package Class::Fields::Fuxor; use strict; no strict 'refs'; use vars qw(@ISA @EXPORT $VERSION); use Carp::Assert; $VERSION = '0.06'; require Exporter; @ISA = qw(Exporter); @EXPORT = qw(add_fields add_field_set has_fields get_fields get_attr ); use constant TRUE => (1==1); use constant FALSE => !TRUE; use constant SUCCESS => TRUE; use constant FAILURE => !SUCCESS; use Class::Fields::Attribs; =pod =head1 NAME Class::Fields::Fuxor - Low level manipuation of object data members =head1 SYNOPSIS # As functions. use Class::Fields::Fuxor; add_fields($class, $attrib, @fields); add_field_set($class, \@fields, \@attribs); has_fields($class); $fields = get_fields($class); $fattr = get_attr($class); # As methods. package Foo; use base qw( Class::Fields::Fuxor ); Foo->add_fields($attrib, @fields); Foo->has_fields; $fields = Foo->get_fields; $fattr = Foo->get_attr; =head1 DESCRIPTION This is a module for low level manipuation of the %FIELDS hash and its accompying %fields::attr hash without actually touching them. Modules like fields.pm, base.pm and public.pm make use of this module. %FIELDS and %fields::attr are currently used to store information about the data members of classes. Since the current data inheritance system, built around pseudo-hashes, is considered a bit twitchy, it is wise to encapsulate and rope it off in the expectation that it will be replaced with something better. Typically one does not want to mess with this stuff and instead uses fields.pm and friends or perhaps Class::Fields. =cut # The %attr hash holds the attributes of the currently assigned fields # per class. The hash is indexed by class names and the hash value is # an array reference. The array is indexed with the field numbers # (minus one) and the values are integer bit masks (or undef). The # size of the array also indicates the next field index to assign for # additional fields in this class. # # BTW %attr is part of fields for legacy reasons. We alias it here to make # life easier. use vars qw(%attr); *attr = \%fields::attr; =pod =over 4 =item B add_fields($class, $attrib, @fields); Adds a bunch of @fields to the given $class using the given $attrib. For example: # Add the public fields 'this' and 'that' to the class Foo. use Class::Fields::Attribs; add_fields('Foo', PUBLIC, qw(this that)); $attrib is built from the constants in Class::Fields::Attribs =cut sub add_fields { my($proto, $attrib, @fields) = @_; add_field_set($proto, \@fields, [($attrib) x @fields]); } =pod =item B add_field_set($class, \@fields, \@attribs); Functionally similar to add_fields(), excepting that it can add a group of fields with different attributes all at once. This is necessary for the proper functioning of fields.pm. Each element in @fields matches up with one in @attribs. Obviously, the two arrays must be the same size. =cut sub add_field_set { # Read the first two parameters. The rest are field names. my($proto, $new_fields, $new_attribs) = @_; assert(@$new_fields == @$new_attribs) if DEBUG; # Quick bail out if nothing is to be added. return SUCCESS unless @$new_fields; my($class) = ref $proto || $proto; my $fields = get_fields($class); my $fattr = get_attr($class); my $next_fno = @$fattr; # Check for existing fields not belonging to base classes. # Indicates a possible module reload. if ($next_fno > $fattr->[0] and ($fields->{$new_fields->[0]} || 0) >= $fattr->[0]) { # Reset the next pointer to let the reload work. $next_fno = $fattr->[0]; } # Go through the fields and attach attributes. foreach my $idx (0..$#{$new_fields}) { my $f = $new_fields->[$idx]; my $attrib = $new_attribs->[$idx]; my $fno = $fields->{$f}; # Allow the module to be reloaded so long as field positions # have not changed. if ($fno and $fno != $next_fno) { require Carp; if ($fno < $fattr->[0]) { Carp::carp("Hides field '$f' in base class") if $^W; } else { Carp::croak("Field name '$f' already in use"); } } $fields->{$f} = $next_fno; $fattr->[$next_fno] = $attrib; $next_fno++; } } =pod =item B has_fields($class); A simple check to see if the given $class has a %FIELDS hash defined. A simple test like (defined %{"$class\::FIELDS"}) will sometimes produce typo warnings because it would create the hash if it was not present before. =cut sub has_fields { my($proto) = shift; my($class) = ref $proto || $proto; my $fglob; return ($fglob = ${"$class\::"}{"FIELDS"} and *$fglob{HASH}) ? TRUE : FALSE; } =pod =item B $fattr = get_attr($class); Get's the field attribute array for the given $class. This is roughly equivalent to $fields::attr{$class} but we put a nice wrapper around it for compatibility and readability. $fattr is an array reference containing the attributes of the fields in the given $class. Each entry in $fattr corresponds to the position indicated by the $class's %FIELDS has. For example: package Foo; use fields qw(this _that); $fattr = get_attr('Foo'); # Get the attributes for '_that' in the class 'Foo'. $that_attribs = print $fattr->[$Foo::FIELDS->{_that}]; When possible, one should avoid using this function since it exposes more implementation detail than I'd like. Class::Fields should provide most of the functionality you'll need. =cut sub get_attr { my($proto) = shift; my($class) = ref $proto || $proto; unless ( defined $attr{$class} ) { $attr{$class} = [1]; } return $attr{$class}; } =pod =item B $fields = get_fields($class); Gets a reference to the %FIELDS hash for the given $class. It will autogenerate a %FIELDS hash if one doesn't already exist. If you don't want this behavior, be sure to check beforehand with has_fields(). When possible, one should avoid using this function since it exposes more implementation detail than I'd like. Class::Fields should provide most of the functionality you'll need. =cut sub get_fields { my($proto) = shift; my($class) = ref $proto || $proto; # Shut up a possible typo warning. () = \%{$class.'::FIELDS'}; return \%{$class.'::FIELDS'}; } =pod =back =head1 AUTHOR Michael G Schwern based heavily on code liberated from the original fields.pm and base.pm. =head1 SEE ALSO L, L, L, L, L, L, L =cut return 'Maybe we should have stopped with Smalltalk.';