Python BeautifulSoup XML Parsing

Question

I've written a simple script to parse XML chat logs using the BeautifulSoup module. The standard soup.prettify() works ok except chat logs have a lot of fluff in them. You can see both the script code and some of the XML input file I'm working with below:

Code

import sys
from BeautifulSoup import BeautifulSoup as Soup

def parseLog(file):
    file = sys.argv[1]
    handler = open(file).read()
    soup = Soup(handler)
    print soup.prettify()

if __name__ == "__main__":
    parseLog(sys.argv[1])

Test XML Input

<?xml version="1.0"?>
<?xml-stylesheet type='text/xsl' href='MessageLog.xsl'?>
<Log FirstSessionID="1" LastSessionID="2"><Message Date="10/31/2010" Time="3:43:48 PM"     DateTime="2010-10-31T20:43:48.937Z" SessionID="1"><From><User FriendlyName="Jon"/></From>    <To><User FriendlyName="Bill"/></To><Text Style="font-family:Segoe UI; color:#000000; ">hey, what's up?</Text></Message>
<Message Date="10/31/2010" Time="3:44:03 PM" DateTime="2010-10-15T20:44:03.421Z" SessionID="1"><From><User FriendlyName="Jon"/></From><To><User FriendlyName="Bill"/></To><Text Style="font-family:Segoe UI; color:#000000; ">Got your message</Text></Message> 
<Message Date="10/31/2010" Time="3:44:31 PM" DateTime="2010-10-15T20:44:31.390Z" SessionID="2"><From><User FriendlyName="Bill"/></From><To><User FriendlyName="Jon"/></To><Text Style="font-family:Segoe UI; color:#000000; ">oh, great</Text></Message>
<Message Date="10/31/2010" Time="3:44:59 PM" DateTime="2010-10-15T20:44:59.281Z" SessionID="2"><From><User FriendlyName="Bill"/></From><To><User FriendlyName="Jon"/></To><Text Style="font-family:Segoe UI; color:#000000; ">hey, i gotta run</Text></Message>

I'm wanting to be able to output this into a format like the following or at least something that is more readable than pure XML:

Jon: Hey, what's up? [10/31/10 @ 3:43p]

Jon: Got your message [10/31/10 @ 3:44p]

Bill: oh, great [10/31/10 @ 3:44p]

etc.. I've heard some decent things about the PyParsing module, maybe it's time to give it a shot.

Solution 1

BeautifulSoup makes getting at attributes and values in xml really simple. I tweaked your example function to use these features.

import sys
from BeautifulSoup import BeautifulSoup as Soup

def parseLog(file):
    file = sys.argv[1]
    handler = open(file).read()
    soup = Soup(handler)
    for message in soup.findAll('message'):
        msg_attrs = dict(message.attrs)
        f_user = message.find('from').user
        f_user_dict = dict(f_user.attrs)
        print "%s: %s [%s @ %s]" % (f_user_dict[u'friendlyname'],
                                    message.find('text').decodeContents(),
                                    msg_attrs[u'date'],
                                    msg_attrs[u'time'])


if __name__ == "__main__":
    parseLog(sys.argv[1])

Solution 2

I'd recommend using the builtin ElementTree module. BeautifulSoup is meant to handle unwell-formed code like hacked up HTML, whereas XML is well-formed and meant to be read by an XML library.

Update: some of my recent reading here suggests lxml as a library built on and enhancing the standard ElementTree.

Comments

[Steven D. Majewski]

Why not XSLT ? That would be the easiest. ( In fact: I see there's an ?xml-stylesheet directive -- what does the default stylesheet look like ? )

[Caley Woods]

I may not always have the XSL stylesheet available, thus the need for something to format the log into something a bit more readable. If I can use the same stylesheet as one that I do have, that might also work.

[Caley Woods]

This works perfectly. What exactly is contained in the dictionary from f_user_dict = dict(f_user.attrs) I assume attributes, i'll have to toy with that piece and see exactly what's there. Thanks again!

[dcolish]

Yup the for all elements, el.attrs will contain a list of tuples of the xml tags attributes. Calling dict on any tuple will make it into a dictionary.

[dcolish]

Oh just wanted to clarify, calling dict on a list of tuples will return a dictionary, not a single tuple: dict([('hello', 'goodbye'), ('foo', 'bar')])

[mamakurka]

I use Beautiful Soup for parsing XML. From the docs: "Beautiful Soup is a Python library for pulling data out of HTML and XML files." Beautiful Soup will use whichever parser you tell it to, including lxml. (see crummy.com/software/BeautifulSoup/bs4/doc/#installing-a-pars‌​er)

[whatnick]

And I have seen plenty of string generated malformed XML lately.