I am attempting to learn WWW::Mechanize in order to avoid paying for API access from a large financial institution. If my app can access their web interface securely then the API is unneccessary. The modules in libwww-perl and HTTP::Headers etc are fresh from CPAN. My sample code:

#!/usr/local/bin/perl
use strict;
use Carp;
use warnings;
use WWW::Mechanize;
my $mech = WWW::Mechanize->new( autocheck => 1 );

my $url = 'https://secure.secret.com/login/login.cfm?someParam=foo';

$mech->get( $url );

print $mech->uri() . "\nStep 1 done. \n";

$mech->submit_form(
     form_number => 0, 
     fields      => {
        login    => 'xxxxx',
        password    => 'xxxx'
    }
);

print $mech->uri() . "\n\nDone.";
[download]

The small sample code above produces the following output on the command line:

https://secure.secret.com/login/login.cfm?someParam=foo
Step 1 done.
Can't locate object method "remove_content_headers" via package "HTTP::Headers" at (eval 14) line 1.

So the get() works on the https url but the object we grab with submit_form() throws an error. Does anyone understand why this should be the case? The source looks like this:

sub remove_content_headers
{
    my $self = shift;
    unless (defined(wantarray)) {
    # fast branch that does not create return object
    delete @$self{grep $entity_header{$_} || /^content-/, keys %$self}
+;
    return;
    }

    my $c = ref($self)->new;
    for my $f (grep $entity_header{$_} || /^content-/, keys %$self) {
    $c->{$f} = delete $self->{$f};
    }
    $c;
}
[download]