Automation problem

Hi all, I am a EE student and I want to automate my design process.
And I want to get some advices from you guys.

Previously when I do the design, it finishes up with a XML file, and I
go to a website

http://www.stateye.org/stateyev4/weblaborator/

Then I choose my file path and press the little "submit" (floppy disk
icon) below. After that, the homepage reloads and the uploaded file(s)
show up. And then I press the little "translation" (tool icon). And
the XML file will be translated in their algorithm to a MATLAB (.m)
file.

What I want to do is that I want to automate this process by just
typing "tcl ABC.xml" in my Linux console, and TCL will go to the
website and do the steps that I mentioned above and get the ABC.m for
me.

I read the HTML of that the homepage. Seems to me that it was written
in the ASP language, just wondering whether I have to download
particular library for this. Or are there any particular skill in
writting this kind of program.

I get stucked in this problem for many days, I really appreciate if
anybody can give me some hints for solving this problem. Or, perhaps
how I can start tackling the problem. Thank you very very much.

On 25 aug, 07:51, tonylo...@gmail.com wrote:
> Hi all, I am a EE student and I want to automate my design process.
> And I want to get some advices from you guys.
>

While I have no specific advice for you, you might take a
look at the various "Wiki reapers" on the Wiki, they show
at least some of the aspects of dealing with browsers/Internet
sites.

Regards,

Arjen

On Aug 25, 6:51 am, tonylo...@gmail.com wrote:
>
> I get stucked in this problem for many days, I really appreciate if
> anybody can give me some hints for solving this problem. Or, perhaps
> how I can start tackling the problem. Thank you very very much.

On the assumption that you are working on a *nix type system, the best
approach would be to use a domain specific tool, such as curl, wget or
even Lynx, to first automate the actual HTTP conversations. You can
then use a shell script or Tcl (or whatever) to tie the tasks
together, ie feed the response of the first HTTP conversation through
filters to prepare the next HTTP conversation.

For an example with curl and Tcl, see:
http://www.flightlab.com/~joe/gutter/packages/curl.html

Dave

On Aug 25, 4:57*am, "dave.joub...@googlemail.com"
<dave.joub...@googlemail.com> wrote:
> On Aug 25, 6:51 am, tonylo...@gmail.com wrote:
>
>
>
> > I get stucked in this problem for many days, I really appreciate if
> > anybody can give me some hints for solving this problem. Or, perhaps
> > how I can start tackling the problem. Thank you very very much.
>
> On the assumption that you are working on a *nix type system, the best
> approach would be to use a domain specific tool, such as curl, wget or
> even Lynx, to first automate the actual HTTP conversations. You can
> then use a shell script or Tcl (or whatever) to tie the tasks
> together, ie feed the response of the first HTTP conversation through
> filters to prepare the next HTTP conversation.
>
> For an example with curl and Tcl, see:http://www.flightlab.com/~joe/gutter/packages/curl.html
>
> Dave



Hi Dave and other esteem expert,

I just read the documentation of Curl, it seems to be a nice choice
for my application.
Seems to me I just need to use the syntax curl -F
upload=@localfilename -F press=OK [url]

But I face two new problems from this,
1. The "floppy" icon only gets a name but doesn't get a value, thus I
cannot trigger the floopy icon by this method.
2. After uploading the file, the homepage reloads, and how can I
interact with the reloaded homepage.

I know that I am very dumb, thank you very much for you help.

Cheers,
Tony