Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 10:44:09 -04:00
Code Issues Packages Projects Releases Wiki Activity

Fix #941 (Raw UTF-8 misinterpreted in HTML with explicitly specified encoding)

Browse Source
This commit is contained in:
Marshall T. Vandegrift 2008-08-11 12:11:10 -04:00
parent 07a6deb2a6
commit 3f6d486fb8
1 changed files with 5 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/calibre/ebooks/lit/reader.py
View File

@ -158,7 +158,7 @@ class UnBinary(object):
continue continue
elif c == '\v': elif c == '\v':
c = '\n' c = '\n'
self.buf.write(c.encode('utf-8')) self.buf.write(c.encode('ascii', 'xmlcharrefreplace'))
elif state == 'get flags': elif state == 'get flags':
if oc == 0: if oc == 0:
@ -258,7 +258,8 @@ class UnBinary(object):
state = 'get attr' state = 'get attr'
elif count > 0: elif count > 0:
if not in_censorship: if not in_censorship:
self.buf.write(unicode(c).encode('utf-8')) self.buf.write(c.encode(
'ascii', 'xmlcharrefreplace'))
count -= 1 count -= 1
if count == 0: if count == 0:
if not in_censorship: if not in_censorship:
@ -310,7 +311,8 @@ class UnBinary(object):
path = self.item_path(doc) path = self.item_path(doc)
if m and frag: if m and frag:
path += m + frag path += m + frag
self.buf.write((u'"%s"' % path).encode('utf-8')) self.buf.write((u'"%s"' % path).encode(
'ascii', 'xmlcharrefreplace'))
state = 'get attr' state = 'get attr'
return index return index