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

Handle invalid bytes when in index_to_soup() for JavascriptRecipe

Browse Source
This commit is contained in:
Kovid Goyal 2014-06-09 15:03:49 +05:30
parent e58cd115e1
commit 7b284b949f
1 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/calibre/web/feeds/jsnews.py
View File

@ -16,6 +16,7 @@ from calibre.web.feeds import feeds_from_index
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
from calibre.web.fetch.javascript import fetch_page, AbortFetch, links_from_selectors from calibre.web.fetch.javascript import fetch_page, AbortFetch, links_from_selectors
from calibre.ebooks.chardet import xml_to_unicode, strip_encoding_declarations from calibre.ebooks.chardet import xml_to_unicode, strip_encoding_declarations
from calibre.utils.cleantext import clean_xml_chars
def image_data_to_url(data, base='cover'): def image_data_to_url(data, base='cover'):
from calibre.utils.imghdr import what from calibre.utils.imghdr import what
@ -221,7 +222,7 @@ class JavascriptRecipe(BasicNewsRecipe):
if raw: if raw:
return html return html
import html5lib import html5lib
root = html5lib.parse(html, treebuilder='lxml', namespaceHTMLElements=False).getroot() root = html5lib.parse(clean_xml_chars(html), treebuilder='lxml', namespaceHTMLElements=False).getroot()
return root return root
# ***************************** Internal API ***************************** # ***************************** Internal API *****************************