Write warcinfo records to the redirect WARC file too

nla · Aug 6, 2024 · 04ca3e6 · 04ca3e6
1 parent 6f72765
commit 04ca3e6
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 1 deletion.
diff --git a/src/au/gov/nla/httrack2warc/Httrack2Warc.java b/src/au/gov/nla/httrack2warc/Httrack2Warc.java
@@ -158,6 +158,10 @@ public void convertDirectory(Path sourceDirectory) throws IOException {
             Set<String> processedFiles = new HashSet<>();
             LinkRewriter linkRewriter = rewriteLinks ? new LinkRewriter(crawl) : null;
 
+            if (redirectWriter.warc != warc) {
+                redirectWriter.warc.writeWarcinfoRecord(UUID.randomUUID(), launchInstant, warcInfo);
+            }
+
             crawl.forEach(record -> {
                 if (isUrlExcluded(record.getUrl())) {
                     log.info("Excluded {}", record.getUrl());

diff --git a/src/au/gov/nla/httrack2warc/RedirectWriter.java b/src/au/gov/nla/httrack2warc/RedirectWriter.java
@@ -13,7 +13,7 @@
  */
 public class RedirectWriter implements Closeable {
     private final String prefix;
-    private final WarcWriter warc;
+    final WarcWriter warc;
 
     public RedirectWriter(String prefix, WarcWriter warc) {
         this.prefix = prefix;