diff options
| author | Leonard Richardson <leonard.richardson@canonical.com> | 2011-02-20 10:04:35 -0500 |
|---|---|---|
| committer | Leonard Richardson <leonard.richardson@canonical.com> | 2011-02-20 10:04:35 -0500 |
| commit | b3ba2b97b2d4c4e5559baadc96f1844753b38df4 (patch) | |
| tree | 61c2113cb8bbfddfe7d4792b9e3cf14213af4fe6 /beautifulsoup/builder/lxml_builder.py | |
| parent | 9f437ea591aeaf16d593350baf081315e56a8b73 (diff) | |
Made the XML treebuilder able to handle basic invalid XML.
Diffstat (limited to 'beautifulsoup/builder/lxml_builder.py')
| -rw-r--r-- | beautifulsoup/builder/lxml_builder.py | 11 |
1 files changed, 4 insertions, 7 deletions
diff --git a/beautifulsoup/builder/lxml_builder.py b/beautifulsoup/builder/lxml_builder.py index e431a62..9f4c0bd 100644 --- a/beautifulsoup/builder/lxml_builder.py +++ b/beautifulsoup/builder/lxml_builder.py @@ -11,7 +11,7 @@ class LXMLTreeBuilderForXML(TreeBuilder): def default_parser(self): # This can either return a parser object or a class, which # will be instantiated with default arguments. - return etree.XMLParser + return etree.XMLParser(target=self, strip_cdata=False, recover=True) def __init__(self, parser=None, empty_element_tags=None): if empty_element_tags is not None: @@ -71,10 +71,6 @@ class LXMLTreeBuilderForXML(TreeBuilder): self.soup.handle_data(content) self.soup.endData(Comment) - def test_fragment_to_document(self, fragment): - """See `TreeBuilder`.""" - return u'<html><body>%s</body></html>' % fragment - class LXMLTreeBuilder(HTMLTreeBuilder, LXMLTreeBuilderForXML): @@ -82,5 +78,6 @@ class LXMLTreeBuilder(HTMLTreeBuilder, LXMLTreeBuilderForXML): def default_parser(self): return etree.HTMLParser - def end(self, name): - self.soup.handle_endtag(name) + def test_fragment_to_document(self, fragment): + """See `TreeBuilder`.""" + return u'<html><body>%s</body></html>' % fragment |