use warnings; use strict; package Jifty::Result; =head1 NAME Jifty::Result - Outcome of running a L =head1 DESCRIPTION C encapsulates the outcome of running a L. Results are also stored on the framework's L object. =cut use base qw/Jifty::Object Class::Accessor::Fast/; __PACKAGE__->mk_accessors(qw(failure action_class message _content)); =head2 new Construct a new action result. This is done automatically when the action is created, and can be accessed via the L. =cut sub new { my $class = shift; my $self = bless {}, $class; $self->failure(0); $self->_content({}); return $self; } =head2 failure [BOOL] Gets or sets if the action succeeded or failed. =head2 success [BOOL] Gets or sets if the action succeeded or failed -- this is an alternate interface from C but has the same effect. =cut sub success { my $self = shift; return 0 if $self->failure(map {not $_} @_); return 1; } =head2 action_class [MESSAGE] Returns the class for the action that this result came from. =head2 message [MESSAGE] Gets or sets the action's response message. This is an informational textual description of the outcome of the action. =head2 error [ERROR] Gets or sets the action's error response. This is an informational textual description of what went wrong with the action, overall. This also automatically sets the result to be a L. =cut sub error { my $self = shift; $self->failure(1) if @_ and $_[0]; $self->{error} = shift if @_; return $self->{error}; } =head2 field_error FIELD [ERROR] [OPTIONS] Gets or sets the error string for a specific field on the action. This also automatically sets the result to be a failure. C is an optional set of key-value pairs; the only currently supported option is C, which sets the L for this field. =cut sub field_error { my $self = shift; my $field = shift; $self->failure(1) if @_ and $_[0]; $self->{field_errors}{ $field } = shift if @_; my %args = @_; $self->{ajax_force_validate}{ $field } = $args{force} if exists $args{force}; return $self->{field_errors}{ $field }; } =head2 field_errors Returns a hash which maps L name to error. =cut sub field_errors { my $self = shift; return %{$self->{field_errors} || {}}; } =head2 field_warning FIELD [WARNING] [OPTIONS] Gets or sets the warning string for a specific field on the action. C is an optional set of key-value pairs; the only currently supported option is C, which sets the L for this field. =cut sub field_warning { my $self = shift; my $field = shift; $self->{field_warnings}{ $field } = shift if @_; my %args = @_; $self->{ajax_force_validate}{ $field } = $args{force} if exists $args{force}; return $self->{field_warnings}{ $field }; } =head2 field_warnings Returns a hash which maps L name to warning. =cut sub field_warnings { my $self = shift; return %{$self->{field_warnings} || {}}; } =head2 ajax_force_validate FIELD [VALUE] Gets or sets the flag which determines if warnings and errors are set using ajax validation, even if the field is empty. By default, validation warnings and errors are I shown for empty fields, as yelling to users about mandatory fields they've not gotten to yet is poor form. You can use this method to force ajax errors to show even on empty fields. =cut sub ajax_force_validate { my $self = shift; my $field = shift; $self->{ajax_force_validate}{ $field } = shift if @_; return $self->{ajax_force_validate}{$field}; } =head2 field_canonicalization_note FIELD [NOTE] Gets or sets a canonicalization note for a specific field on the action. =cut sub field_canonicalization_note { my $self = shift; my $field = shift; $self->{field_canonicalization_notes}{ $field } = shift if @_; return $self->{field_canonicalization_notes}{ $field }; } =head2 field_canonicalization_notes Returns a hash which maps L name to canonicalization notes. =cut sub field_canonicalization_notes { my $self = shift; return %{$self->{field_canonicalization_notes} || {}}; } =head2 content [KEY [, VALUE]] Gets or sets the content C. This is used when actions need to return values. If not C is passed, it returns an anonymous hash of all of the C and C pairs. =cut sub content { my $self = shift; return $self->_content unless @_; my $key = shift; $self->_content->{$key} = shift if @_; return $self->_content->{$key}; } =head2 as_hash This returns the results as a hash to be given directly to the end user (usually via REST or webservices). The difference between C<< $result->as_hash >> and C<%$result> is that the latter will expand everything as deeply as possible. The former won't inflate C columns, among other things. =cut sub as_hash { my $self = shift; my $out = { success => $self->success, failure => $self->failure, action_class => $self->action_class, message => $self->message, error => $self->error, field_errors => { $self->field_errors }, field_warnings => { $self->field_warnings }, content => $self->_recurse_object_to_data($self->content), }; for (keys %{$out->{field_errors}}) { delete $out->{field_errors}->{$_} unless $out->{field_errors}->{$_}; } for (keys %{$out->{field_warnings}}) { delete $out->{field_warnings}->{$_} unless $out->{field_warnings}->{$_}; } return $out; } sub _recurse_object_to_data { my $self = shift; my $o = shift; return $o if !ref($o); if (ref($o) eq 'ARRAY') { return [ map { $self->_recurse_object_to_data($_) } @$o ]; } elsif (ref($o) eq 'HASH') { my %h; $h{$_} = $self->_recurse_object_to_data($o->{$_}) for keys %$o; return \%h; } return $self->_object_to_data($o); } sub _object_to_data { my $self = shift; my $o = shift; if ($o->can('jifty_serialize_format')) { return $o->jifty_serialize_format($self); } # As the last resort, return the object itself and expect the # $accept-specific renderer to format the object as e.g. YAML or JSON data. return $o; } 1;