Context Note: I am releasing a succession of around 15 Perl 5 object 
modules.  Six of these are complete and documented, and will be 
submitted quickly.  The others are working but not fully documented, 
so I am holding them back for the moment.  All of these modules, with 
info on how they can be used together (and working examples), are 
available at http://www.DarrenDuncan.net.  They have temporary names 
in the DDuncan::* name space.  They all require 5.004 for 
consistency, although some can do with less.
-----------------------------------------------

Here is #3:

Name                 DSLI  Description 
Info
-------------------  ---- 
----------------------------------------------  -------
Method::ParamParser  bdpO  complex parameter parsing for subclass 
methods  DUNCAND

If you have suggestions of alternate names for this module, I would 
be happy to hear them.  Likewise, I appreciate suggestions for a 
better brief description for use in the module list.

For a good description of my module, I have provided part of its POD 
at the end of this letter.  The rest of the POD is on my site.

Currently, at least 2 of my other modules use this one.

// Darren Duncan

----------------------------------------------

=head1 NAME

DDuncan::MethodParamParser - Perl inheritable module for tidying up 
the parameter
lists passed to methods.

I<Please note that the path name of this module is temporary, something that
works until a more appropriate and integrated name can be found.  A possibility
could be "Method::Params">

=head1 DEPENDENCIES

=head2 Perl Version

5.004

=head2 Standard Modules

=item I<none>

=head2 Nonstandard Modules

=item I<none>

=head1 SYNOPSIS

        use DDuncan::MethodParamParser 1.01;
        @ISA = qw( DDuncan::MethodParamParser );

        sub textfield {
                my $self = shift( @_ );
                my $rh_params = $self->params_to_hash( \@_, 0,
                        [ 'name', 'value', 'size', 'maxlength' ],
                        { 'default' => 'value' } );
                $rh_params->{'type'} = 'text';
                return( $self->make_html_tag( 'input', $rh_params ) );
        }

        sub AUTOLOAD {
                my $self = shift( @_ );
                my $rh_params = $self->params_to_hash( \@_, 0, 
'text', {}, 'text' );
                my $ra_text = delete( $rh_params->{'text'} );
                $AUTOLOAD =~ m/([^:]*)$/;
                my $tag_name = $1;
                return( $self->make_html_tag( $tag_name, $rh_params, 
$ra_text ) );
        }

=head1 DESCRIPTION

This Perl 5 object class implements two methods which inherited classes can use
to tidy up parameter lists for their own methods and functions.  The 
two methods
differ in that one returns a HASH ref containing named parameters and the other
returns an ARRAY ref containing positional parameters.

Both methods can process the same kind of input parameter formats:

=item I<empty list>

=item value

=item value1, value2, ...

=item name1 => value1, name2 => value2, ...

=item -name1 => value1, -name2 => value2, ...

=item { -name1 => value1, name2 => value2, ... }

=item { name1 => value1, -name2 => value2, ... }, valueR

Those examples included single or multiple positional parameters, single or
multiple named parameters, and a HASH ref containing named parameters (with
optional "remaining" value afterwards).  That list of input variations is not
exhaustive.  Named parameters can either be prefixed with "-" or left natural.

We assume that the parameters are named when either they come as a HASH ref or
the first parameter begins with a "-".  We assume that they are positional if
there is exactly one of them.  Otherwise we are in doubt and rely on 
an optional
argument to the tidying method that tells us which to guess by default.

We assume that any "value" may be an array ref (aka "multiple" values under the
same name) and hence we don't do anything special with them, passing 
them as is.

If the source and destination are both positional, then they are identical.

=head1 SYNTAX

This class does not export any functions or methods, so you need to call them
using indirect notation.  This means using B<Class->function()> for 
functions and
B<$object->method()> for methods.  If you are inheriting this class 
for your own
modules, then that often means something like B<$self->method()>.

However, if you feel like breaking the "indirect notation" rules, you can still
call the method using a dummy value instead of the implicit 
object/class that is
usually passed.  This won't break anything as the object isn't used.  However,
this practice is depreciated as of Perl 5.004 and probably won't work later.

=head1 FUNCTIONS AND METHODS

=head2 params_to_hash( SOURCE, DEF, NAMES[, RENAME[, REM]] )
=head2 params_to_array( SOURCE, DEF, NAMES[, RENAME[, REM]] )

=head1 PARAMETERS

The arguments for the above methods are the same, so they are 
discussed together
here:

=item

The first argument, SOURCE, is an ARRAY ref containing the original parameters
that were passed to the method which calls this one.  It is safe to pass "\@_"
because we don't modify the argument at all.  If SOURCE isn't a valid ARRAY ref
then its default value is [].

=item

The second argument, DEF, is a boolean/scalar that tells us whether, when in
doubt over whether SOURCE is in positional or named format, what to guess by
default.  A value of 0, the default, means we guess named, and a 
value of 1 means
we assume positional.

=item

The third argument, NAMES, is an ARRAY ref (or SCALAR) that provides 
the names to
use when SOURCE and our return value are not in the same format (named or
positional).  This is because positional parameters don't know what their names
are and named parameters (hashes) don't know what order they belong 
in; the NAMES
array provides the missing information to both.  The first name in 
NAMES matches
the first value in a positional SOURCE, and so-on.  Likewise, the order of
argument names in NAMES determines the sequence for positional output when the
SOURCE is named.

=item

The optional fourth argument, RENAME, is a HASH ref that allows us to 
interpret a
variety of names from a SOURCE in named format as being aliases for 
one enother.
The keys in the hash are names to look for and the values are what to 
rename them
to.  Keys are matched irregardless of whether the SOURCE names have 
"-" in front
of them or not.  If several SOURCE names are renamed to the same hash 
value, then
all but one are lost; the SOURCE should never contain more than one 
alias for the
same parameter anyway.  One way to explicitely delete a parameter is 
to rename it
with "", as parameters with that name are discarded.

=item

The optional fifth argument, REM, is only used in circumstances where the first
element of SOURCE is a HASH ref containing the actual named parameters that
SOURCE would otherwise be.  If SOURCE has a second, "remaining" 
element following
the HASH ref, then REM says what its name is.  Remaining parameters 
with the same
name as normal parameters (post renaming and "-" substitution) take precedence.
The default value for REM is "", and it is discarded unless renamed.

Reply via email to