Flatten an element tree node into an xpath ordered dict.

flatten_xml.py

from __future__ import absolute_import, division, print_function, unicode_literals

from collections import OrderedDict


def flatten_xml(node):
    """
    Produce an ordered dictionary of elements enountered.

    Keys are valid xpath selectors.
    """
    output = OrderedDict()
    children = {}  # by tag
    if node:  # has children
        for child in node:
            children.setdefault(child.tag, []).append(flatten_xml(child))
    elif node.text:
        output[''] = node.text
    for attrib, value in node.attrib.iteritems():
        output['@{}'.format(attrib)] = value
    for tag, child_xpaths in children.iteritems():
        path = tag if len(child_xpaths) == 1 else tag + '[{}]'
        for i, child_xpath in enumerate(child_xpaths, 1):
            child_path = path.format(i)
            for key, value in child_xpath.iteritems():
                output['{}/{}'.format(child_path, key) if key else child_path] = value
    return output