How to convert an XML string to a dictionary?

Question

xmltodict (full disclosure: I wrote it) does exactly that:

xmltodict.parse("""
<?xml version="1.0" ?>
<person>
  <name>john</name>
  <age>20</age>
</person>""")
# {u'person': {u'age': u'20', u'name': u'john'}}

This is a great module that someone created. I've used it several times. http://code.activestate.com/recipes/410469-xml-as-dictionary/

Here is the code from the website just in case the link goes bad.

from xml.etree import cElementTree as ElementTree

class XmlListConfig(list):
    def __init__(self, aList):
        for element in aList:
            if element:
                # treat like dict
                if len(element) == 1 or element[0].tag != element[1].tag:
                    self.append(XmlDictConfig(element))
                # treat like list
                elif element[0].tag == element[1].tag:
                    self.append(XmlListConfig(element))
            elif element.text:
                text = element.text.strip()
                if text:
                    self.append(text)


class XmlDictConfig(dict):
    '''
    Example usage:

    >>> tree = ElementTree.parse('your_file.xml')
    >>> root = tree.getroot()
    >>> xmldict = XmlDictConfig(root)

    Or, if you want to use an XML string:

    >>> root = ElementTree.XML(xml_string)
    >>> xmldict = XmlDictConfig(root)

    And then use xmldict for what it is... a dict.
    '''
    def __init__(self, parent_element):
        if parent_element.items():
            self.update(dict(parent_element.items()))
        for element in parent_element:
            if element:
                # treat like dict - we assume that if the first two tags
                # in a series are different, then they are all different.
                if len(element) == 1 or element[0].tag != element[1].tag:
                    aDict = XmlDictConfig(element)
                # treat like list - we assume that if the first two tags
                # in a series are the same, then the rest are the same.
                else:
                    # here, we put the list in dictionary; the key is the
                    # tag name the list elements all share in common, and
                    # the value is the list itself 
                    aDict = {element[0].tag: XmlListConfig(element)}
                # if the tag has attributes, add those to the dict
                if element.items():
                    aDict.update(dict(element.items()))
                self.update({element.tag: aDict})
            # this assumes that if you've got an attribute in a tag,
            # you won't be having any text. This may or may not be a 
            # good idea -- time will tell. It works for the way we are
            # currently doing XML configuration files...
            elif element.items():
                self.update({element.tag: dict(element.items())})
            # finally, if there are no child tags and no attributes, extract
            # the text
            else:
                self.update({element.tag: element.text})

Example usage:

tree = ElementTree.parse('your_file.xml')
root = tree.getroot()
xmldict = XmlDictConfig(root)

//Or, if you want to use an XML string:

root = ElementTree.XML(xml_string)
xmldict = XmlDictConfig(root)

The following XML-to-Python-dict snippet parses entities as well as attributes following this XML-to-JSON "specification". It is the most general solution handling all cases of XML.

from collections import defaultdict

def etree_to_dict(t):
    d = {t.tag: {} if t.attrib else None}
    children = list(t)
    if children:
        dd = defaultdict(list)
        for dc in map(etree_to_dict, children):
            for k, v in dc.items():
                dd[k].append(v)
        d = {t.tag: {k:v[0] if len(v) == 1 else v for k, v in dd.items()}}
    if t.attrib:
        d[t.tag].update(('@' + k, v) for k, v in t.attrib.items())
    if t.text:
        text = t.text.strip()
        if children or t.attrib:
            if text:
              d[t.tag]['#text'] = text
        else:
            d[t.tag] = text
    return d

It is used:

from xml.etree import cElementTree as ET
e = ET.XML('''
<root>
  <e />
  <e>text</e>
  <e name="value" />
  <e name="value">text</e>
  <e> <a>text</a> <b>text</b> </e>
  <e> <a>text</a> <a>text</a> </e>
  <e> text <a>text</a> </e>
</root>
''')

from pprint import pprint
pprint(etree_to_dict(e))

The output of this example (as per above-linked "specification") should be:

{'root': {'e': [None,
                'text',
                {'@name': 'value'},
                {'#text': 'text', '@name': 'value'},
                {'a': 'text', 'b': 'text'},
                {'a': ['text', 'text']},
                {'#text': 'text', 'a': 'text'}]}}

Not necessarily pretty, but it is unambiguous, and simpler XML inputs result in simpler JSON. :)

---

Update

If you want to do the reverse, emit an XML string from a JSON/dict, you can use:

try:
  basestring
except NameError:  # python3
  basestring = str

def dict_to_etree(d):
    def _to_etree(d, root):
        if not d:
            pass
        elif isinstance(d, basestring):
            root.text = d
        elif isinstance(d, dict):
            for k,v in d.items():
                assert isinstance(k, basestring)
                if k.startswith('#'):
                    assert k == '#text' and isinstance(v, basestring)
                    root.text = v
                elif k.startswith('@'):
                    assert isinstance(v, basestring)
                    root.set(k[1:], v)
                elif isinstance(v, list):
                    for e in v:
                        _to_etree(e, ET.SubElement(root, k))
                else:
                    _to_etree(v, ET.SubElement(root, k))
        else:
            raise TypeError('invalid type: ' + str(type(d)))
    assert isinstance(d, dict) and len(d) == 1
    tag, body = next(iter(d.items()))
    node = ET.Element(tag)
    _to_etree(body, node)
    return ET.tostring(node)

pprint(dict_to_etree(d))

This lightweight version, while not configurable, is pretty easy to tailor as needed, and works in old pythons. Also it is rigid - meaning the results are the same regardless of the existence of attributes.

import xml.etree.ElementTree as ET

from copy import copy

def dictify(r,root=True):
    if root:
        return {r.tag : dictify(r, False)}
    d=copy(r.attrib)
    if r.text:
        d["_text"]=r.text
    for x in r.findall("./*"):
        if x.tag not in d:
            d[x.tag]=[]
        d[x.tag].append(dictify(x,False))
    return d

So:

root = ET.fromstring("<erik><a x='1'>v</a><a y='2'>w</a></erik>")

dictify(root)

Results in:

{'erik': {'a': [{'x': '1', '_text': 'v'}, {'y': '2', '_text': 'w'}]}}

The most recent versions of the PicklingTools libraries (1.3.0 and 1.3.1) support tools for converting from XML to a Python dict.

The download is available here: PicklingTools 1.3.1

There is quite a bit of documentation for the converters here: the documentation describes in detail all of the decisions and issues that will arise when converting between XML and Python dictionaries (there are a number of edge cases: attributes, lists, anonymous lists, anonymous dicts, eval, etc. that most converters don't handle). In general, though, the converters are easy to use. If an 'example.xml' contains:

<top>
  <a>1</a>
  <b>2.2</b>
  <c>three</c>
</top>

Then to convert it to a dictionary:

>>> from xmlloader import *
>>> example = file('example.xml', 'r')   # A document containing XML
>>> xl = StreamXMLLoader(example, 0)     # 0 = all defaults on operation
>>> result = xl.expect XML()
>>> print result
{'top': {'a': '1', 'c': 'three', 'b': '2.2'}}

There are tools for converting in both C++ and Python: the C++ and Python do indentical conversion, but the C++ is about 60x faster