selecting attribute values from lxml

I want to use an xpath expression to get the value of an attribute.

I expected the following to work

from lxml import etree

for customer in etree.parse('file.xml').getroot().findall('BOB'):
    print customer.find('./@NAME')

but this gives an error :

Traceback (most recent call last):
  File "bob.py", line 22, in <module>
    print customer.find('./@ID')
  File "lxml.etree.pyx", line 1409, in lxml.etree._Element.find (src/lxml/lxml.etree.c:39972)
  File "/usr/local/lib/python2.7/dist-packa开发者_运维技巧ges/lxml/_elementpath.py", line 272, in find
    it = iterfind(elem, path, namespaces)
  File "/usr/local/lib/python2.7/dist-packages/lxml/_elementpath.py", line 262, in iterfind
    selector = _build_path_iterator(path, namespaces)
  File "/usr/local/lib/python2.7/dist-packages/lxml/_elementpath.py", line 246, in _build_path_iterator
    selector.append(ops[token[0]](_next, token))
KeyError: '@'

Am I wrong to expect this to work?

find and findall only implement a subset of XPath. Their presence is meant to provide compatibility with other ElementTree implementations (like ElementTree and cElementTree).

The xpath method, in contrast, provides full access to XPath 1.0:

print customer.xpath('./@NAME')[0]

However, you could instead use get:

print customer.get('NAME')

or attrib:

print customer.attrib['NAME']

As a possible useful addition, this is how to get the value of an attribute in the case that the element has more than one, and it is the only difference with respect to another element. E.g., given the following file.xml:

<?xml version ="1.0" encoding="UTF-8"?>
    <level1>
      <level2 first_att='att1' second_att='foo'>8</level2>
      <level2 first_att='att2' second_att='bar'>8</level2>
    </level1>

One can access the attribute 'bar' with:

import lxml.etree as etree
tree = etree.parse("test_file.xml")
print tree.xpath("//level1/level2[@first_att='att2']/@second_att")[0]