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

Update Amazon metadata download plugin to handle amazon.com website change that was preventing any metadata from being downloaded

Browse Source 
Fixes a bug in the default html5lib lxml treebuilder that caused it to
fail on pages that have comments with -- or trailing hyphens.
This commit is contained in:
Kovid Goyal 2015-11-21 13:31:32 +05:30
parent 9a9e5d5d7b
commit c68b9b7d64
1 changed files with 5 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
src/html5lib/treebuilders/etree_lxml.py
View File

@ -309,6 +309,11 @@ class TreeBuilder(_base.TreeBuilder):
if (parent == self.document and if (parent == self.document and
self.document._elementTree.getroot()[-1].tag == comment_type): self.document._elementTree.getroot()[-1].tag == comment_type):
warnings.warn("lxml cannot represent adjacent comments beyond the root elements", DataLossWarning) warnings.warn("lxml cannot represent adjacent comments beyond the root elements", DataLossWarning)
if data['data']:
# lxml cannot handle comment text that contains -- or endswith -
# Should really check if changes happened and issue a data loss
# warning, but that's a fairly big performance hit.
data['data'] = data['data'].replace('--', '\u2010\u2010').rstrip('-')
super(TreeBuilder, self).insertComment(data, parent) super(TreeBuilder, self).insertComment(data, parent)
def insertRoot(self, token): def insertRoot(self, token):