=encoding utf-8

=head1 NAME

Unicode::GCString - String as Sequence of UAX #29 Grapheme Clusters

=head1 SYNOPSIS

    use Unicode::GCString;
    $gcstring = Unicode::GCString->new($string);
    
=head1 DESCRIPTION

Unicode::GCString treats Unicode string as a sequence of
I<extended grapheme clusters> defined by Unicode Standard Annex #29 [UAX #29].

B<Grapheme cluster> is a sequence of Unicode character(s) that consists of one
B<grapheme base> and optional B<grapheme extender> and/or
B<“prepend” character>.  It is close in that people consider as I<character>.

=head2 Public Interface

=head3 Constructors

=over 4

=item new (STRING, [KEY => VALUE, ...])

=item new (STRING, [LINEBREAK])

I<Constructor>.
Create new grapheme cluster string (Unicode::GCString object) from
Unicode string STRING.

About optional KEY => VALUE pairs see L<Unicode::LineBreak/Options>.
On second form, L<Unicode::LineBreak> object LINEBREAK controls
breaking features.

B<Note>:
The first form was introduced by release 2012.10.

=item copy

I<Copy constructor>.
Create a copy of grapheme cluster string.
Next position of new string is set at beginning.

=back

=head3 Sizes

=over 4

=item chars

I<Instance method>.
Returns number of Unicode characters grapheme cluster string includes,
i.e. length as Unicode string.

=item columns

I<Instance method>.
Returns total number of columns of grapheme clusters
defined by built-in character database.
For more details see L<Unicode::LineBreak/DESCRIPTION>.

=item length

I<Instance method>.
Returns number of grapheme clusters contained in grapheme cluster string.

=back

=head3 Operations as String

=over 4

=item as_string

=item C<">OBJECTC<">

I<Instance method>.
Convert grapheme cluster string to Unicode string explicitly.

=item cmp (STRING)

=item STRING C<cmp> STRING

I<Instance method>.
Compare strings.  There are no oddities.
One of each STRING may be Unicode string.

=item concat (STRING)

=item STRING C<.> STRING

I<Instance method>.
Concatenate STRINGs.  One of each STRING may be Unicode string.
Note that number of columns (see columns()) or grapheme clusters
(see length()) of resulting string is not always equal to sum of both
strings.
Next position of new string is that set on the left value.

=item join ([STRING, ...])

I<Instance method>.
Join STRINGs inserting grapheme cluster string.
Any of STRINGs may be Unicode string.

=item substr (OFFSET, [LENGTH, [REPLACEMENT]])

I<Instance method>.
Returns substring of grapheme cluster string.
OFFSET and LENGTH are based on grapheme clusters.
If REPLACEMENT is specified, substring is replaced by it.
REPLACEMENT may be Unicode string.

Note:
This method cannot return the lvalue, unlike built-in substr().

=back

=head3 Operations as Sequence of Grapheme Clusters

=over 4

=item as_array

=item C<@{>OBJECTC<}>

=item as_arrayref

I<Instance method>.
Convert grapheme cluster string to an array of grapheme clusters.

=item eos

I<Instance method>.
Test if current position is at end of grapheme cluster string.

=item item ([OFFSET])

I<Instance method>.
Returns OFFSET-th grapheme cluster.
If OFFSET was not specified, returns next grapheme cluster.

=item next

=item C<E<lt>>OBJECTC<E<gt>>

I<Instance method>, iterative.
Returns next grapheme cluster and increment next position.

=item pos ([OFFSET])

I<Instance method>.
If optional OFFSET is specified, set next position by it.
Returns next position of grapheme cluster string.

=back

=begin comment

=head4 Methods planned to be deprecated

=over 4

=item flag ([OFFSET, [VALUE]])

I<Instance method>.
Get or set flag value of OFFEST-th grapheme cluster.
If OFFSET was not specified, returns flag value of next grapheme cluster.
Flag value is an non-zero integer not greater than 255 and initially is 0.

Predefined flags are:

=over 4

=item Unicode::LineBreak::ALLOW_BEFORE

Allow line breaking just before this grapheme cluster.

=item Unicode::LineBreak::PROHIBIT_BEFORE

Prohibit line breaking just before this grapheme cluster.

=back

=item lbclass ([OFFSET])

I<Instance method>.
Returns Line Breaking Class (See L<Unicode::LineBreak>) of the first
character of OFFSET-th grapheme cluster.
If OFFSET was not specified, returns class of next grapheme cluster.

B<Note>:
Use lbc().

=item lbclass_ext ([OFFSET])

I<Instance method>.
Returns Line Breaking Class (See L<Unicode::LineBreak>) of the last
grapheme extender of OFFSET-th grapheme cluster.  If there are no
grapheme extenders or its class is CM, value of lbclass() is returned.

B<Note>:
Use lbcext().

=back

=end comment

=head3 Miscelaneous

=over 4

=item lbc

I<Instance method>.
Returns Line Breaking Class (See L<Unicode::LineBreak>) of the first
character of first grapheme cluster.

=item lbcext

I<Instance method>.
Returns Line Breaking Class (See L<Unicode::LineBreak>) of the last
grapheme extender of last grapheme cluster.
If there are no grapheme extenders or its class is CM, value of last
grapheme base will be returned.

=back

=head1 CAVEATS

=over 4

=item *

The grapheme cluster should not be referred to as "grapheme"
even though Larry does.

=item *

On Perl around 5.10.1, implicit conversion from Unicode::GCString object to
Unicode string sometimes let C<"utf8_mg_pos_cache_update"> cache be confused.

For example, instead of doing

    $sub = substr($gcstring, $i, $j);

do

    $sub = substr("$gcstring", $i, $j);

    $sub = substr($gcstring->as_string, $i, $j);

=item *

This module implements I<default> algorithm for determining grapheme cluster
boundaries.  Tailoring mechanism has not been supported yet.

=back

=head1 VERSION

Consult $VERSION variable.

=head2 Incompatible Changes

=over 4

=item Release 2013.10

=over 4

=item *

The new() method can take non-Unicode string argument.
In this case it will be decoded by iso-8859-1 (Latin 1) character set.
That method of former releases would die with non-Unicode inputs.

=back

=back

=head1 SEE ALSO

[UAX #29]
Mark Davis (ed.) (2009-2013).
I<Unicode Standard Annex #29: Unicode Text Segmentation>, Revisions 15-23.
L<http://www.unicode.org/reports/tr29/>.

=head1 AUTHOR

Hatuka*nezumi - IKEDA Soji <hatuka(at)nezumi.nu>

=head1 COPYRIGHT

Copyright (C) 2009-2013 Hatuka*nezumi - IKEDA Soji.

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

=cut