualTo(okUrlWith304.path))
                 .willReturn(WireMock.aResponse()
@@ -117,6 +131,8 @@ class HttpFetcherImplFetchTest {
                         .withHeader("Keep-Alive", "max=4, timeout=30")
                         .withBody("Hello")
                         ));
+
+
         wireMockServer.start();
 
     }
@@ -134,20 +150,31 @@ class HttpFetcherImplFetchTest {
     public void setUp() throws IOException {
         fetcher = new HttpFetcherImpl(new UserAgent("test.marginalia.nu", "test.marginalia.nu"));
         warcFile = Files.createTempFile(getClass().getSimpleName(), ".warc");
-        warcRecorder = new WarcRecorder(warcFile, fetcher);
+        warcRecorder = new WarcRecorder(warcFile);
     }
 
     @AfterEach
     public void tearDown() throws IOException {
+        var stats = fetcher.getPoolStats();
+        assertEquals(0, stats.getLeased());
+        assertEquals(0, stats.getPending());
+
+        System.out.println(stats);
+
         fetcher.close();
         warcRecorder.close();
         Files.deleteIfExists(warcFile);
     }
 
 
+    @Test
+    public void testFoo() {
+        fetcher.fetchSitemapUrls("https://www.marginalia.nu/sitemap.xml", new CrawlDelayTimer(100));
+    }
+
     @Test
     public void testOk_NoProbe() throws IOException {
-        var result = fetcher.fetchContent(okUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+        var result = fetcher.fetchContent(okUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultOk.class, result);
         Assertions.assertTrue(result.isOk());
@@ -158,12 +185,29 @@ class HttpFetcherImplFetchTest {
         Assertions.assertInstanceOf(WarcResponse.class, warcRecords.get(1));
 
         WarcResponse response = (WarcResponse) warcRecords.get(1);
-        assertEquals("0", response.headers().first("X-Has-Cookies").orElse("0"));
+        assertEquals("0", response.http().headers().first("X-Has-Cookies").orElse("0"));
+    }
+
+    @Test
+    public void testOkSetsCookie() throws IOException {
+        var cookies = new DomainCookies();
+        var result = fetcher.fetchContent(okUrlSetsCookie, warcRecorder, cookies, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+
+        Assertions.assertInstanceOf(HttpFetchResult.ResultOk.class, result);
+        Assertions.assertTrue(result.isOk());
+
+        List<WarcRecord> warcRecords = getWarcRecords();
+        assertEquals(2, warcRecords.size());
+        Assertions.assertInstanceOf(WarcRequest.class, warcRecords.get(0));
+        Assertions.assertInstanceOf(WarcResponse.class, warcRecords.get(1));
+
+        WarcResponse response = (WarcResponse) warcRecords.get(1);
+        assertEquals("1", response.http().headers().first("X-Has-Cookies").orElse("0"));
     }
 
     @Test
     public void testOk_FullProbe() {
-        var result = fetcher.fetchContent(okUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
+        var result = fetcher.fetchContent(okUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultOk.class, result);
         Assertions.assertTrue(result.isOk());
@@ -171,7 +215,7 @@ class HttpFetcherImplFetchTest {
 
     @Test
     public void testOk304_NoProbe() {
-        var result = fetcher.fetchContent(okUrlWith304, warcRecorder, new CrawlDelayTimer(1000), new ContentTags(etag, lastModified), HttpFetcher.ProbeType.DISABLED);
+        var result = fetcher.fetchContent(okUrlWith304, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), new ContentTags(etag, lastModified), HttpFetcher.ProbeType.DISABLED);
 
         Assertions.assertInstanceOf(HttpFetchResult.Result304Raw.class, result);
         System.out.println(result);
@@ -180,7 +224,7 @@ class HttpFetcherImplFetchTest {
 
     @Test
     public void testOk304_FullProbe() {
-        var result = fetcher.fetchContent(okUrlWith304, warcRecorder, new CrawlDelayTimer(1000), new ContentTags(etag, lastModified), HttpFetcher.ProbeType.FULL);
+        var result = fetcher.fetchContent(okUrlWith304, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), new ContentTags(etag, lastModified), HttpFetcher.ProbeType.FULL);
 
         Assertions.assertInstanceOf(HttpFetchResult.Result304Raw.class, result);
         System.out.println(result);
@@ -188,7 +232,7 @@ class HttpFetcherImplFetchTest {
 
