Just getting numbers from a string

wmkolcz wrote: I used screen scraping to get html from a number of our web sites that we are converting to a CMS. I was asked to grab one block of code that includes links to individuals using a numeric ID as their primary key. This is  where my 'expertise' ends...lol. Lets say you have a list block like: <ul> <li><a href="details.cfm?id=20>Steve String</a></li> <li><a href="details.cfm?id=50>Mary String</a></li> <li><a href="details.cfm?id=120>Jerry String</a></li> </ul> I want to strip out the ID's to only produce: 20 50 120 or 20,50,120 (would be optimal). How I can use probably regex to remove all non numeric numbers from a block of string?

The following assumes ColdFusion 8 or newer (Rematch). I have also assumed that the IDs are 2 or 3 digits long. You can easily adapt the code as appropriate.

<cfsavecontent variable="block"><ul>

<li><a href="details.cfm?id=20>Steve String</a></li>

<li><a href="details.cfm?id=50>Mary String</a></li>

<li><a href="details.cfm?id=120>Jerry String</a></li>

</ul>

</cfsavecontent>

<!--- Raw list is of the form =20>, =100>, etc.  --->

<cfset rawList = arrayToList(REMatch("(=[0-9]{2,3}>)",block))>

<cfset numberList = replaceList(rawList,"=,>",",")>

<cfoutput>#numberList#</cfoutput>

Ok, I used  <cfset cleanDocs = reReplace( theDocs, "[^[:digit:]]", ' ', "all") /> which infact did remove all the non numeric characters but leaves me with duplicates (this is a weird escape(url) in with the hyperlinks so the ID is in there twice. Anyone know how to remove duplicates from a string?