Converts a Python dictionary or other native data type into a valid XML string.
Converts a Python dictionary or other native data type into a valid XML string.
Supports item (int, float, long, decimal.Decimal, bool, str, unicode, datetime, none and other number-like objects) and collection (list, set, tuple and dict, as well as iterable and dict-like objects) data types, with arbitrary nesting for the collections. Items with a datetime type are converted to ISO format strings. Items with a none type become empty XML elements.
The root object passed into the dicttoxml method can be any of the supported data types.
To satisfy XML syntax, the method prepends an <?xml version="1.0" encoding="UTF-8" ?> element and wraps the output in a <root> ... </root> element. However, this can be disabled to create XML snippets. Alternately, a custom root element can be specified by passing in the optional custom_root=foobar argument.
For lists of items, if each item is also a collection data type (lists, dict), the elements of that item are wrapped in a generic <item> ... </item> element.
Each element includes an optional type attribute with the data type. By default, the type attribute it included but it can be excluded by passing an optional attr_type=False argument when calling the dicttoxml method.
Note: datetime data types are converted into ISO format strings, and unicode and datetime data types get a str attribute.
Python -> XML
integer int
long long
float float
Decimal number
string str
unicode str
datetime str
None null
boolean bool
list list
set list
tuple list
dict dict
Elements with an unsupported data type raise a TypeError exception.
If an element name is invalid XML, it is rendered with the name "key" and the invalid name is included as a name attribute. E.g. { "^.{0,256}$": "foo" } would be rendered <key name="^.{0,256}$">foo</key>. An exception is element names with spaces, which are converted to underscores.
This module should work in Python 2.6+ and Python 3.
The dicttoxml module is published on the Python Package Index, so you can install it using pip or easy_install.
pip install dicttoxml
Or:
easy_install dicttoxml
Alternately, you can download the tarballed installer - dicttoxml-[VERSION].tar.gz - for this package from the dist directory on github and uncompress it. Then, from a terminal or command window, navigate into the unzipped folder and type the command:
python setup.py install
That should be all you need to do.
Once installed, import the library into your script and convert a dict into xml by running the dicttoxml function:
>>> import dicttoxml
>>> xml = dicttoxml.dicttoxml(some_dict)
Alternately, you can import the dicttoxml() function from the library.
>>> from dicttoxml import dicttoxml
>>> xml = dicttoxml(some_dict)
That's it!
Let's say you want to fetch a JSON object from a URL and convert it into XML. Here's how you can do that:
>>> import json
>>> import urllib
>>> import dicttoxml
>>> page = urllib.urlopen('http://quandyfactory.com/api/example')
>>> content = page.read()
>>> obj = json.loads(content)
>>> print(obj)
{u'mylist': [u'foo', u'bar', u'baz'], u'mydict': {u'foo': u'bar', u'baz': 1}, u'ok': True}
>>> xml = dicttoxml.dicttoxml(obj)
>>> print(xml)
<?xml version="1.0" encoding="UTF-8" ?><root><mylist><item type="str">foo</item><item type="str">bar</item><item type="str">baz</item></mylist><mydict><foo type="str">bar</foo><baz type="int">1</baz></mydict><ok type="bool">true</ok></root>
It's that simple.
By default, dicttoxml includes a type attribute for each element. Starting in version 1.4, you can turn this off by passing an optional attr_type=False argument to the dicttoxml method.
Using our example:
>>> xml = dicttoxml.dicttoxml(obj, attr_type=False)
>>> print(xml)
<?xml version="1.0" encoding="UTF-8" ?><root><mydict><foo>bar</foo><baz>1</baz></mydict><mylist><item>foo</item><item>bar</item><item>baz</item></mylist><ok>true</ok></root>
As you can see, the only difference is that the type attributes are now absent.
By default, dicttoxml wraps all the elements in a <root> ... </root> element. Starting in version 1.5, you can change the name of the root element to something else by passing an optional custom_root=some_custom_root argument to the dicttoxml method.
Using our example:
>>> xml = dicttoxml.dicttoxml(obj, custom_root=some_custom_root)
>>> print(xml)
<?xml version="1.0" encoding="UTF-8" ?><some_custom_root><mydict><foo>bar</foo><baz>1</baz></mydict><mylist><item>foo</item><item>bar</item><item>baz</item></mylist><ok>true</ok></some_custom_root>
As you can see, the name of the root element has changed to some_custom_root.
You can also create an XML snippet for inclusion into another XML document, rather than a full XML document itself.
Continuing with the example from above:
>>> xml_snippet = dicttoxml.dicttoxml(obj, root=False)
>>> print(xml_snippet)
<mylist><item type="str">foo</item><item type="str">bar</item><item type="str">baz</item></mylist><mydict><foo type="str">bar</foo><baz type="int">1</baz></mydict><ok type="bool">true</ok>
With the optional root argument set to False, the method converts the dict into XML without including an <?xml> prolog or a <root> element to enclose all the other elements.
As they say, Python comes with batteries included. You can easily syntax-check and pretty-print your XML using Python's xml.dom.minidom module.
