Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-06-23 15:30:45 -04:00
Code Issues Packages Projects Releases Wiki Activity

Heuristic processing: Fix the italicize common patterns algorithm breaking on some HTML markup. Fixes #922317 (Private bug)

Browse Source
This commit is contained in:
Kovid Goyal 2012-01-28 16:43:25 +05:30
parent 565fc2d479
commit bf034f4c5b
1 changed files with 5 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
src/calibre/ebooks/conversion/utils.py
View File

@ -157,7 +157,7 @@ class HeuristicProcessor(object):
ITALICIZE_STYLE_PATS = [ ITALICIZE_STYLE_PATS = [
ur'(?msu)(?<=[\s>"“\'‘])_(?P<words>[^_]+)_', ur'(?msu)(?<=[\s>"“\'‘])_(?P<words>[^_]+)_',
ur'(?msu)(?<=[\s>"“\'‘])/(?P<words>[^/\*>]+)/', ur'(?msu)(?<=[\s>"“\'‘])/(?P<words>[^/\*><]+)/',
ur'(?msu)(?<=[\s>"“\'‘])~~(?P<words>[^~]+)~~', ur'(?msu)(?<=[\s>"“\'‘])~~(?P<words>[^~]+)~~',
ur'(?msu)(?<=[\s>"“\'‘])\*(?P<words>[^\*]+)\*', ur'(?msu)(?<=[\s>"“\'‘])\*(?P<words>[^\*]+)\*',
ur'(?msu)(?<=[\s>"“\'‘])~(?P<words>[^~]+)~', ur'(?msu)(?<=[\s>"“\'‘])~(?P<words>[^~]+)~',
@ -172,8 +172,11 @@ class HeuristicProcessor(object):
for word in ITALICIZE_WORDS: for word in ITALICIZE_WORDS:
html = re.sub(r'(?<=\s|>)' + re.escape(word) + r'(?=\s|<)', '<i>%s</i>' % word, html) html = re.sub(r'(?<=\s|>)' + re.escape(word) + r'(?=\s|<)', '<i>%s</i>' % word, html)
def sub(mo):
return '<i>%s</i>'%mo.group('words')
for pat in ITALICIZE_STYLE_PATS: for pat in ITALICIZE_STYLE_PATS:
html = re.sub(pat, lambda mo: '<i>%s</i>' % mo.group('words'), html) html = re.sub(pat, sub, html)
return html return html