     @Test
     public void testBadStatus_NoProbe() throws IOException {
-        var result = fetcher.fetchContent(badHttpStatusUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+        var result = fetcher.fetchContent(badHttpStatusUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultOk.class, result);
         Assertions.assertFalse(result.isOk());
@@ -202,7 +246,7 @@ class HttpFetcherImplFetchTest {
 
     @Test
     public void testBadStatus_FullProbe() {
-        var result = fetcher.fetchContent(badHttpStatusUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
+        var result = fetcher.fetchContent(badHttpStatusUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultOk.class, result);
         Assertions.assertFalse(result.isOk());
@@ -212,7 +256,7 @@ class HttpFetcherImplFetchTest {
 
     @Test
     public void testRedirect_NoProbe() throws URISyntaxException, IOException {
-        var result = fetcher.fetchContent(redirectUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+        var result = fetcher.fetchContent(redirectUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultRedirect.class, result);
         assertEquals(new EdgeUrl("http://localhost:18089/test.html.bin"), ((HttpFetchResult.ResultRedirect) result).url());
@@ -225,7 +269,7 @@ class HttpFetcherImplFetchTest {
 
     @Test
     public void testRedirect_FullProbe() throws URISyntaxException {
-        var result = fetcher.fetchContent(redirectUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
+        var result = fetcher.fetchContent(redirectUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultRedirect.class, result);
         assertEquals(new EdgeUrl("http://localhost:18089/test.html.bin"), ((HttpFetchResult.ResultRedirect) result).url());
@@ -238,7 +282,7 @@ class HttpFetcherImplFetchTest {
     public void testFetchTimeout_NoProbe() throws IOException, URISyntaxException {
         Instant requestStart = Instant.now();
 
-        var result = fetcher.fetchContent(timeoutUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+        var result = fetcher.fetchContent(timeoutUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultException.class, result);
 
@@ -262,7 +306,7 @@ class HttpFetcherImplFetchTest {
 
     @Test
     public void testRangeResponse() throws IOException {
-        var result = fetcher.fetchContent(okRangeResponseUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+        var result = fetcher.fetchContent(okRangeResponseUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultOk.class, result);
         Assertions.assertTrue(result.isOk());
@@ -279,7 +323,7 @@ class HttpFetcherImplFetchTest {
     @Test
     public void testFetchTimeout_Probe() throws IOException, URISyntaxException {
         Instant requestStart = Instant.now();
-        var result = fetcher.fetchContent(timeoutUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
+        var result = fetcher.fetchContent(timeoutUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
         Instant requestEnd = Instant.now();
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultException.class, result);
@@ -302,7 +346,7 @@ class HttpFetcherImplFetchTest {
     @Test
     public void testKeepaliveUrl() {
         // mostly for smoke testing and debugger utility
-        var result = fetcher.fetchContent(keepAliveUrl, warcRecorder, new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
+        var result = fetcher.fetchContent(keepAliveUrl, warcRecorder, new DomainCookies(), new CrawlDelayTimer(1000), ContentTags.empty(), HttpFetcher.ProbeType.DISABLED);
 
         Assertions.assertInstanceOf(HttpFetchResult.ResultOk.class, result);
         Assertions.assertTrue(result.isOk());
@@ -319,6 +363,13 @@ class HttpFetcherImplFetchTest {
             WarcXEntityRefused.register(reader);
 
             for (var record : reader) {
+                // Load the body, we need to do this before we close the reader to have access to the content.
+                if (record instanceof WarcRequest req) {
+                    req.http();
+                } else if (record instanceof WarcResponse rsp) {
+                    rsp.http();
+                }
+
                 records.add(record);
             }
         }

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

@@ -1,12 +1,12 @@
 package nu.marginalia.crawl.retreival;
 
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.warc.WarcRecorder;
 import nu.marginalia.model.EdgeDomain;
 import nu.marginalia.model.EdgeUrl;
 import org.apache.hc.client5.http.classic.HttpClient;
-import org.apache.hc.client5.http.cookie.BasicCookieStore;
+import org.apache.hc.client5.http.classic.methods.HttpGet;
 import org.apache.hc.client5.http.impl.classic.HttpClients;
-import org.apache.hc.core5.http.io.support.ClassicRequestBuilder;
 import org.junit.jupiter.api.AfterEach;
 import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.Test;
@@ -45,7 +45,7 @@ class CrawlerWarcResynchronizerTest {
 
     @Test
     void run() throws IOException, URISyntaxException {
-        try (var oldRecorder = new WarcRecorder(fileName, new BasicCookieStore())) {
+        try (var oldRecorder = new WarcRecorder(fileName)) {
             fetchUrl(oldRecorder, "https://www.marginalia.nu/");
             fetchUrl(oldRecorder, "https://www.marginalia.nu/log/");
             fetchUrl(oldRecorder, "https://www.marginalia.nu/feed/");
@@ -55,7 +55,7 @@ class CrawlerWarcResynchronizerTest {
 
         var crawlFrontier = new DomainCrawlFrontier(new EdgeDomain("www.marginalia.nu"), List.of(), 100);
 
-        try (var newRecorder = new WarcRecorder(outputFile, new BasicCookieStore())) {
+        try (var newRecorder = new WarcRecorder(outputFile)) {
             new CrawlerWarcResynchronizer(crawlFrontier, newRecorder).run(fileName);
         }
 
@@ -78,10 +78,10 @@ class CrawlerWarcResynchronizerTest {
     }
 
     void fetchUrl(WarcRecorder recorder, String url) throws NoSuchAlgorithmException, IOException, URISyntaxException, InterruptedException {
-        var req = ClassicRequestBuilder.get(new java.net.URI(url))
-                .addHeader("User-agent", "test.marginalia.nu")
-                .addHeader("Accept-Encoding", "gzip")
-                .build();
-        recorder.fetch(httpClient, req);
+        HttpGet request = new HttpGet(url);
+        request.addHeader("User-agent", "test.marginalia.nu");
+        request.addHeader("Accept-Encoding", "gzip");
+
+        recorder.fetch(httpClient, new DomainCookies(), request);
     }
 }

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

@@ -2,6 +2,7 @@ package nu.marginalia.crawl.retreival.fetcher;
 
 import com.sun.net.httpserver.HttpServer;
 import nu.marginalia.crawl.fetcher.ContentTags;
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.HttpFetcher;
 import nu.marginalia.crawl.fetcher.HttpFetcherImpl;
 import nu.marginalia.crawl.retreival.CrawlDelayTimer;
@@ -88,7 +89,7 @@ class ContentTypeProberTest {
 
     @Test
     void probeContentTypeOk() throws Exception {
-        HttpFetcher.ContentTypeProbeResult result = fetcher.probeContentType(htmlEndpoint, new CrawlDelayTimer(50), ContentTags.empty());
+        HttpFetcher.ContentTypeProbeResult result = fetcher.probeContentType(htmlEndpoint, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
 
         System.out.println(result);
 
@@ -97,7 +98,7 @@ class ContentTypeProberTest {
 
     @Test
     void probeContentTypeRedir() throws Exception {
-        HttpFetcher.ContentTypeProbeResult result = fetcher.probeContentType(htmlRedirEndpoint, new CrawlDelayTimer(50), ContentTags.empty());
+        HttpFetcher.ContentTypeProbeResult result = fetcher.probeContentType(htmlRedirEndpoint, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
 
         System.out.println(result);
 
@@ -106,7 +107,7 @@ class ContentTypeProberTest {
 
     @Test
     void probeContentTypeBad() throws Exception {
-        HttpFetcher.ContentTypeProbeResult result = fetcher.probeContentType(binaryEndpoint, new CrawlDelayTimer(50), ContentTags.empty());
+        HttpFetcher.ContentTypeProbeResult result = fetcher.probeContentType(binaryEndpoint, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
 
         System.out.println(result);
 
@@ -115,7 +116,7 @@ class ContentTypeProberTest {
 
     @Test
     void probeContentTypeTimeout() throws Exception {
-        HttpFetcher.ContentTypeProbeResult result = fetcher.probeContentType(timeoutEndpoint, new CrawlDelayTimer(50), ContentTags.empty());
+        HttpFetcher.ContentTypeProbeResult result = fetcher.probeContentType(timeoutEndpoint, new DomainCookies(), new CrawlDelayTimer(50), ContentTags.empty());
 
         System.out.println(result);
 

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

@@ -1,11 +1,11 @@
 package nu.marginalia.crawl.retreival.fetcher;
 
 import com.sun.net.httpserver.HttpServer;
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.warc.WarcRecorder;
 import org.apache.hc.client5.http.classic.HttpClient;
-import org.apache.hc.client5.http.cookie.BasicCookieStore;
+import org.apache.hc.client5.http.classic.methods.HttpGet;
 import org.apache.hc.client5.http.impl.classic.HttpClients;
-import org.apache.hc.core5.http.io.support.ClassicRequestBuilder;
 import org.junit.jupiter.api.*;
 import org.netpreserve.jwarc.WarcReader;
 import org.netpreserve.jwarc.WarcRequest;
@@ -51,14 +51,14 @@ class WarcRecorderFakeServerTest {
                 os.write("<html><body>hello</body></html>".getBytes());
                 os.flush();
                 try {
-                    TimeUnit.SECONDS.sleep(1);
+                    TimeUnit.SECONDS.sleep(2);
                 } catch (InterruptedException e) {
                     throw new RuntimeException(e);
                 }
                 os.write(":".getBytes());
                 os.flush();
                 try {
-                    TimeUnit.SECONDS.sleep(1);
+                    TimeUnit.SECONDS.sleep(2);
                 } catch (InterruptedException e) {
                     throw new RuntimeException(e);
                 }
@@ -89,24 +89,22 @@ class WarcRecorderFakeServerTest {
         fileNameWarc = Files.createTempFile("test", ".warc");
         fileNameParquet = Files.createTempFile("test", ".parquet");
 
-        client = new WarcRecorder(fileNameWarc, new BasicCookieStore());
+        client = new WarcRecorder(fileNameWarc);
     }
 
     @AfterEach
     public void tearDown() throws Exception {
+
         client.close();
         Files.delete(fileNameWarc);
     }
 
     @Test
     public void fetchFast() throws Exception {
-        client.fetch(httpClient,
-                ClassicRequestBuilder
-                        .get(new java.net.URI("http://localhost:14510/fast"))
-                        .addHeader("User-agent", "test.marginalia.nu")
-                        .addHeader("Accept-Encoding", "gzip")
-                        .build()
-        );
+        HttpGet request = new HttpGet("http://localhost:14510/fast");
+        request.addHeader("User-agent", "test.marginalia.nu");
+        request.addHeader("Accept-Encoding", "gzip");
+        client.fetch(httpClient, new DomainCookies(), request);
 
         Map<String, String> sampleData = new HashMap<>();
         try (var warcReader = new WarcReader(fileNameWarc)) {
@@ -127,11 +125,13 @@ class WarcRecorderFakeServerTest {
     public void fetchSlow() throws Exception {
         Instant start = Instant.now();
 
+        HttpGet request = new HttpGet("http://localhost:14510/slow");
+        request.addHeader("User-agent", "test.marginalia.nu");
+        request.addHeader("Accept-Encoding", "gzip");
+
         client.fetch(httpClient,
-                ClassicRequestBuilder.get(new java.net.URI("http://localhost:14510/slow"))
-                        .addHeader("User-agent", "test.marginalia.nu")
-                        .addHeader("Accept-Encoding", "gzip")
-                        .build(),
+                new DomainCookies(),
+                request,
                 Duration.ofSeconds(1)
         );
         Instant end = Instant.now();
@@ -149,6 +149,8 @@ class WarcRecorderFakeServerTest {
             });
         }
 
+        System.out.println(
+                Files.readString(fileNameWarc));
         System.out.println(sampleData);
 
         // Timeout is set to 1 second, but the server will take 5 seconds to respond,

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

@@ -2,14 +2,14 @@ package nu.marginalia.crawl.retreival.fetcher;
 
 import nu.marginalia.UserAgent;
 import nu.marginalia.crawl.fetcher.ContentTags;
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.warc.WarcRecorder;
 import nu.marginalia.model.EdgeUrl;
 import nu.marginalia.parquet.crawldata.CrawledDocumentParquetRecordFileReader;
 import nu.marginalia.parquet.crawldata.CrawledDocumentParquetRecordFileWriter;
 import org.apache.hc.client5.http.classic.HttpClient;
-import org.apache.hc.client5.http.cookie.BasicCookieStore;
+import org.apache.hc.client5.http.classic.methods.HttpGet;
 import org.apache.hc.client5.http.impl.classic.HttpClients;
-import org.apache.hc.core5.http.io.support.ClassicRequestBuilder;
 import org.junit.jupiter.api.AfterEach;
 import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.Test;
@@ -41,7 +41,7 @@ class WarcRecorderTest {
         fileNameWarc = Files.createTempFile("test", ".warc");
         fileNameParquet = Files.createTempFile("test", ".parquet");
 
-        client = new WarcRecorder(fileNameWarc, new BasicCookieStore());
+        client = new WarcRecorder(fileNameWarc);
     }
 
     @AfterEach
@@ -52,12 +52,12 @@ class WarcRecorderTest {
 
     @Test
     void fetch() throws NoSuchAlgorithmException, IOException, URISyntaxException, InterruptedException {
-        client.fetch(httpClient,
-                ClassicRequestBuilder.get(new java.net.URI("https://www.marginalia.nu/"))
-                        .addHeader("User-agent", "test.marginalia.nu")
-                        .addHeader("Accept-Encoding", "gzip")
-                        .build()
-        );
+
+        HttpGet request = new HttpGet("https://www.marginalia.nu/");
+        request.addHeader("User-agent", "test.marginalia.nu");
+        request.addHeader("Accept-Encoding", "gzip");
+
+        client.fetch(httpClient, new DomainCookies(), request);
 
         Map<String, String> sampleData = new HashMap<>();
         try (var warcReader = new WarcReader(fileNameWarc)) {
@@ -78,8 +78,9 @@ class WarcRecorderTest {
     @Test
     public void flagAsSkipped() throws IOException, URISyntaxException {
 
-        try (var recorder = new WarcRecorder(fileNameWarc, new BasicCookieStore())) {
+        try (var recorder = new WarcRecorder(fileNameWarc)) {
             recorder.writeReferenceCopy(new EdgeUrl("https://www.marginalia.nu/"),
+                    new DomainCookies(),
                     "text/html",
                     200,
                     "<?doctype html><html><body>test</body></html>".getBytes(),
@@ -102,8 +103,9 @@ class WarcRecorderTest {
     @Test
     public void flagAsSkippedNullBody() throws IOException, URISyntaxException {
 
-        try (var recorder = new WarcRecorder(fileNameWarc, new BasicCookieStore())) {
+        try (var recorder = new WarcRecorder(fileNameWarc)) {
             recorder.writeReferenceCopy(new EdgeUrl("https://www.marginalia.nu/"),
+                    new DomainCookies(),
                     "text/html",
                     200,
                     null,
@@ -114,8 +116,9 @@ class WarcRecorderTest {
 
     @Test
     public void testSaveImport() throws URISyntaxException, IOException {
-        try (var recorder = new WarcRecorder(fileNameWarc, new BasicCookieStore())) {
+        try (var recorder = new WarcRecorder(fileNameWarc)) {
             recorder.writeReferenceCopy(new EdgeUrl("https://www.marginalia.nu/"),
+                    new DomainCookies(),
                     "text/html",
                     200,
                     "<?doctype html><html><body>test</body></html>".getBytes(),
@@ -138,23 +141,23 @@ class WarcRecorderTest {
 
     @Test
     public void testConvertToParquet() throws NoSuchAlgorithmException, IOException, URISyntaxException, InterruptedException {
-        client.fetch(httpClient, ClassicRequestBuilder
-                .get(new java.net.URI("https://www.marginalia.nu/"))
-                .addHeader("User-agent", "test.marginalia.nu")
-                .addHeader("Accept-Encoding", "gzip")
-                .build());
+        HttpGet request1 = new HttpGet("https://www.marginalia.nu/");
+        request1.addHeader("User-agent", "test.marginalia.nu");
+        request1.addHeader("Accept-Encoding", "gzip");
 
-        client.fetch(httpClient, ClassicRequestBuilder
-                .get(new java.net.URI("https://www.marginalia.nu/log/"))
-                .addHeader("User-agent", "test.marginalia.nu")
-                .addHeader("Accept-Encoding", "gzip")
-                .build());
+        client.fetch(httpClient, new DomainCookies(), request1);
 
