process XMLHTTP response returning poorly formed html

[dandiebolt]

Using xmlhttp I am accessing a document from the web that is not xml
and is in fact not even proper html even though it is supposed to be
(unbalanced tags). Here is the type of code I am using:

url="http://www.domain.com/page.html";
var xmlhttp = new ActiveXObject("Msxml2.XMLHTTP");
xmlhttp.open("GET", url, false);
xmlhttp.send();
var xmlResp = xmlhttp.responseXML;

I want to create an array that holds the contents of every paragraph
<P> tag. The paragraphs are well formed with both opening and closing
tags <P> and </P> but the document as a whole is not valid xml or html.
How can I process the so-called xml response to extract the contents of
each paragraph into a unique element of the array? I would like to use
DOM methods to extract the paragraphs rather than parse up the text
with xmlhttp.responseText. Any help would be appreciated.

 

[strout]

It will be easy to parse by regular expression.

 

[dandiebolt]

It will be easy to parse by regular expression.

How? The only think I know about the document is that the information I
need is in between successive <P> and </P> tags. I was reluctant to
use regexp because I have any more structure than what is described and
I don't control the format of the page source.

 

[Tim Williams]

Why not automate IE to load the page and then grab your content once
IE has done it's job ? Presumably that will "fix" any irregularities
in the source.

Depends on where you want to do this, but is an option. Or just use
response.text and MSHTML ?

Tim.

 

[Jim Ley]

Using xmlhttp I am accessing a document from the web that is not xml
and is in fact not even proper html even though it is supposed to be
(unbalanced tags).

There's nothin inherent in unbalanced tags that would make something
not valid HTML - html fullly allows lots of closing elements as
optional.

How can I process the so-called xml response to extract the contents of
each paragraph into a unique element of the array? I would like to use
DOM methods to extract the paragraphs rather than parse up the text
with xmlhttp.responseText. Any help would be appreciated.

using the browser to parse the responseText as html will give you a
DOM of it, there's no other reasonable solution.

Jim.