Today I had the unfortunate luck of having to integrate an XML web service into an application I’m working on.
As you might already know, parsing JSON data with Python is really simple thanks to some great built in tools. I wish I could say the same thing for working with XML! Unfortunately, while Python does include built in XML processing tools, they’re not exactly the easiest thing in the world to work with.
I’ve written about XML parsing in Python before (a really long time ago!), but after spending an hour hunting around for the most elegant solution earlier today, I figured I’d write a quick post as future reference to myself.
When I last worked with XML in Python, I recommended that users try to use the lxml library. LXML is pretty great because it allows you do some advanced things with XML, but on the downside: it requires C libraries (yuck!) that are annoying to install locally, and adds another dependency to your project.
Instead of going the LXML route, I’d like to suggest using Python’s built in xml.etree module. It’s got a (relatively) simple interface that makes parsing simple XML documents fairly easy – and since that’s all I needed to do, it seemed like a good solution.
If you’re looking for a quick way to extract XML data, read on.
Anyhow, let’s say you’ve got a simple XML document that looks something like the below:
<?xml version='1.0' encoding='UTF-8'?> <!DOCTYPE xgdresponse SYSTEM 'xgdresponse.dtd'> <xgdresponse version='1.0'> <transid>2771709</transid> <errorcode>0</errorcode> <response> <result> <element>666</element> <errorcode>0</errorcode> <value>SOMETHING IMPORTANT!</value> </result> </response> </xgdresponse>
Now, let’s say you need to extract the
value element out of this XML tree.
After lots of experimentation, this is the simplest way I found to do the data
"""Extract the `value` element from the XML tree, if it exists.""" from xml.etree import ElementTree as ET xml = """ <?xml version='1.0' encoding='UTF-8'?> <!DOCTYPE xgdresponse SYSTEM 'xgdresponse.dtd'> <xgdresponse version='1.0'> <transid>2771709</transid> <errorcode>0</errorcode> <response> <result> <element>666</element> <errorcode>0</errorcode> <value>SOMETHING IMPORTANT!</value> </result> </response> </xgdresponse> """.strip() value = ET.fromstring(xml).find('response/result/value') if value: print 'Found value:', value.text
$ parse.py Found value: SOMETHING IMPORTANT!
The way this works is that we:
- Load our XML document into memory, and construct an XML
- We then use the
findmethod, passing in an XPath selector, which allows us to specify what element we’re trying to extract.
- If the element can’t be found,
- If the element can be found, then we’ll use the
.textproperty on our element object to grab the data out of the desired XML element.
All in all, not so bad!
The next time you’re looking to quickly extract some data out of an XML document, give it a try!
Got a better method? Drop me a line.