paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-09-16 21:55:37 -05:00
Code Issues Packages Projects Releases Wiki Activity

Extends the regex to find dates in documents as reported by @isaacsando

Browse Source
This commit is contained in:
Wolf-Bastian Pöttner
2018-02-08 09:01:55 +01:00
parent 178361b247
commit 07d06d9aee
4 changed files with 31 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/paperless_tesseract/parsers.py
View File

@@ -208,11 +208,13 @@ class RasterisedDocumentParser(DocumentParser):
# - XX/YY/ZZZZ with XX + YY being 1 or 2 and ZZZZ being 2 or 4 digits
# - XX-YY-ZZZZ with XX + YY being 1 or 2 and ZZZZ being 2 or 4 digits
# - XX. MONTH ZZZZ with XX being 1 or 2 and ZZZZ being 2 or 4 digits
# - MONTH ZZZZ
# - MONTH ZZZZ, with ZZZZ being 4 digits
# - MONTH XX, ZZZZ with XX being 1 or 2 and ZZZZ being 4 digits
m = re.search(
r'\b([0-9]{1,2})[\.\/-]([0-9]{1,2})[\.\/-]([0-9]{4}|[0-9]{2})\b|' +
r'\b([0-9]{1,2}\. [^ ]{3,9} ([0-9]{4}|[0-9]{2}))\b|' +
r'\b([^ ]{3,9} [0-9]{4})\b', text)
r'\b([0-9]{1,2}[\. ]+[^ ]{3,9} ([0-9]{4}|[0-9]{2}))\b|' +
r'\b([\w]{3,9} [0-9]{1,2}, ([0-9]{4}))\b|' +
r'\b([\w]{3,9} [0-9]{4})\b', text)
if m is None:
return None