Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-10-30 18:22:25 -04:00
Code Issues Packages Projects Releases Wiki Activity

...

Browse Source
This commit is contained in:
Kovid Goyal 2011-08-24 17:48:58 -06:00
commit c63ff81d96
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

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

@ -315,7 +315,7 @@ class HeuristicProcessor(object):
supports a range of html markup and text files
'''
# define the pieces of the regex
lookahead = "(?<=.{"+str(length)+u"}([a-zäëïöüàèìòùáćéíóńśúâêîôûçąężıãõñæøþðßě,:)\IA\u00DF]|(?<!\&\w{4});))" # (?<!\&\w{4});) is a semicolon not part of an entity
lookahead = "(?<=.{"+str(length)+u"}([a-zäëïöüàèìòùáćéíóńśúâêîôûçąężıãõñæøþðßôľščťžňďěřů,:)\IA\u00DF]|(?<!\&\w{4});))" # (?<!\&\w{4});) is a semicolon not part of an entity
em_en_lookahead = "(?<=.{"+str(length)+u"}[\u2013\u2014])"
soft_hyphen = u"\xad"
line_ending = "\s*</(span|[iubp]|div)>\s*(</(span|[iubp]|div)>)?"