Javascript example of taking a webpage and performing text analysis

LeperKing

Hi Folks,

Is it possible for javascript to take "www.website.com/index.html" as a string, and then counting the number of times "apple" is in the webpage?

Thanks,
LK.

galwayguy22

I'm fairly sure you could use an Iframe for this.

Set the src of the iframe to be the website you want to analyze, then use some regular expressions on the contents of the iframe.

galwayguy22

No, that doesn't seem to work. You don't have permission to read web pages not on your own domain

<script>
function show()
{
 alert(document.getElementById('ifrm').contentWindow.document.body.innerHTML)
 
}
</script>
<iframe width=600 height=600 frameborder=1 src="http://www.apple.com/" id=ifrm></iframe>
<button onclick="show()" >Show HTML</button>

Get this error in firefox:
Error: Permission denied to get property Window.innerHTML
Source File: file:///C:/Documents%20and%20Settings/MyActualName/Desktop/iframesample.html
Line: 4


You could always do it server side. I'm fairly sure, again, in PHP you can stream and read the contents of any url. Eg. file_get_contents

LeperKing

Ok, thanks for that. I won't have server side capabilities for this. Thanks for checking....

LK.

Eoin

You could use an AJAX request to read in the contents of a remote page and then do your calculations. Will still be client side.

Liam Byrne

galwayguy22 wrote: »

I'm fairly sure, again, in PHP you can stream and read the contents of any url. Eg. file_get_contents

Be aware that PHP5 switched this off by default for security reasons. Some hosts might have re-enabled it, but it's off by default.

CuLT

Could also do it with a simple enough bit of java, if you want to run it from your home computer or server or wherever.

Starting point:

import javax.swing.text.html.*;
import javax.swing.text.*;
import java.net.*;
import java.io.*;
import java.util.*;