Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-08-11 09:13:57 -04:00
Code Issues Packages Projects Releases Wiki Activity

AZW3 Input: Handle files with garbage bytes in their table of contents. Fixes #1297713 [private](https://bugs.launchpad.net/calibre/+bug/1297713)

Browse Source
This commit is contained in:
Kovid Goyal 2014-03-26 15:12:57 +05:30
parent a258f4a201
commit 99db7985bd
1 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
src/calibre/ebooks/metadata/toc.py
View File

@ -13,6 +13,7 @@ from lxml.builder import ElementMaker
from calibre.constants import __appname__, __version__
from calibre.ebooks.BeautifulSoup import BeautifulSoup
from calibre.ebooks.chardet import xml_to_unicode
from calibre.utils.cleantext import clean_xml_chars
NCX_NS = "http://www.daisy.org/z3986/2005/ncx/"
CALIBRE_NS = "http://calibre.kovidgoyal.net/2009/metadata"
@ -136,7 +137,7 @@ class TOC(list):
try:
if not os.path.exists(toc):
bn = os.path.basename(toc)
bn = bn.replace('_top.htm', '_toc.htm') # Bug in BAEN OPF files
bn = bn.replace('_top.htm', '_toc.htm') # Bug in BAEN OPF files
toc = os.path.join(os.path.dirname(toc), bn)
self.read_html_toc(toc)
@ -258,6 +259,7 @@ class TOC(list):
text = ''
c[1] += 1
item_id = 'num_%d'%c[1]
text = clean_xml_chars(text)
elem = E.navPoint(
E.navLabel(E.text(re.sub(r'\s+', ' ', text))),
E.content(src=unicode(np.href)+(('#' + unicode(np.fragment))