HTML::Form Phenomena

dReKurCe has asked for the wisdom of the Perl Monks concerning the following question:

Grettings Monks; When using two slightly different strategies :one that involves passing the result of a Form query into a sub and one outputing the result immediately. Here's the code and the results.

#! /usr/bin/perl

use LWP;
use HTTP::Request;
use HTTP::Response;
use HTML::Form;

$con=new LWP::UserAgent;
$con->agent('n3u7digit');
$request=new HTTP::Request(GET=>"http://www.<site>.org");
$reply=$con->request($request);

@forms=HTML::Form->parse($reply);

for $formh(@forms){@keys=keys %$formh;
                        for $name(@keys){
                        print "\t---$name\n";
                        }
                   }
[download]

....the second version ..

for $formh(@forms){@keys=values %$formh;
                        for $name(@keys){
                        process($name);
                        print "--$name--\n";


}                                                                     
+           sub process{$return=shift;return $return;}
[download]

the output: CASE 1

---enctype
---action
---method
---attr
---inputs
---enctype
---action
---method
---attr
---inputs
[download]

CASE 2

--application/x-www-form-urlencoded--
--http:/<site>.org/--
--GET--
--HASH(0x84d3d4c)--
--ARRAY(0x85190e4)--
--application/x-www-form-urlencoded--
--http://<site>.org/en/video?destination=node%xxxxxx--
--POST---
--HASH(0x852232c)--
--ARRAY(0x852ac40)--
[download]

So the utility of a closure seems evident, but why?

#! /usr/bin/perl

use LWP;
use HTTP::Request;
use HTTP::Response;
use HTML::Form;

$con=new LWP::UserAgent;
$con->agent('n3u7digit');
$request=new HTTP::Request(GET=>"http://www.<site>.org");
$reply=$con->request($request);

@forms=HTML::Form->parse($reply);



for $formh(@forms){@keys=values %$formh;
                        for $name(@keys){
                        process($name);
                        print "--$name--\n";
           }
sub process{
        $return=shift;
        return $return;
        }
[download]

Comment on HTML::Form Phenomena Select or Download Code

Replies are listed 'Best First'.
Re: HTML::Form Phenomena by eric256 (Parson) on Mar 28, 2007 at 13:37 UTC
You are missing a closing } in your example, that would lead me to beleive that this isn't the code actualy producing that result....or just a cut and paste error, either way it would be good to add your } and enlose all your code and results in `<code>`</code> tags. I am wondering if you are also misusing HTML::Form the documentation says to use $form->inputs() to get at the fields and your are accessing its hash directly. I would think that the behavour of objects when you don't follow their interface is undefined as you are seeing. I am willing to bet however that the output you are getting is the result of some sort of string overloading in the first case that you arn't getting in the second. ___________ Eric Hodges	[reply] [d/l]
Re: HTML::Form Phenomena by Anno (Deacon) on Mar 29, 2007 at 08:35 UTC
In your first code example you are looping over the keys of a hash `@keys=keys %$formh;` [download] In the second, you loop over the values `@keys=values %$formh;` [download] Accordingly the output looks different. The call to `process()` in the second case does nothing and has nothing to do with the difference. It looks like you are running without `strict` and `warnings`. Switch them on, they're helpful. Anno Update: Added pointer to `strict` and `warnings`	[reply] [d/l] [select]