Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-07 18:24:30 -04:00
Code Issues Packages Projects Releases Wiki Activity

Edit book: Fix position syncing incorrect when in HTML files that contain <meta> tags with charset encoding declarations over more than a single line

Browse Source
This commit is contained in:
Kovid Goyal 2013-12-23 09:57:59 +05:30
parent 5ba7de5fc7
commit d72cfadb6f
1 changed files with 11 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
src/calibre/ebooks/oeb/polish/parsing.py
View File

@ -18,7 +18,7 @@ from html5lib.ihatexml import InfosetFilter, DataLossWarning
from html5lib.html5parser import HTMLParser
from calibre import xml_replace_entities
from calibre.ebooks.chardet import xml_to_unicode, strip_encoding_declarations
from calibre.ebooks.chardet import xml_to_unicode, ENCODING_PATS
from calibre.ebooks.oeb.parse_utils import fix_self_closing_cdata_tags
from calibre.utils.cleantext import clean_xml_chars
@ -599,6 +599,16 @@ def parse_html5(raw, decoder=None, log=None, discard_namespaces=False, line_numb
raise ValueError('Failed to parse correctly, root has tag: %s and prefix: %s' % (root.tag, root.prefix))
return root
def strip_encoding_declarations(raw):
# A custom encoding stripper that preserves line numbers
limit = 10*1024
for pat in ENCODING_PATS:
prefix = raw[:limit]
suffix = raw[limit:]
prefix = pat.sub(lambda m: '\n' * m.group().count('\n'), prefix)
raw = prefix + suffix
return raw
def parse(raw, decoder=None, log=None, line_numbers=True, linenumber_attribute=None, replace_entities=True):
if isinstance(raw, bytes):
raw = xml_to_unicode(raw)[0] if decoder is None else decoder(raw)