How to convert an xml string to a dictionary in Py

Question

I have a program that reads an xml document from a socket. I have the xml document stored in a string which I would like to convert directly to a Python dictionary, the same way it is done in Django's simplejson library.

Take as an example:

str ="<?xml version="1.0" ?><person><name>john</name><age>20</age></person"
dic_xml = convert_to_dic(str)

Then dic_xml would look like {'person' : { 'name' : 'john', 'age' : 20 } }

Answer 1

I have a recursive method to get a dictionary from a lxml element

    def recursive_dict(element):
        return (element.tag.split('}')[1],
                dict(map(recursive_dict, element.getchildren()),
                     **element.attrib))

Answer 2

You can do this quite easily with lxml. First install it:

[sudo] pip install lxml

Here is a recursive function I wrote that does the heavy lifting for you:

from lxml import objectify as xml_objectify


def xml_to_dict(xml_str):
    """ Convert xml to dict, using lxml v3.4.2 xml processing library """
    def xml_to_dict_recursion(xml_object):
        dict_object = xml_object.__dict__
        if not dict_object:
            return xml_object
        for key, value in dict_object.items():
            dict_object[key] = xml_to_dict_recursion(value)
        return dict_object
    return xml_to_dict_recursion(xml_objectify.fromstring(xml_str))

xml_string = """<?xml version="1.0" encoding="UTF-8"?><Response><NewOrderResp>
<IndustryType>Test</IndustryType><SomeData><SomeNestedData1>1234</SomeNestedData1>
<SomeNestedData2>3455</SomeNestedData2></SomeData></NewOrderResp></Response>"""

print xml_to_dict(xml_string)

The below variant preserves the parent key / element:

def xml_to_dict(xml_str):
    """ Convert xml to dict, using lxml v3.4.2 xml processing library, see http://lxml.de/ """
    def xml_to_dict_recursion(xml_object):
        dict_object = xml_object.__dict__
        if not dict_object:  # if empty dict returned
            return xml_object
        for key, value in dict_object.items():
            dict_object[key] = xml_to_dict_recursion(value)
        return dict_object
    xml_obj = objectify.fromstring(xml_str)
    return {xml_obj.tag: xml_to_dict_recursion(xml_obj)}

If you want to only return a subtree and convert it to dict, you can use Element.find() to get the subtree and then convert it:

xml_obj.find('.//')  # lxml.objectify.ObjectifiedElement instance

See the lxml docs here. I hope this helps!

Answer 3

The code from http://code.activestate.com/recipes/410469-xml-as-dictionary/ works well, but if there are multiple elements that are the same at a given place in the hierarchy it just overrides them.

I added a shim between that looks to see if the element already exists before self.update(). If so, pops the existing entry and creates a lists out of the existing and the new. Any subsequent duplicates are added to the list.

Not sure if this can be handled more gracefully, but it works:

import xml.etree.ElementTree as ElementTree

class XmlDictConfig(dict):
    def __init__(self, parent_element):
        if parent_element.items():
            self.updateShim(dict(parent_element.items()))
        for element in parent_element:
            if len(element):
                aDict = XmlDictConfig(element)
                if element.items():
                    aDict.updateShim(dict(element.items()))
                self.updateShim({element.tag: aDict})
            elif element.items():
                self.updateShim({element.tag: dict(element.items())})
            else:
                self.updateShim({element.tag: element.text.strip()})

    def updateShim (self, aDict ):
        for key in aDict.keys():
            if key in self:
                value = self.pop(key)
                if type(value) is not list:
                    listOfDicts = []
                    listOfDicts.append(value)
                    listOfDicts.append(aDict[key])
                    self.update({key: listOfDicts})

                else:
                    value.append(aDict[key])
                    self.update({key: value})
            else:
                self.update(aDict)

Answer 4

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'}}

Answer 5

@dibrovsd: Solution will not work if the xml have more than one tag with same name

On your line of thought, I have modified the code a bit and written it for general node instead of root:

from collections import defaultdict
def xml2dict(node):
    d, count = defaultdict(list), 1
    for i in node:
        d[i.tag + "_" + str(count)]['text'] = i.findtext('.')[0]
        d[i.tag + "_" + str(count)]['attrib'] = i.attrib # attrib gives the list
        d[i.tag + "_" + str(count)]['children'] = xml2dict(i) # it gives dict
     return d

Answer 6

From @K3---rnc response (the best for me) I've added a small modifications to get an OrderedDict from an XML text (some times order matters):

def etree_to_ordereddict(t):
d = OrderedDict()
d[t.tag] = OrderedDict() if t.attrib else None
children = list(t)
if children:
    dd = OrderedDict()
    for dc in map(etree_to_ordereddict, children):
        for k, v in dc.iteritems():
            if k not in dd:
                dd[k] = list()
            dd[k].append(v)
    d = OrderedDict()
    d[t.tag] = OrderedDict()
    for k, v in dd.iteritems():
        if len(v) == 1:
            d[t.tag][k] = v[0]
        else:
            d[t.tag][k] = v
if t.attrib:
    d[t.tag].update(('@' + k, v) for k, v in t.attrib.iteritems())
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

Following @K3---rnc example, you can use it:

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_ordereddict(e))

Hope it helps ;)