urls found in stylesheets have extra quotes

[Jean-Paul Calderone]

If a css url is quoted, the quotes confuse the link extraction logic:

>>> from lxml.html import fromstring
>>> fromstring("""
... <style>background: url('/foo/bar.png');</style>
... """)
<Element html at b7687bfc>
>>> tree = _
>>> list(tree.iterlinks())
[(<Element style at b7143ecc>, None, "'/foo/bar.png'", 16)]
>>> tree.make_links_absolute('http://example.com/')
>>> list(tree.iterlinks())
[(<Element style at b7508aac>, None, "http://example.com/'/foo/bar.png'", 16)]

[scoder (scoder)]

Fixed in current trunk:

>>> import lxml.etree as et
>>> print et.__version__, et.LIBXML_VERSION
2.3.dev (2, 7, 6)

>>> from lxml.html import fromstring
>>> tree = fromstring("""
... <style>background: url('/foo/bar.png');</style>
... """)
>>> list(tree.iterlinks())
[(<Element style at 2526d50>, None, '/foo/bar.png', 17)]
>>> tree.make_links_absolute('http://example.com/')
>>> list(tree.iterlinks())
[(<Element style at 2526d50>, None, 'http://example.com/foo/bar.png', 17)]

[scoder (scoder)]

Fixed in lxml 2.3alpha1.