extract content from file

darkhalf · September 19th, 2006, 10:09 AM

Hi,

using xmlHTTP through an hta, I'm able to get back a broken web page that sadly has errors, and I can't fix it (outside of my control)... Now I get the text back, but I would like to know how could I extract a section of the text that is not corrupt... In essense I would need to get everything below a comment, and then take that bit, and through it in a Div so I can Dom it...

darkhalf · September 19th, 2006, 10:15 AM

Hi,

To clarify, the data that I want is in the last table... so if I use regex (!<table>.*</table>!s), how do I get the last table from the text?

joefawcett · September 19th, 2006, 10:38 AM

A different approach is to load into a HTML Dom document:

Code:

var oDoc = new ActiveXObject("htmlfile"); 
oDoc.open();
var sHtml = "<html><head><title>Test</title></head><body>Hello 
World</body></html>"; 
oDoc.write(sHtml);
oDoc.close();
alert(oDoc.documentElement.outerHTML);

Obviously you'll need to find the table, show the actual HTML if still stuck.

Using your method with regexp you need to get the second match object after perform the search.

http://msdn.microsoft.com/library/de...b8ed696d0a.asp

--

Joe (Microsoft MVP - XML)

darkhalf · September 19th, 2006, 11:31 AM

No that approach cannot work... as the page that I'm getting back was very poorly written, and there is javscript within the <body> that doesn't work, and generates errors once it loads... my best solution is to do the regex, and use the exec to pull build the array... I need help to build the regex expression, as !<table>.*</table>!s does not work @ all...

How can make a regex to grab the tables...?

darkhalf · September 19th, 2006, 12:29 PM

Ok, I've managed to make this one... which works fine (ish) for the first table.

var myRe = new RegExp("<table((.|\n)*?)</table>","ig");

Problem now, is that there are embedded tables... so this does not work... I just need to get the last table tags of the text... any suggestions?