- From: Matt Gray <g3021906@student.anu.edu.au>
- Date: Sun, 13 Sep 1998 05:30:48 -0400 (EDT)
- To: www-validator@w3.org
Hi,
I've been using the patch provided by Bernhard Weisshuhn on this newgroup
to use the lwp-request call. I've found that in order to handle server
errors properly I had to take out some 'die' lines in the code, otherwise
the script will exit too early without printing any of the error messages
to the user.
I don't know if this was a problem when using the original 'get' command,
or whether it is something particular to my computer (running Linux Redhat
5.1).
The changes are as follows:
--------------------------
Old code:
--------------------------
open( URI, "$grabber \"$uri_escaped\" |" ) || die "couldn't retrieve
uri: $!";
@file = <URI>;
close( URI ) || die "couldn't close uri retrieval pipe: $!";
--------------------------
New code:
--------------------------
open( URI, "$grabber \"$uri_escaped\" |" );
# || die "couldn't retrieve uri: $!";
@file = <URI>;
close( URI );
#|| die "couldn't close uri retrieval pipe: $!";
Regards,
Matt Gray.
>Enjoy, and feel free to comment or correct.
>
>regards,
>Bernhard Weisshuhn
>
>Index: check
>===================================================================
>RCS file: /sources/public/validator/httpd/cgi-bin/check,v
>retrieving revision 1.7
>diff -r1.7 check
>24c24
>< $abs_svc_uri = "http://validator.w3.org/";
>---
>> $abs_svc_uri = 'http://'.$ENV{'HTTP_HOST'}.'/';
>28c28
>< $grabber = "/usr/local/lib/libwww-perl/get";
>---
>> $grabber = "/usr/local/bin/lwp-request -Use";
>318a319,322
>> if (/^([0-9][0-9][0-9])\s+(.*)/ ) {
>> ( $response, $message ) = ($1,$2);
>> }
>>
>
Received on Sunday, 13 September 1998 14:50:25 UTC