Jeff

While not exactly like eBay, i tried a similar scheme for a client... 
a general-purpose "harvest & resolve URLs" so that all linked URLs 
would be in a frame ore even a popup window (the client's window was 
to remain active as the control center).

The client wanted to provide a "master control panel" to news sites 
such as cnn, fox, msnbc, etc.

What I found is simply this:

   If a target URL wants to thwart harvesting, it can do so quite easily, and in
   such a way that it is impractical (if not impossible) to detect and 
resolve by
   programming"

Such things as:

   metatags that reload/repoint a url

   JavaScript onLoad, onUnload, timer events which do the same

   hidden links, e.g. a dropdown (or other) that invokes a javascript function,
   which dynamically creates and executes a link

I encountered some pretty clever ways to hide links.

You just had to analyze each page.

Then, once you had implemented a scheme, the target URL could thwart 
you again. by simply changing its methods... you are always one step 
(at least) behind.

It becomes increasingly complex as you traverse a hierarchy of 
imbedded links where many different methods are used.

What you end up doing (or trying to do) is writing an intelligent 
browser/javascript interpreter

Dick



At 10:34 AM -0500 12/15/00, Jeffry Houser wrote:
>  An ecommerce site that I work on sometimes does eBay auctions.  They
>want the winning bidder to be able to go to their site, plug in their name
>and auction number, have a script go out to ebay and grab the
>auction info and create an order for the person.
>
>  I did some plugging around eBay in hopes that they actually encourage
>this sort of thing and I could grab the data via wddx.  I found what I
>was expecting to find, which was nothing. 
>
>  I told the client that the first thing we would have to do was
>de-construct the eBay URLs so that we would be able to create a URL
>directly to the auction page.  This could take hours, or weeks, or be
>completely hopeless.  I haven't looked at eBay from a 'how do they do
>it' standpoint.  Once we have the URL I can grab the page, via CFHTTP,
>and parse it out to get the necesarry information.
>
>   Has anyone had any experience with this or know of any sites that
>perform similiar actions?  In the past hasn't eBay tried to block 3rd
>party 'auction search' programs that automatically search their site
>for auctions?  Wouldn't an attempt at deconstructing their URL's border
>on hacking?

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
        Structure your ColdFusion code with Fusebox. Get the official book at 
http://www.fusionauthority.com/bkinfo.cfm

Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/
Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists

Reply via email to