I think, yes, it is possible, but not as easy as 1,2,3. One need to investigate the structure of two initial pages, get their elements, figure out what elements to click to get to the third page, and then grab the information from there.

Now, the basic question is: what if they change the page structures? If it is in house application, and for one-two users, and programmer is always present, then one can go this way. But if it is an application for distribution, then it is not a reliable approach. It would be better to have the Web site involved into the agreement for providing the information following the documented procedure.




>Thanks for your help Yuri.
>
>This is the page where I need to do a search to get the results:
>
>http://www.sunbiz.org/jlilist.html
>
>Let's say I search there for the "a", all the names starting with "a" will come up, but it's a summary of the results, so I have to click on the one that I want and then I get the detail info for that record that looks like this page:
>
>http://www.sunbiz.org/scripts/jlidet.exe?action=DETLIST&inquiry_number=J07000392871&inquiry_date=047035296001000000&return_number=J03900007479&return_date=045691776000300000
>
>This is where the information that I want is, get some dates and other info, create a record in my own table and then go to the next record on the web page, until all the 'a' have been worked, then search on 'b' and so on and so forth. All this needs to be automatic, is this possible?