Again, continuing with our example:
>>> from xml.dom.minidom import parseString
>>> dom = parseString(xml)
>>> print(dom.toprettyxml())
<?xml version="1.0" ?>
<root>
<mylist type="list">
<item type="str">foo</item>
<item type="str">bar</item>
<item type="str">baz</item>
</mylist>
<mydict type="dict">
<foo type="str">bar</foo>
<baz type="int">1</baz>
</mydict>
<ok type="bool">true</ok>
</root>
This makes the XML easier to read. If it is not well-formed, the xml parser will raise an exception.
Starting in version 1.1, you can set an optional ids parameter so that dicttoxml gives each element a unique id attribute.
With the ids flag on, the function generates a unique randomly-generated ID for each element based on the parent element in the form parent_unique. For list items, the id is in the form parent_unique_index.
Continuing with our example:
>>> xml_with_ids = dicttoxml.dicttoxml(obj, ids=True)
>>> print(parseString(xml_with_ids).toprettyxml())
<?xml version="1.0" ?>
<root>
<mylist id="root_160980" type="list">
<item id="mylist_609405_1" type="str">foo</item>
<item id="mylist_609405_2" type="str">bar</item>
<item id="mylist_609405_3" type="str">baz</item>
</mylist>
<mydict id="root_140407" type="dict">
<foo id="mydict_260437" type="str">bar</foo>
<baz id="mydict_111194" type="int">1</baz>
</mydict>
<ok id="root_612831" type="bool">true</ok>
</root>
Note that the default XML output remains the same as previous, so as not to break compatibility for existing uses.
Starting in version 1.3, dicttoxml accepts dict-like objects that are derived from the dict base class and treats them like dicts. For example:
>>> import collections
>>> dictlike = collections.OrderedDict({'foo': 1, 'bar': 2, 'baz': 3})
>>> xml = dicttoxml.dicttoxml(dictlike)
>>> print(xml)
<?xml version="1.0" encoding="UTF-8" ?><root><baz type="int">3</baz><foo type="int">1</foo><bar type="int">2</bar></root>
Also starting in version 1.3, dicttoxml accepts iterable objects and treats them like lists. For example:
>>> myiterator = xrange(1,11)
>>> xml = dicttoxml.dicttoxml(myiterator)
>>> print(xml)
'<?xml version="1.0" encoding="UTF-8" ?><root><item type="int">1</item><item type="int">2</item><item type="int">3</item><item type="int">4</item><item type="int">5</item><item type="int">6</item><item type="int">7</item><item type="int">8</item><item type="int">9</item><item type="int">10</item></root>'
As always, this remains compatible with arbitrary nesting of objects and types.
You can also enable debugging information.
>>> import dicttoxml
>>> dicttoxml.set_debug()
Debug mode is on. Events are logged at: dicttoxml.log
>>> xml = dicttoxml.dicttoxml(some_dict)
By default, debugging information is logged to dicttoxml.log, but you can change this:
>>> dicttoxml.set_debug(filename='some_other_filename.log')
Debug mode is on. Events are logged at: some_other_filename.log
To turn debug mode off, just call set_debug with an argument of False:
>>> dicttoxml.set_debug(False)
Debug mode is off.
If you encounter any errors in the code, please file an issue on github: https://github.com/quandyfactory/dicttoxml/issues.
Upload failed (500): Internal Server Error message when I try to upload the code. I'm incrementing the version by one and reinstalling it to see if that fixes the issue.isinstance(True, numbers.Number) returns True. I modified the get_xml_type() function to test for boolean before testing for numbers.Number. Thanks to badsequel for identifying and reporting the issue.decimal.Decimal as a number. This is done by replacing type(val).__name__ in ('int', 'long') with the more generic isinstance(val, number.Number). Thanks to jmagnusson for finding and fixing the error.convert_dict(). Thanks to jmagnusson for finding and fixing the error.logging with LOG. Thanks to mfriedenhagen for identifying the issue with the logger, and to seyhuns for supplying a pull request that could be merged automatically.get_xml_type().k -> key, v -> val).MANIFEST.in and setup.py so the licence and readme are properly included in the distribution.MANIFEST.in to include the LICENCE.txt and README.markdown files in the distribution, as per issue #15.dicttoxml() accepts [None] as a parameter and returns a valid XML object, as per issue #13.dicttoxml() now accepts None as a parameter and returns a valid XML object, as per issue #13.key_is_valid_xml() function to test if a key is valid XMLconvert_kv(), convert_bool() and convert_none() functions to test whether the key is a valid XML name and, if it is not, to render it as <key name="{invalidname}">value</key>. This addresses issue 10.ids argument to give each element a unique, randomly generated id attribute.type attribute.logging module.convert_list.root argument (default True) on whether to wrap the generated XML in an XML declaration and a root element.license attribute in setup.py.notify() function to debug_notify() and made it more comprehensive.Copyright 2012 by Ryan McGreal.
Released under the GNU General Public Licence, Version 2:
http://www.gnu.org/licenses/old-licenses/gpl-2.0.html