The London Perl and Raku Workshop takes place on 26th Oct 2024. If your company depends on Perl, please consider sponsoring and/or attending.

NAME

Class::DBI::FromCGI - Update Class::DBI data using CGI::Untaint

SYNOPSIS

  package Film;
  use Class::DBI::FromCGI;
  use base 'Class::DBI'; 
  # set up as any other Class::DBI class.

  __PACKAGE__->untaint_columns(
    printable => [qw/Title Director/],
    integer   => [qw/DomesticGross NumExplodingSheep],
    date      => [qw/OpeningDate/],
  );

  # Later on, over in another package ...

  my $h = CGI::Untaint->new;
  my $film = Film->retrieve('Godfather II');
  unless ($film->update_from_cgi($h => @columns_to_update)) {
    my %errors = $film->cgi_update_errors;
    while (my ($field, $problem) = each %errors) {                              
      warn "Problem with $field: $problem\n";
    }
  }

DESCRIPTION

Lots of times, Class::DBI is used in web-based applications. (In fact, coupled with a templating system that allows you to pass objects, such as Template::Toolkit, Class::DBI is very much your friend for these.)

And, as we all know, one of the most irritating things about writing web-based applications is the monotony of writing much of the same stuff over and over again. And, where there's monotony there's a tendency to skip over stuff that we all know is really important, but is a pain to write - like Taint Checking and sensible input validation. (Especially as we can still show a 'working' application without it!). So, we now have CGI::Untaint to take care of a lot of that for us.

It so happens that CGI::Untaint also plays well with Class::DBI. All you need to do is to 'use Class::DBI::FromCGI' in your class (or in your local Class::DBI subclass that all your other classes inherit from. You do do that, don't you?).

Then, in each class in which you want to use this, you declare how you want to untaint each column:

  __PACKAGE__->untaint_columns(
    printable => [qw/Title Director/],
    integer   => [qw/DomesticGross NumExplodingSheep],
    date      => [qw/OpeningDate/],
  );

(where the keys are the CGI::Untaint package to be used, and the values a listref of the relevant columns).

Then, when you want to update based on the values coming in from a web-based form, you just call:

  $obj->update_from_cgi($h => @columns_to_update);

If every value passed in gets through the CGI::Untaint process, the object will be updated (but not committed, in case you want to do anything else with it). Otherwise the update will fail (there are no partial updates), and $obj->cgi_update_errors will tell you what went wrong (as a hash of problem field => error from CGI::Untaint).

Doesn't that make your life so much easier?

NOTE

Don't try to update the value of your primary key. Class::DBI doesn't like that. If you try to do this it will be silently skipped.

Note that this means, on the other hand, that you can do:

  $obj->update_from_cgi($h => $obj->columns('All'));

In fact, this is the behaviour if you don't pass it any columns at all, and just do: $obj->update_from_cgi($h);

ANOTHER NOTE

If you haven't set up any 'untaint_column' information for a column which you later attempt to untaint, then we try to call $self->column_type to ascertain the default handler to use. Currently this will only use if you're using Class::DBI::mysql, and only for certain column types.

SEE ALSO

Class::DBI. CGI::Untaint. Template.

AUTHOR

Tony Bowden. <tmtm@kasei.com>.

FEEDBACK

I'd love to hear from you if you start using this. I'd particularly like to hear any suggestions as to how to make it even better / easier etc.

COPYRIGHT

Copyright (C) 2001 Kasei. All rights reserved.

This module is free software; you can redistribute it and/or modify it under the same terms as Perl itself.