creat a DOM from an html document

[Mark Harrison]

I thought I saw a package that would create a DOM from html, with
allowances that it would do a "best effort" job to parse
non-perfectly formed html.

Now I can't seem to find this... does anybody have a recommendation
as to a good package to look at?

Many TIA!
Mark

[Mark Harrison]

Re: creat a DOM from an html document

Mark Harrison <mh@pixar.com > wrote:[color=blue]
> Now I can't seem to find this... does anybody have a recommendation
> as to a good package to look at?[/color]

Ahh, it's BeautifulSoup.. .

Thanks All!!

[John J. Lee]

Re: creat a DOM from an html document

Mark Harrison <mh@pixar.com > writes:
[color=blue]
> Mark Harrison <mh@pixar.com > wrote:[color=green]
> > Now I can't seem to find this... does anybody have a recommendation
> > as to a good package to look at?[/color]
>
> Ahh, it's BeautifulSoup.. .[/color]

Strictly that's not THE DOM, just A document object model. The DOM
proper is a standardised interface, which BeautifulSoup does not
implement. You could build a DOM using BeautifulSoup, though.


John

[Paul Boddie]

Re: creat a DOM from an html document

John J. Lee wrote:[color=blue]
> Mark Harrison <mh@pixar.com > writes:
>[color=green]
> > Ahh, it's BeautifulSoup.. .[/color]
>
> Strictly that's not THE DOM, just A document object model. The DOM
> proper is a standardised interface, which BeautifulSoup does not
> implement. You could build a DOM using BeautifulSoup, though.[/color]

For a certain value of standardised, libxml2dom provides "the DOM" for
HTML:

import urllib, libxml2dom
f = urllib.urlopen( "http://www.python.org" )
s = f.read(); f.close()
d = libxml2dom.pars eString(s, html=1)
print "There are", len(d.xpath("//table")), "tables in the document."

See http://www.python.org/pypi/libxml2dom for more information.

Paul

[Xavier Morel]

Re: creat a DOM from an html document

Mark Harrison wrote:[color=blue]
> I thought I saw a package that would create a DOM from html, with
> allowances that it would do a "best effort" job to parse
> non-perfectly formed html.
>
> Now I can't seem to find this... does anybody have a recommendation
> as to a good package to look at?
>
> Many TIA!
> Mark[/color]
While it doesn't generate a W3C DOM, BeautifulSoup is probably your best
bet for parsing less-than-perfect HTML and get something useable out of it.

Once you have your (parsed) document, you can either use it as is or try
to convert it to a valid W3C DOM though.