Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Fix regex for py3

Browse Source
This commit is contained in:
Kovid Goyal 2019-08-30 07:56:18 +05:30
parent 260c330da3
commit b0fe64571e
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
src/calibre/ebooks/conversion/preprocess.py
View File

@ -602,7 +602,7 @@ class HTMLPreProcessor(object):
end_rules.append( end_rules.append(
# Un wrap using punctuation # Un wrap using punctuation
(re.compile(( (re.compile((
r'(?<=.{%i}([a-zäëïöüàèìòùáćéíĺóŕńśúýâêîôûçąężıãõñæøþðßěľščťžňďřů,:)\\IA\u00DF]' r'(?<=.{%i}([a-zäëïöüàèìòùáćéíĺóŕńśúýâêîôûçąężıãõñæøþðßěľščťžňďřů,:)\\IAß]'
r'|(?<!\&\w{4});))\s*(?P<ital></(i|b|u)>)?\s*(</p>\s*<p>\s*)+\s*(?=(<(i|b|u)>)?' r'|(?<!\&\w{4});))\s*(?P<ital></(i|b|u)>)?\s*(</p>\s*<p>\s*)+\s*(?=(<(i|b|u)>)?'
r'\s*[\w\d$(])') % length, re.UNICODE), wrap_lines), r'\s*[\w\d$(])') % length, re.UNICODE), wrap_lines),
) )

2
src/calibre/ebooks/conversion/utils.py
View File

@ -367,7 +367,7 @@ class HeuristicProcessor(object):
# define the pieces of the regex # define the pieces of the regex
# (?<!\&\w{4});) is a semicolon not part of an entity # (?<!\&\w{4});) is a semicolon not part of an entity
lookahead = "(?<=.{"+unicode_type(length)+"}([a-zა-ჰäëïöüàèìòùáćéíĺóŕńśúýâêîôûçąężıãõñæøþðßěľščťžňďřů,:)\\IA\u00DF]|(?<!\\&\\w{4});))" lookahead = "(?<=.{"+unicode_type(length)+r"}([a-zა-ჰäëïöüàèìòùáćéíĺóŕńśúýâêîôûçąężıãõñæøþðßěľščťžňďřů,:)\\IAß]|(?<!\&\w{4});))"
em_en_lookahead = "(?<=.{"+unicode_type(length)+"}[\u2013\u2014])" em_en_lookahead = "(?<=.{"+unicode_type(length)+"}[\u2013\u2014])"
soft_hyphen = "\xad" soft_hyphen = "\xad"
line_ending = "\\s*(?P<style_close></(span|[iub])>)?\\s*(</(p|div)>)?" line_ending = "\\s*(?P<style_close></(span|[iub])>)?\\s*(</(p|div)>)?"