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

Replace common line break patterns with line breaks

Browse Source
This commit is contained in:
Kovid Goyal 2008-02-21 20:49:27 +00:00
parent debc7c26de
commit c8cce1e355
1 changed files with 6 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/libprs500/ebooks/lrf/html/convert_from.py
View File

@ -96,12 +96,18 @@ class HTMLConverter(object):
# remove <p> tags from within <a> tags # remove <p> tags from within <a> tags
(re.compile(r'<a.*?>(.*?)</a\s*>', re.DOTALL|re.IGNORECASE), (re.compile(r'<a.*?>(.*?)</a\s*>', re.DOTALL|re.IGNORECASE),
lambda match: re.compile(r'<\s*?p.*?>', re.IGNORECASE).sub('', match.group())), lambda match: re.compile(r'<\s*?p.*?>', re.IGNORECASE).sub('', match.group())),
# Replace common line break patterns with line breaks
(re.compile(r'<p>(&nbsp;|\s)*</p>', re.IGNORECASE), lambda m: '<br />'),
# Replace entities # Replace entities
(re.compile(ur'&(\S+?);'), partial(entity_to_unicode, (re.compile(ur'&(\S+?);'), partial(entity_to_unicode,
exceptions=['lt', 'gt', 'amp'])), exceptions=['lt', 'gt', 'amp'])),
# Remove comments from within style tags as they can mess up BeatifulSoup # Remove comments from within style tags as they can mess up BeatifulSoup
(re.compile(r'(<style.*?</style>)', re.IGNORECASE|re.DOTALL), (re.compile(r'(<style.*?</style>)', re.IGNORECASE|re.DOTALL),
strip_style_comments), strip_style_comments),
] ]
# Fix Baen markup # Fix Baen markup
BAEN = [ BAEN = [