-        client.fetch(httpClient, ClassicRequestBuilder
-                .get(new java.net.URI("https://www.marginalia.nu/sanic.png"))
-                .addHeader("User-agent", "test.marginalia.nu")
-                .addHeader("Accept-Encoding", "gzip")
-                .build());
+        HttpGet request2 = new HttpGet("https://www.marginalia.nu/log/");
+        request2.addHeader("User-agent", "test.marginalia.nu");
+        request2.addHeader("Accept-Encoding", "gzip");
+
+        client.fetch(httpClient, new DomainCookies(), request2);
+
+        HttpGet request3 = new HttpGet("https://www.marginalia.nu/sanic.png");
+        request3.addHeader("User-agent", "test.marginalia.nu");
+        request3.addHeader("Accept-Encoding", "gzip");
+
+        client.fetch(httpClient, new DomainCookies(), request3);
 
         CrawledDocumentParquetRecordFileWriter.convertWarc(
                 "www.marginalia.nu",

code/processes/crawling-process/test/nu/marginalia/crawling/HttpFetcherTest.java

@@ -1,6 +1,7 @@
 package nu.marginalia.crawling;
 
 import nu.marginalia.crawl.fetcher.ContentTags;
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.HttpFetcher;
 import nu.marginalia.crawl.fetcher.HttpFetcherImpl;
 import nu.marginalia.crawl.fetcher.warc.WarcRecorder;
@@ -31,7 +32,7 @@ class HttpFetcherTest {
     void fetchUTF8() throws Exception {
         var fetcher = new HttpFetcherImpl("nu.marginalia.edge-crawler");
         try (var recorder = new WarcRecorder()) {
-            var result = fetcher.fetchContent(new EdgeUrl("https://www.marginalia.nu"), recorder, new CrawlDelayTimer(100), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
+            var result = fetcher.fetchContent(new EdgeUrl("https://www.marginalia.nu"), recorder, new DomainCookies(), new CrawlDelayTimer(100), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
             if (DocumentBodyExtractor.asString(result) instanceof DocumentBodyResult.Ok bodyOk) {
                 System.out.println(bodyOk.contentType());
             }
@@ -49,7 +50,7 @@ class HttpFetcherTest {
         var fetcher = new HttpFetcherImpl("nu.marginalia.edge-crawler");
 
         try (var recorder = new WarcRecorder()) {
-            var result = fetcher.fetchContent(new EdgeUrl("https://www.marginalia.nu/robots.txt"), recorder, new CrawlDelayTimer(100), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
+            var result = fetcher.fetchContent(new EdgeUrl("https://www.marginalia.nu/robots.txt"), recorder, new DomainCookies(), new CrawlDelayTimer(100), ContentTags.empty(), HttpFetcher.ProbeType.FULL);
             if (DocumentBodyExtractor.asString(result) instanceof DocumentBodyResult.Ok bodyOk) {
                 System.out.println(bodyOk.contentType());
             }

code/processes/crawling-process/test/nu/marginalia/crawling/retreival/CrawlerMockFetcherTest.java

@@ -3,10 +3,7 @@ package nu.marginalia.crawling.retreival;
 import crawlercommons.robots.SimpleRobotRules;
 import nu.marginalia.crawl.CrawlerMain;
 import nu.marginalia.crawl.DomainStateDb;
-import nu.marginalia.crawl.fetcher.ContentTags;
-import nu.marginalia.crawl.fetcher.HttpFetcher;
-import nu.marginalia.crawl.fetcher.HttpFetcherImpl;
-import nu.marginalia.crawl.fetcher.SitemapRetriever;
+import nu.marginalia.crawl.fetcher.*;
 import nu.marginalia.crawl.fetcher.warc.WarcRecorder;
 import nu.marginalia.crawl.retreival.CrawlDelayTimer;
 import nu.marginalia.crawl.retreival.CrawlerRetreiver;
@@ -137,7 +134,7 @@ public class CrawlerMockFetcherTest {
         }
 
         @Override
-        public HttpFetchResult fetchContent(EdgeUrl url, WarcRecorder recorder, CrawlDelayTimer timer, ContentTags tags, ProbeType probeType) {
+        public HttpFetchResult fetchContent(EdgeUrl url, WarcRecorder recorder, DomainCookies cookies, CrawlDelayTimer timer, ContentTags tags, ProbeType probeType) {
             logger.info("Fetching {}", url);
             if (mockData.containsKey(url)) {
                 byte[] bodyBytes = mockData.get(url).documentBodyBytes;

code/processes/crawling-process/test/nu/marginalia/crawling/retreival/CrawlerRetreiverTest.java

@@ -16,7 +16,6 @@ import nu.marginalia.model.crawldata.CrawledDocument;
 import nu.marginalia.model.crawldata.CrawledDomain;
 import nu.marginalia.model.crawldata.SerializableCrawlData;
 import nu.marginalia.slop.SlopCrawlDataRecord;
-import org.apache.hc.client5.http.cookie.BasicCookieStore;
 import org.jetbrains.annotations.NotNull;
 import org.junit.jupiter.api.*;
 import org.netpreserve.jwarc.*;
@@ -180,7 +179,7 @@ class CrawlerRetreiverTest {
                 new EdgeDomain("www.marginalia.nu"),
                 List.of(), 100);
         var resync = new CrawlerWarcResynchronizer(revisitCrawlFrontier,
-                new WarcRecorder(tempFileWarc2, new BasicCookieStore())
+                new WarcRecorder(tempFileWarc2)
         );
 
         // truncate the size of the file to simulate a crash
@@ -456,7 +455,7 @@ class CrawlerRetreiverTest {
                 List.of(), 100);
 
         var resync = new CrawlerWarcResynchronizer(revisitCrawlFrontier,
-                new WarcRecorder(tempFileWarc3, new BasicCookieStore())
+                new WarcRecorder(tempFileWarc3)
         );
 
         // truncate the size of the file to simulate a crash
@@ -507,7 +506,7 @@ class CrawlerRetreiverTest {
     }
 
     private void doCrawlWithReferenceStream(CrawlerMain.CrawlSpecRecord specs, CrawlDataReference reference) {
-        try (var recorder = new WarcRecorder(tempFileWarc2, new BasicCookieStore());
+        try (var recorder = new WarcRecorder(tempFileWarc2);
              var db = new DomainStateDb(tempFileDb)
         ) {
             new CrawlerRetreiver(httpFetcher, new DomainProber(d -> true), specs, db, recorder).crawlDomain(new DomainLinks(), reference);
@@ -519,7 +518,7 @@ class CrawlerRetreiverTest {
 
     @NotNull
     private DomainCrawlFrontier doCrawl(Path tempFileWarc1, CrawlerMain.CrawlSpecRecord specs) {
-        try (var recorder = new WarcRecorder(tempFileWarc1, new BasicCookieStore());
+        try (var recorder = new WarcRecorder(tempFileWarc1);
              var db = new DomainStateDb(tempFileDb)
         ) {
             var crawler = new CrawlerRetreiver(httpFetcher, new DomainProber(d -> true), specs, db, recorder);

code/tools/integration-test/test/nu/marginalia/IntegrationTest.java

@@ -10,6 +10,7 @@ import nu.marginalia.api.searchquery.model.results.PrototypeRankingParameters;
 import nu.marginalia.converting.processor.DomainProcessor;
 import nu.marginalia.converting.writer.ConverterBatchWriter;
 import nu.marginalia.crawl.fetcher.ContentTags;
+import nu.marginalia.crawl.fetcher.DomainCookies;
 import nu.marginalia.crawl.fetcher.HttpFetcherImpl;
 import nu.marginalia.crawl.fetcher.warc.WarcRecorder;
 import nu.marginalia.functions.searchquery.QueryFactory;
@@ -43,7 +44,6 @@ import nu.marginalia.process.control.FakeProcessHeartbeat;
 import nu.marginalia.storage.FileStorageService;
 import nu.marginalia.test.IntegrationTestModule;
 import nu.marginalia.test.TestUtil;
-import org.apache.hc.client5.http.cookie.BasicCookieStore;
 import org.junit.jupiter.api.AfterEach;
 import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.Test;
@@ -121,11 +121,12 @@ public class IntegrationTest {
     public void run() throws Exception {
 
         /** CREATE WARC */
-        try (WarcRecorder warcRecorder = new WarcRecorder(warcData, new BasicCookieStore())) {
+        try (WarcRecorder warcRecorder = new WarcRecorder(warcData)) {
             warcRecorder.writeWarcinfoHeader("127.0.0.1", new EdgeDomain("www.example.com"),
                     new HttpFetcherImpl.DomainProbeResult.Ok(new EdgeUrl("https://www.example.com/")));
 
             warcRecorder.writeReferenceCopy(new EdgeUrl("https://www.example.com/"),
+                    new DomainCookies(),
                     "text/html", 200,
                     """
                             <html>