: to adjust the setting according to the recomandations given but all what
: I get is the script loaded as a regular text in my browser. Executing
: the script from the command line (to see the error messages) give the
: following message:
: The archive has not been setup correctly! Sorry!
marc-search.cgi is a CGI script, and it gets much of the information
it needs from environment variables that are set by your web
server. You could set these yourself and run the script from
the command line, but there's little point in doing so.
: I have placed the cgi script into the archive directory. Could it be
: somewhere else?
This will not work, and explains why you're seeing the script loaded
as regular text in your browser. You must install it in a 'scriptaliased'
directory, which is to say, a directory that has been designated
in your web server configuration files as a place for cgi-scripts.
On many systems, this is simply http://www.some.site/cgi-bin/
but you should check with your server administrator to find out for certain.
: #! /usr/aix3-PD/bin/perl
provided that this is a perl V5 binary, this should be fine, though
there should be no space between the ! and the /
: my $server = 'http://fserv.wiz.uni-kassel.de';
: could be "www.wiz.uni-kassel.de" but doesn't seam to make a differenz.
This is fine, though I imagine you'd want to use the form that you
give out when distributing the URL for your site. In other words,
if you think of your site as www.wiz, then that's what I'd use.
: my $help = $server . '';
: SHould be added later - may this stay free for the moment ?
Yes. If you produce a translation of the help file in German, I
would be very grateful for a copy (which I could then make available
to those who are interested).
: my $doc_root = '/kww/www';
: According to the mail I have read this should be the path to the
: root directory of the www server ?
: I have problems with this section: The above absolute path would we
: the URL of:
: http://fserv.wiz.uni-kassel.de/kww/index.html
: The complete absolute address of the file "index.html" in our directory is:
: cserv:/kww/www
$doc_root should be set to whatever directory is the absolute top
level for your web server's document area. On your system (which
I've just taken a quick peek at), you should give the path which
corresponds to http://fserv.wiy.uni-kassel.de/index.html.
This is what happens 'behind the scenes':
someone initiates a search of your archive.
The web server sets an environment variable for the URL from which the
search was initiated and executes the CGI script.
The script removes the http:// part from the front of this environment
variable so that it has the path (relative to the top level of your
server's document area) for your archive.
The script concatenates the value in $doc_root with this relative path
to produce the absolute path for the directory where your archive
is located and begins the search.
Let's say that your archive is available at
http://www.fserv.wiz.uni-kassel.de/kww/irrig_l/index.html
and that this page has a link pointing to the marc-search.cgi
script in a properly configured script aliased directory.
Clicking on that link would set the environment variable to the
(hypothetial) URL given above. when the script executes, it will
remove the first and last parts, leaving 'kww/irrig_l' to be concatenated
with the value in $doc_root, which I'm guessing would be /www.
When marc-search begins its search, then, it will look in the
directory /www/kww/irrig_l to find the files associated with your
archive.
: my $script = $server . '/kww/www/irrig_l/db/marc-search.cgi';
Again, this needs to supply the path for the script in a scriptaliased
directory.
: my $cgilib = '/kww/www/irrig_l/db';
:
: I downloaded the "cgi-lib.pl" from somewhere (v 1.2 , 1996)
My copy, released January 21, 1996 is at version 2.4
check http://www.bio.cam.ac.uk/cgi-lib/
: Does the program require the "maillist.html" file ?
maillist.html is the index file mhonarc generates for your archive.
You should add the pointer to the cgi-script to this document
by editing your mhonarc resource file.
: I want to search in more archives (every month one) so is it possible to
: search in all aerchives (ddirectories)?
At present you can only search one directory at a time. Multiple
directories is somewhere on my to do list, but since I personally
have no need for this, I'll confess that it's near the bottom.
Do let me know if you produce a german translation of the help file.
best wishes,
Eric
--
Eric D. Friedman
friedman(_at_)uci(_dot_)edu
