python lxml find tag

Question

I'm using lxml to parse an html that has a facebook comments tag that looks like that:

<fb:comments id="fb_comments"  href="http://example.com" num_posts="5" width="600"></fb:comments>

I am trying to select it to get the href value but when i do a cssselect('fb:comments') i get the following error:

The pseudo-class Symbol(u'comments', 3) is unknown

Is there a way to do it?

Edit: The code:

from lxml.html import fromstring
html = '...'
parser = fromstring(html)
parser.cssselect('fb:comments')  #raises the exception

Please provide your python code. – Mariusz Jamro Jun 19 '12 at 14:18 — Mariusz Jamro, Jun 19 '12 at 14:18

score 3 · Accepted Answer · answered Jun 19 '12 at 14:30

3

The cssselect() method parses the document using given CSS selector expression. In your case the colon character (:) is a XML namespace prefix separator (i.e. <namespace:tagname/>) which is confused with CSS pseudo-class syntax (i.e. tagname:pseudo-class).

According to lxml manual you should use namespace-prefix|element syntax in cssselect()in order to to find a tag (comments) with a namespace prefix (fb). So:

from lxml.html import fromstring
html = '...'
parser = fromstring(html)
parser.cssselect('fb|comments')

answered Jun 19 '12 at 14:30

Mariusz Jamro

30,615
24
120
162

great! thanks, but that is returning `Undefined namespace prefix` any idea why? – applechief Jun 19 '12 at 14:35
1

Maybe try with just `comments` (w/o namespace) ? – Mariusz Jamro Jun 19 '12 at 14:39
See http://stackoverflow.com/questions/6860013/xhtml-namespace-issues-with-cssselect-in-lxml on how to use namespaces with cssselect. – Simon Sapin Jul 28 '12 at 08:11

python lxml find tag

1 Answers1