14

I want to parse HTML with lxml using XPath expressions. My problem is matching for the contents of a tag:

For example given the 

<a href="http://something">Example</a>

element I can match the href attribute using 

.//a[@href='http://something']

but the given the expression

.//a[.='Example']

or even

.//a[contains(.,'Example')]

lxml throws the 'invalid node predicate' exception.

What am I doing wrong?

EDIT:

Example code:

from lxml import etree
from cStringIO import StringIO

html = '<a href="http://something">Example</a>'
parser = etree.HTMLParser()
tree   = etree.parse(StringIO(html), parser)

print tree.find(".//a[text()='Example']").tag

Expected output is 'a'. I get 'SyntaxError: invalid node predicate'

Giacomo1968
  • 25,759
  • 11
  • 71
  • 103
akosch
  • 4,326
  • 7
  • 59
  • 80

1 Answers1

23

I would try with:

.//a[text()='Example']

using xpath() method:

tree.xpath(".//a[text()='Example']")[0].tag

If case you would like to use iterfind(), findall(), find(), findtext(), keep in mind that advanced features like value comparison and functions are not available in ElementPath.

lxml.etree supports the simple path syntax of the find, findall and findtext methods on ElementTree and Element, as known from the original ElementTree library (ElementPath). As an lxml specific extension, these classes also provide an xpath() method that supports expressions in the complete XPath syntax, as well as custom extension functions.

systempuntoout
  • 71,966
  • 47
  • 171
  • 241
  • I don't want to find the link based on href, but based on the text it contains: "Example" in the above example :) .//a[@href='http://something'] works the way it is... – akosch Apr 14 '10 at 13:59
  • 1
    you need to remove an = .//a[text()='Example'] – Greg Apr 14 '10 at 14:20
  • Thanks for your suggestion, but this one raises "SyntaxError: invalid node predicate" too – akosch Apr 14 '10 at 14:20
  • Thank you: with XPath() it really works. Strangely enough @href works in both cases. – akosch Apr 15 '10 at 01:09
  • @systempuntoout Then is `.//a[text()='Example']` invalid in this case? – Learner Nov 20 '15 at 16:26