[Catalyst] FormValidator usage / enhancements

[Will Hawes]

Carl Franks wrote:
> I'm experimenting with how best to use the FormValidator plug-in.
> I think it's far too complicated, particularly on pages where it's not
> appropriate to use it in conjunction with FillInForm and you actually
> have to use the API !
> 
> What I've done at the moment is create a plug-in which subclasses
> Catalyst::Plugin::FormValidator and looks like this:
> 
> package Catalyst::Plugin::FormValidator::MergeRequest;
> use strict;
> use NEXT;
> use base 'Catalyst::Plugin::FormValidator';
> 
> sub form {
>     my $c = shift;
> 
>     if (@_) {
>         $c->NEXT::form( $c, @_ );
> 
>         # logic here
>         # see below for discussion
>     }
> 
>     return $c->{form};
> }
> 
> 1;
> 
> What I want the 'login' part to do is:
> For 'invalid' input, delete it from the request parameters.
> For 'valid' input, put it into the request parameters (because it may
> have been changed by FormValidator filters).
> 
> To do this, is it "acceptable" to change values and delete keys from
> the hash-ref returned by $c->req->parameters() ? Will this break
> anything?
> 
> I'm also unsure about 'unknown' input. As the Data::FormValidator docs
> say, "Whether or not this indicates an error in the user input is
> application dependant".
> Maybe add another method to optionally strip unknowns from the request
> parameters?
> $c->delete_unknown_params
> 
> The purpose of this, is so only the 'auto' routine needs to deal with
> $c->form stuff such as invalid and valid
> The action routine can simply use $c->req->params as normal, assuming
> that all is well.
> 
> Thoughts? Ideas?
> 
> Carl
> 
> _______________________________________________
> Catalyst mailing list
> Catalyst at lists.rawmode.org
> http://lists.rawmode.org/mailman/listinfo/catalyst

I'm not sure I understand what you're trying to do, but you should 
probably consider making a copy of $c->req->params locally if you want 
to modify it.

Modifying c->req->params sounds like a recipe for disaster. If and when 
it causes anything to break it will make debugging much harder.