0x2620/pandora_scripts
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

better year detection

Browse source
This commit is contained in:
j 2019-06-10 10:32:58 +01:00
parent c0cff7c541
commit c001e6fed7
1 changed files with 4 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

3
cleanup_collection.py
View file

@ -34,11 +34,12 @@ for d in api.findDocuments({
edit['title'] = title edit['title'] = title
year = re.compile(' \d{4} ').findall(document['title']) year = re.compile(' \d{4} ').findall(document['title'])
if year and not document.get('date'): if year and year[0] in ('1', '2') and not document.get('date'):
document['date'] = edit['date'] = year[0] document['date'] = edit['date'] = year[0]
if document.get('date') and document['date'] in title: if document.get('date') and document['date'] in title:
title = title.replace(document['date'], '').strip() title = title.replace(document['date'], '').strip()
if title:
edit['title'] = title edit['title'] = title
if edit: if edit: