(nutch) branch master updated: NUTCH-3063 Support for "addBinaryContent" from REST API

[snagel]

This is an automated email from the ASF dual-hosted git repository.

snagel pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/nutch.git


The following commit(s) were added to refs/heads/master by this push:
     new ac03cf164 NUTCH-3063 Support for "addBinaryContent" from REST API
ac03cf164 is described below

commit ac03cf1646f5af152daeb9f0bef3fec2b51739c2
Author: Sebastian Nagel <sna...@apache.org>
AuthorDate: Fri Sep 6 21:59:51 2024 +0200

    NUTCH-3063 Support for "addBinaryContent" from REST API
    
    Apply patch contributed by Isabelle Giguere.
---
 src/java/org/apache/nutch/indexer/IndexingJob.java | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/src/java/org/apache/nutch/indexer/IndexingJob.java 
b/src/java/org/apache/nutch/indexer/IndexingJob.java
index c3ddb4ae9..fc2c44a06 100644
--- a/src/java/org/apache/nutch/indexer/IndexingJob.java
+++ b/src/java/org/apache/nutch/indexer/IndexingJob.java
@@ -312,6 +312,8 @@ public class IndexingJob extends NutchTool implements Tool {
     boolean filter = false;
     boolean normalize = false;
     boolean isSegment = false;
+    boolean addBinaryContent = false;
+    boolean base64 = false;
     String params= null;
     Configuration conf = getConf();
 
@@ -405,12 +407,18 @@ public class IndexingJob extends NutchTool implements 
Tool {
     if(args.containsKey("filter")){
       filter = true;
     }
+    if (args.containsKey("addBinaryContent")) {
+      addBinaryContent = true;
+      if (args.containsKey("base64")) {
+          base64 = true;
+      }
+    }
     if(args.containsKey("params")){
       params = (String)args.get("params");
     }
     setConf(conf);
     index(crawlDb, linkdb, segments, noCommit, deleteGone, params, filter,
-        normalize);
+        normalize, addBinaryContent, base64);
     Map<String, Object> results = new HashMap<>();
     results.put(Nutch.VAL_RESULT, 0);
     return results;