iwannamassage.py v0.2

# Fetch and parse an HTML page into a DOM-like tree of elements.
# Uses my new favorite XML parser, ElementTree by the effbot.
# Rather than use ElementTree's interface to the Tidy command
# (which runs Tidy in a subshell), I'm using M.-A. Lemburg's version
# that turns Tidy into a Python extention.

import sys, os, types, tempfile
from urllib2 import urlopen, URLError
from mx.Tidy import Tidy
from elementtree import ElementTree, HTMLTreeBuilder
from fileadaptor import FileAdaptor

class TidyError(Exception): pass

class TidyFilter:
    def __init__(self, src):
        print "TidyFilter __init__"
        self.src = src
   
        # mx.tidy.Tidy must have a true file object for input and output.
        # If the data source is not a true file object, use FileAdaptor to turn it into one.
        if not isinstance(self.src, types.FileType):
            self.inFileObj = FileAdaptor(self.src).file()
        else:
            self.inFileObj = self.src

        # Set up a temp file for output.
        self.outFileObj = os.tmpfile()

        # Use mx.Tidy.tidy to convert our input HTML into clean
        # output XHTML.
        nerrors, nwarnings, outputdata, errordata = 
            Tidy.tidy(self.inFileObj, self.outFileObj, output_xhtml=1)

        # Raise an exception if Tidy got an error
        if nerrors:
            raise TidyError

        # Make sure the temp output file is ready for input        
        self.outFileObj.flush()
        self.outFileObj.seek(0,0)
        return

    def read(self, size=None):
        return self.outFileObj.read(size)    

def main(url):

    # Get an ElementTree HTML parser.  
    htmlParserObj = HTMLTreeBuilder.TreeBuilder()

    # Open the URL, getting a file-like URL object.
    try:
        urlObj = urlopen(url)
    except URLError:
        print "Error opening URL '%s'" % url
        raise
    
    # Tell ElementTree to parse our input XHTML source, using an
    # HTML parser.  This will give us a top-level ElementTree object.
    treeObj = ElementTree.parse(TidyFilter(urlObj), parser=htmlParserObj)

    # For now, just dump out the tree so we can see it.
    ElementTree.dump(treeObj)
    return

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

© Copyright 2003 Michael Kent. Last update: 6/26/2003; 12:14:26 PM. This theme is based on the SoundWaves (blue) Manila theme.