My web app (for IIS reasons) uses meta refreshes to redirect the user around. I test these redirects with the code below. I've used a regex as my refresh template is fixed and very, very simple. However, if yours isn't/aren't then you should replace the regex with a call to something like HTML::TokeParser.

Update: fixed silly mistake as hilighted below. Also fixed what was, for my test suite, a logic error. The final get call must be to $expected_url and not $url. If your test quite works differently then use $url instead :).
sub meta_refresh
{
    my $mech = shift;
    my $expected_url = shift;
    
    my $url;
    
    if($mech->content() =~ /<meta http-equiv="refresh" content="0;url=
+([^"]*)"/)
    {
        $url = $1;
    }

    cmp_ok($expected_url, 'eq', $url, "The meta refresh returns the ex
+pected URL");
    
    $mech->get( $expected_url );
    ok($mech->success(), "URL loaded successfully");
}

[download]
So call it like this: 
   # Code to cause the refresh to appear not shown.
   
   # Check the refresh and follow
   meta_refresh($mech, '/index.cgi?rm=home');

[download]

Replies are listed 'Best First'.
Re: WWW::Mechanize follow meta refreshes
by merlyn (Sage) on Apr 13, 2005 at 14:38 UTC 
    $mech->content() =~ /<meta http-equiv="refresh" content="0;url=([^"]*)
+"/;
my $url = $1;

    [download]
    Never never never use $1 without having tested the match. If the match fails, you're using a previous $1 from a previous successful match. Oops!

    -- Randal L. Schwartz, Perl hacker
    Be sure to read my standard disclaimer if this is a reply.

[reply]
[d/l]
      Oops!! :).

      I have updated the node.
[reply]
Re: WWW::Mechanize follow meta refreshes
by Kanji (Parson) on Apr 14, 2005 at 15:19 UTC
    I've used a regex as my refresh template is fixed and very, very simple. However, if yours isn't/aren't then you should replace the regex with a call to something like HTML::TokeParser.

    This is actually built into WWW::Mechanize (well, LWP...) for you, so you can do something like:-

    if ($mech->response and my $refresh = $mech->response->header('Refresh
+'))
{
    my($delay, $uri) = split /;url=/i, $refresh;

    $uri ||= $mech->uri; # No URL; reload current URL.

    sleep $delay;

    $mech->get($uri);
}

    [download]

    $delay should probably be validated to protect against malformed META refresh tags, and there's a whole other headache about potential loops if you hack WWW::Mechanize to follow refreshes automatically.

        --k.


[reply]
[d/l]
      The snippet I provided is from my test suite. I'll be first to admit that it isn't great as I've only just started hacking away with Mechanize (and wondered why I didn't start sooner ;P).

      Anyway, from a testing perspective is it not better to follow the expected url and not the url in the template? Its only a minor point but are you not then reporting on a mistaken redirect but continuing as normal otherwise? I feel this is better but would welcome your comments.

      I do like the delay bit but, for my testing purposes, I would also pass that into the function. Something like: 
       meta_refresh($mech, '/index.cgi?rm=home', 5);

      [download]
      Or whatever :). I would also then, personally, have a default delay (of some time determined by the particular project) and simply validate the delay as being correct (for the same reasons as with the URL).

      Its funny, I only wrote this function because IIS, at the time, couldn't handle HTTP redirects and would crash (no really). Its *fixed now* but I don't have the time to rework my app again :).
[reply]
[d/l]
Re: WWW::Mechanize follow meta refreshes
by jbrugger (Parson) on Apr 14, 2005 at 06:10 UTC
    As it does with JavaScript as well, see www::mechanize reloading page, so you'd remove a line (or a part of it) to stop JavsScript from loading another page.
    "We all agree on the necessity of compromise. We just can't agree on when it's necessary to compromise." - Larry Wall.
[reply]
Re: WWW::Mechanize follow meta refreshes
by mhi (Friar) on Aug 28, 2005 at 20:46 UTC
    Thanks for posting this. It was just right to get me on my way to becoming a first-time user of WWW::Mechanize.

    I've put together a little script that will do some pre-fetching on a web-application. The app has a result-cache and will output a refresh whenever it encounters a cache-miss and does its calculations.

    Instead of parsing the refresh-URLs or somesuch, I simply limited the number of refreshs the script will perform. Perhaps somemonk will find a useful snippet of code herein. 

    #!/usr/bin/perl -w
use strict;
use WWW::Mechanize; 

my $maxrefreshs=5;
my $debug=0;
my @urls=(
   "http://www.whatever.xy/cgi-bin/cgi.pl?ACTION=surnames",
   "http://www.whatever.xy/cgi-bin/cgi.pl?ACTION=unisearch&MATCHSTRING
+=foo",
   "http://www.whatever.xy/cgi-bin/cgi.pl?ACTION=path&STARTNODE=I1&END
+NODE=I1257",
); 

my $refreshs=0;
my $mech= new WWW::Mechanize; 
foreach my $url (@urls){
   while($refreshs < $maxrefreshs){
      $mech->get($url);
      my $c=$mech->content;
      $debug and print $c; 
      if($c =~/<meta\s+http-equiv="refresh"\s+content="\d+;\s*url=([^"
+]*)"/mi){
         $url=($1 or $url);
         ++$refreshs;
      }else{
         last;
      }
   }
}

    [download]
[reply]
[d/l]

Back to Cool Uses for Perl