Very Computer

Wello World,

I'm trying to track stock markets by automatically opening URL's,
saving the html files and post-processing them (searching for quotes).

Currently I'm doing it the usual way (Remote Control of UNIX Netscape):

        netscape -id $my_id -noraise -remote "openURL..."
  or    netscape-remote -id $my_id -noraise -remote "openURL..."
        (Sun Sparc, UNIX, Solaris 2.5, using the remote.c implementation)

Unfortunately, if an error occurs (e.g. 'unable to connect to server'),
a message window from Netscape pops up. But I don't like error windows
on my desktop. I would like to control all errors from my own master process.
I'm looking for a functionality like this:

  openURL(http://www.....) - but don't pop up error windows
  saveAs(/home...filename.htm) - if the document has been loaded

I tried the error handler functions from Java (onError), but, as the Java
Manual says, messages like 'unable to connect to server' are Navigator
Errors, not JavaScript Errors, and so they cannot be escaped by an
'onError' function.

Is there any way of stopping Netscape from raising error message windows,
or is there any direct way to open a URL and save the contents to a file
without invoking the Netscape browser ? Is there perhaps a command-line
based tool like 'ftp' that can read http files ?

Thanks for any comments,

Luke Farmer

The easiest way is probably
  lynx -dump <URL>

By the way, you should have posted this in a browsers newsgroup, not a
servers newsgroup.

--

  Statistical Laboratory,  16 Mill Lane, Cambridge, CB2 1SB, England
  "The Bishop of Huntingdon and Postman Pat each opened new school extensions"
                                            (Cambridge Weekly News, 28-May-97)

There is a tool called wget that would be useful, and there are various
modules for perl that will retrieve URLs.  LWP, e.g.
--
Scott Rose              
<URL:http://www.cs.wisc.edu/~condon/family/rose.html>
Bicycling Community Page  <URL:http://danenet.wicip.org/bcp/>