Matěj Suchánek has uploaded a new change for review. (
https://gerrit.wikimedia.org/r/329761 )
Change subject: Fix and improve default regexes
......................................................................
Fix and improve default regexes
- Remove unneccessary flags.
- Clean up 'header' using multiline.
- Expand 'pre' to support HTML attributes (mostly 'style').
- Update 'property' to support parameters (currently, it supports
"|from=" but it might support more in the future).
- Localize 'property' and 'invoke' using magic words.
Change-Id: Ib805bf70cb1cc99711138d7d6c7e40971f31b602
---
M pywikibot/textlib.py
1 file changed, 9 insertions(+), 7 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/pywikibot/core
refs/changes/61/329761/2
diff --git a/pywikibot/textlib.py b/pywikibot/textlib.py
index 9f7782e..2908ee1 100644
--- a/pywikibot/textlib.py
+++ b/pywikibot/textlib.py
@@ -221,13 +221,13 @@
_regex_cache.update({
'comment': re.compile(r'(?s)<!--.*?-->'),
# section headers
- 'header': re.compile(r'\r?\n=+.+=+ *\r?\n'),
+ 'header': re.compile(r'(?m)^=+.+=+ *$'),
# preformatted text
- 'pre': re.compile(r'(?ism)<pre>.*?</pre>'),
+ 'pre': re.compile(r'(?is)<pre[ >].*?</pre>'),
'source': re.compile(r'(?is)<source .*?</source>'),
- 'score': re.compile(r'(?ism)<score[ >].*?</score>'),
+ 'score': re.compile(r'(?is)<score[ >].*?</score>'),
# inline references
- 'ref': re.compile(r'(?ism)<ref[ >].*?</ref>'),
+ 'ref': re.compile(r'(?is)<ref[ >].*?</ref>'),
'template': NESTED_TEMPLATE_REGEX,
# lines that start with a space are shown in a monospace font and
# have whitespace preserved.
@@ -247,11 +247,13 @@
site.validLanguageLinks() +
list(site.family.obsolete.keys()))),
# Wikibase property inclusions
- 'property': re.compile(r'(?i)\{\{\s*#property:\s*p\d+\s*\}\}'),
+ 'property': (r'(?i)\{\{\s*#(%s):\s*p\d+.*?\}\}',
+ lambda site:
'|'.join(site.getmagicwords('property'))),
# Module invocations (currently only Lua)
- 'invoke': re.compile(r'(?i)\{\{\s*#invoke:.*?}\}'),
+ 'invoke': (r'(?i)\{\{\s*#(%s):.*?\}\}',
+ lambda site: '|'.join(site.getmagicwords('invoke'))),
# categories
- 'category': ('\[\[ *(?:%s)\s*:.*?\]\]',
+ 'category': (r'\[\[ *(?:%s)\s*:.*?\]\]',
lambda site: '|'.join(site.namespaces[14])),
# files
'file': (FILE_LINK_REGEX,
--
To view, visit https://gerrit.wikimedia.org/r/329761
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: Ib805bf70cb1cc99711138d7d6c7e40971f31b602
Gerrit-PatchSet: 2
Gerrit-Project: pywikibot/core
Gerrit-Branch: master
Gerrit-Owner: Matěj Suchánek <matejsuchane...@gmail.com>
Gerrit-Reviewer: jenkins-bot <>
_______________________________________________
MediaWiki-commits mailing list
MediaWiki-commits@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
