[hobbit] About Web monitor

[Ralph Mitchell]

On 8/23/07, Henrik Stoerner <henrik at hswn.dk> wrote:
> On Wed, Aug 22, 2007 at 12:35:30PM +0800, case wrote:
> > This's a browser/server architecture, we have Crystal Report 8  used(10.2.10.7).
> > One IE browser with URL
> > "10.2.10.7/tcard/tempSchl.rpt?user0 at c.rpt=rpt_read&password0 at c.rpt=rpt_read&user1=rpt_read&password1=rpt_read&init=actx"
> > on client launched and called Crystal Report Viewer, a new IE window would be opened.
> >
> > How can we monitor this new IE window or content in this new IE window?
>
> It's difficult. What I normally do in these cases is to use something
> like wireshark (ethereal) to trace the requests going to the server, and
> see if I can just grab one of the pages without going through the whole
> series of screens that the user does. If that is not possible, then the
> only way forward is to write a script to mimic the behaviour of your
> browser, which can be quite tricky to do - getting everything right with
> dynamic URL's, Javascript "onLoad()" functions, cookies etc. can be a lot
> of work.

There are a couple of tools that can help:

1) LiveHTTPHeaders plugin for Firefox shows you exactly what headers
flow between the browser and the server;

2) curl is pretty good at grabbing web pages.  It has a library and
extensions for using it within Perl, PHP, tcl, C, etc, as well as the
command line version for use in shell scripts.

Ralph Mitchell