From e76c8689e9778f4635b5aa81c643bd57a61dd3b9 Mon Sep 17 00:00:00 2001
From: cube <cube@github.com>
Date: Thu, 19 Dec 2024 00:06:16 +0100
Subject: [PATCH] Add timeout for Tika client as suggested in
 https://github.com/paperless-ngx/paperless-ngx/discussions/8509

---
 .github/workflows/ci.yml      |  1 +
 src/paperless_tika/parsers.py | 10 ++++++++--
 2 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 8bd31467b..401fffb15 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -12,6 +12,7 @@ on:
   pull_request:
     branches-ignore:
       - 'translations**'
+  workflow_dispatch:
 
 env:
   # This is the version of pipenv all the steps will use
diff --git a/src/paperless_tika/parsers.py b/src/paperless_tika/parsers.py
index f51a03916..40aa8e581 100644
--- a/src/paperless_tika/parsers.py
+++ b/src/paperless_tika/parsers.py
@@ -33,7 +33,10 @@ class TikaDocumentParser(DocumentParser):
 
     def extract_metadata(self, document_path, mime_type):
         try:
-            with TikaClient(tika_url=settings.TIKA_ENDPOINT) as client:
+            with TikaClient(
+                tika_url=settings.TIKA_ENDPOINT,
+                timeout=settings.CELERY_TASK_TIME_LIMIT,
+            ) as client:
                 parsed = client.metadata.from_file(document_path, mime_type)
                 return [
                     {
@@ -54,7 +57,10 @@ class TikaDocumentParser(DocumentParser):
         self.log.info(f"Sending {document_path} to Tika server")
 
         try:
-            with TikaClient(tika_url=settings.TIKA_ENDPOINT) as client:
+            with TikaClient(
+                tika_url=settings.TIKA_ENDPOINT,
+                timeout=settings.CELERY_TASK_TIME_LIMIT,
+            ) as client:
                 try:
                     parsed = client.tika.as_text.from_file(document_path, mime_type)
                 except httpx.HTTPStatusError as err: