Paths in ul-li

.gitignore

bin
local
include
lib
man
.Python
*.pyc
*.swp

animals.html

<ul>
  <li>Animals
    <ul>
      <li>Birds</li>
      <li>Mammals
	<ul>
          <li>Elephant</li>
          <li><a href="#">Mouse</a></li>
	</ul>
      </li>
      <li>Reptiles</li>
    </ul>
  </li>
  <li>Plants
    <ul>
      <li>Flowers
	<ul>
          <li>Rose</li>
          <li>Tulip</li>
	</ul>
      </li>
      <li>Trees</li>
    </ul>
  </li>
</ul>

requirements.txt

PyZen==0.3.2
lxml==2.3.4
nose==1.1.2
rednose==0.3.3

test_tree.py

from lxml import etree
from nose.tools import assert_equal
from tree import tree_paths

def test_should_find_tree_paths():
    with open('animals.html') as f:
        tree = etree.parse(f)
    paths = tree_paths(tree)
    
    assert_equal([
            '/Animals',
            '/Animals/Birds',
            '/Animals/Mammals',
            '/Animals/Mammals/Elephant',
            '/Animals/Mammals/Mouse',
            '/Animals/Reptiles',
            '/Plants',
            '/Plants/Flowers',
            '/Plants/Flowers/Rose',
            '/Plants/Flowers/Tulip',
            '/Plants/Trees'
        ],
        paths.keys())
    element = paths['/Animals/Mammals/Mouse']
    assert_equal('a', element.tag)

tree.py

from collections import OrderedDict

def path_to(elem):
    ancestors_having_text = elem.xpath('./ancestor-or-self::*[boolean(normalize-space(text()))]')
    return [e.text.strip() for e in ancestors_having_text ]

def path_as_string(path):
    return '/' + '/'.join(path)
    
def tree_paths(tree):
    elements_having_text = tree.xpath('.//*[boolean(normalize-space(text()))]')
    return OrderedDict([ (path_as_string(path_to(l)), l) for l in elements_having_text ])