From aa3d91a338a66b39774acdfdc092be5f1672ff61 Mon Sep 17 00:00:00 2001
From: jonaswinkler <17569239+jonaswinkler@users.noreply.github.com>
Date: Mon, 22 Feb 2021 11:38:16 +0100
Subject: [PATCH] associate error messages with documents

---
 .../management/commands/document_archiver.py  | 20 ++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)

diff --git a/src/documents/management/commands/document_archiver.py b/src/documents/management/commands/document_archiver.py
index 6a6056ed7..497e8fe7e 100644
--- a/src/documents/management/commands/document_archiver.py
+++ b/src/documents/management/commands/document_archiver.py
@@ -4,7 +4,9 @@ import multiprocessing
 import logging
 import os
 import shutil
+import sys
 import uuid
+from io import TextIOBase
 
 import tqdm
 from django import db
@@ -12,7 +14,6 @@ from django.conf import settings
 from django.core.management.base import BaseCommand
 from django.db import transaction
 from filelock import FileLock
-from whoosh.writing import AsyncWriter
 
 from documents.models import Document
 from ... import index
@@ -24,11 +25,28 @@ from ...parsers import get_parser_class_for_mime_type
 logger = logging.getLogger("paperless.management.archiver")
 
 
+class LoggerWriter(TextIOBase):
+
+    def __init__(self, doc: Document):
+        self.doc = doc
+
+    def write(self, message):
+        # if statement reduces the amount of newlines that are
+        # printed to the logger
+        if message != '\n':
+            logger.error(
+                f"Document {self.doc} (ID: {self.doc.pk}): {message}"
+            )
+
+
 def handle_document(document_id):
     document = Document.objects.get(id=document_id)
 
     mime_type = document.mime_type
 
+    # redirect errors to the log and associate them with the current document
+    sys.stderr = LoggerWriter(document)
+
     parser_class = get_parser_class_for_mime_type(mime_type)
 
     if not parser_class: