timeout message implementation for GROBID and ingest workers

2 files changed, 18 insertions, 0 deletions

diff --git a/python/sandcrawler/grobid.py b/python/sandcrawler/grobid.py
index 08e3a96..f329a73 100644
--- a/python/sandcrawler/grobid.py
+++ b/python/sandcrawler/grobid.py
@@ -87,6 +87,15 @@ class GrobidWorker(SandcrawlerWorker):
         self.sink = sink
         self.consolidate_mode = 2
 
+    def timeout_response(self, task):
+        default_key = task['sha1hex']
+        return dict(
+            status="error-timeout",
+            error_msg="internal GROBID worker timeout",
+            source=task,
+            key=default_key,
+        )
+
     def process(self, record):
         default_key = record['sha1hex']
         if record.get('warc_path') and record.get('warc_offset'):
diff --git a/python/sandcrawler/ingest.py b/python/sandcrawler/ingest.py
index 5cb3ef8..0be7653 100644
--- a/python/sandcrawler/ingest.py
+++ b/python/sandcrawler/ingest.py
@@ -229,6 +229,15 @@ class IngestFileWorker(SandcrawlerWorker):
         result.pop('key', None)
         return result
 
+    def timeout_response(self, task):
+        print("[TIMEOUT]", file=sys.stderr)
+        return dict(
+            request=task,
+            hit=False,
+            status="timeout",
+            error_message="ingest worker internal timeout",
+        )
+
     def process(self, request):
 
         # backwards compatibility