paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-04-02 13:45:10 -05:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #1591 from paperless-ngx/fix/1583-tika-str

Browse Source 
Fix: Parsing Tika documents fails with AttributeError
This commit is contained in:
shamoon 2022-09-14 21:52:49 -07:00 committed by GitHub
commit 8ed401aec1
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 11 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
src/paperless_tika/parsers.py
View File

@ -1,4 +1,5 @@
import os import os
from pathlib import Path
import dateutil.parser import dateutil.parser
import requests import requests
@ -28,6 +29,11 @@ class TikaDocumentParser(DocumentParser):
def extract_metadata(self, document_path, mime_type): def extract_metadata(self, document_path, mime_type):
tika_server = settings.TIKA_ENDPOINT tika_server = settings.TIKA_ENDPOINT
# tika does not support a PathLike, only strings
# ensure this is a string
document_path = str(document_path)
try: try:
parsed = parser.from_file(document_path, tika_server) parsed = parser.from_file(document_path, tika_server)
except Exception as e: except Exception as e:
@ -47,10 +53,14 @@ class TikaDocumentParser(DocumentParser):
for key in parsed["metadata"] for key in parsed["metadata"]
] ]
def parse(self, document_path, mime_type, file_name=None): def parse(self, document_path: Path, mime_type, file_name=None):
self.log("info", f"Sending {document_path} to Tika server") self.log("info", f"Sending {document_path} to Tika server")
tika_server = settings.TIKA_ENDPOINT tika_server = settings.TIKA_ENDPOINT
# tika does not support a PathLike, only strings
# ensure this is a string
document_path = str(document_path)
try: try:
parsed = parser.from_file(document_path, tika_server) parsed = parser.from_file(document_path, tika_server)
except Exception as err: except Exception as err: