=head1 NAME

PPIx::Regexp::Token::Assertion - Represent a simple assertion.

=head1 SYNOPSIS

 use PPIx::Regexp::Dumper;
 PPIx::Regexp::Dumper->new( 'qr{\bfoo\b}smx' )
     ->print();

=head1 INHERITANCE

C<PPIx::Regexp::Token::Assertion> is a
L<PPIx::Regexp::Token|PPIx::Regexp::Token>.

C<PPIx::Regexp::Token::Assertion> has no descendants.

=head1 DESCRIPTION

This class represents one of the simple assertions; that is, those that
are not defined via parentheses. This includes the zero-width assertions
C<^>, C<$>, C<\b>, C<\B>, C<\A>, C<\Z>, C<\z> and C<\G>, as well as:

=over

=item * The C<\z> assertion added in Perl 5.005,

=item * The C<\K> assertion added in Perl 5.009005,

=item * The C<\b{gcb}> assertion (and friends) added in Perl 5.021009.
Similar braced constructions (like C<\b{foo}>) are unknown tokens.

=back

=head1 METHODS

This class provides no public methods beyond those provided by its
superclass.

=cut

package PPIx::Regexp::Token::Assertion;

use strict;
use warnings;

use base qw{ PPIx::Regexp::Token };

use PPIx::Regexp::Constant qw{
    COOKIE_CLASS
    COOKIE_LOOKAROUND_ASSERTION
    LITERAL_LEFT_CURLY_ALLOWED
    MINIMUM_PERL
    TOKEN_LITERAL
    TOKEN_UNKNOWN
    @CARP_NOT
};

use constant KEEP_EXPLANATION	=> 'In s///, keep everything before the \\K';

our $VERSION = '0.088';

# Return true if the token can be quantified, and false otherwise
# sub can_be_quantified { return };

my @braced_assertions = (
    [ qr< \\ [bB] [{] (?: g | gcb | wb | sb ) [}] >smx, '5.021009' ],
    [ qr< \\ [bB] [{] (?: lb ) [}] >smx, '5.023007' ],
    [ qr< \\ [bB] [{] .*? [}] >smx, undef, TOKEN_UNKNOWN,
	{ error => 'Unknown bound type' },
    ],
);

=head2 is_matcher

This method returns a true value because an assertion actually matches
something.

=cut

sub is_matcher { return 1; }

sub perl_version_introduced {
    my ( $self ) = @_;
    return ( $self->{perl_version_introduced} ||=
	$self->_perl_version_introduced() );
}

{

    my %perl_version_introduced = (
	'\\K' => '5.009005',
	'\\z' => '5.005',
    );

    sub _perl_version_introduced {
	my ( $self ) = @_;
	my $content = $self->content();
	foreach my $item ( @braced_assertions ) {
	    $content =~ m/ \A $item->[0] \z /smx
		and return $item->[1];
	}
	return $perl_version_introduced{ $content } || MINIMUM_PERL;
    }

}

sub perl_version_removed {
    my ( $self ) = @_;
    return ( $self->{perl_version_removed} ||=
	$self->_perl_version_removed() );
}

sub _perl_version_removed {
    my ( $self ) = @_;
    if ( '\\K' eq $self->content() ) {
	my $parent = $self;
	while ( $parent = $parent->parent() ) {
	    $parent->isa( 'PPIx::Regexp::Structure::Assertion' )
		and return '5.031003';
	}
    }
    return $self->SUPER::perl_version_removed();
}

{
    my %explanation = (
	'$'	=> 'Assert position is at end of string or newline',
	'\\A'	=> 'Assert position is at beginning of string',
	'\\B'	=> 'Assert position is not at word/nonword boundary',
	'\\B{gcb}'	=> 'Assert position is not at grapheme cluster boundary',
	'\\B{g}'	=> 'Assert position is not at grapheme cluster boundary',
	'\\B{lb}'	=> 'Assert position is not at line boundary',
	'\\B{sb}'	=> 'Assert position is not at sentence boundary',
	'\\B{wb}'	=> 'Assert position is not at word boundary',
	'\\G'	=> 'Assert position is at pos()',
	'\\K'	=> KEEP_EXPLANATION,
	'\\Z'	=> 'Assert position is at end of string, or newline before end',
	'\\b'	=> 'Assert position is at word/nonword boundary',
	'\\b{gcb}'	=> 'Assert position is at grapheme cluster boundary',
	'\\b{g}'	=> 'Assert position is at grapheme cluster boundary',
	'\\b{lb}'	=> 'Assert position is at line boundary',
	'\\b{sb}'	=> 'Assert position is at sentence boundary',
	'\\b{wb}'	=> 'Assert position is at word boundary',
	'\\z'	=> 'Assert position is at end of string',
	'^'	=> 'Assert position is at beginning of string or after newline',
    );

    sub __explanation {
	return \%explanation;
    }
}

# An un-escaped literal left curly bracket can always follow this
# element.
sub __following_literal_left_curly_disallowed_in {
    return LITERAL_LEFT_CURLY_ALLOWED;
}

# By logic we should handle '$' here. But
# PPIx::Regexp::Token::Interpolation needs to process it to see if it is
# a sigil. If it is not, that module is expected to make it into an
# assertion. This is to try to keep the order in which the tokenizers
# are called non-critical, and try to keep all processing for a
# character in one place. Except for the back slash, which gets in
# everywhere.
#
## my %assertion = map { $_ => 1 } qw{ ^ $ };
my %assertion = map { $_ => 1 } qw{ ^ };
my %escaped = map { $_ => 1 } qw{ b B A Z z G K };

sub __PPIX_TOKENIZER__regexp {
    my ( undef, $tokenizer, $character ) = @_;

    # Inside a character class, these are all literals.
    my $make = $tokenizer->cookie( COOKIE_CLASS ) ?
	TOKEN_LITERAL :
	__PACKAGE__;

    # '^' and '$'. Or at least '^'. See note above for '$'.
    $assertion{$character}
	and return $tokenizer->make_token( 1, $make );

    $character eq '\\' or return;

    defined ( my $next = $tokenizer->peek( 1 ) ) or return;

    # Handle assertions of the form \b{gcb} and friends, introduced in
    # Perl 5.21.9. These are not recognized inside square bracketed
    # character classes, where \b is not an assertion but a backspace
    # character.
    if ( __PACKAGE__ eq $make ) {	# Only outside [...]
	foreach my $item ( @braced_assertions ) {
	    my $end = $tokenizer->find_regexp( qr/ \A $item->[0] /smx )
		or next;
	    $item->[2]
		or return $end;
	    return $tokenizer->make_token( $end, $item->[2], $item->[3] );
	}
    }

    # We special-case '\K' because it was retracted inside look-around
    # assertions in 5.31.3.
    if ( 'K' eq $next && __PACKAGE__ eq $make &&
	$tokenizer->__cookie_exists( COOKIE_LOOKAROUND_ASSERTION ) ) {
	return $tokenizer->make_token( 2, $make, {
		perl_version_removed	=> '5.031003',
		explanation		=> KEEP_EXPLANATION .
		    '; retracted inside look-around assertion',
	    },
	);
    }

    $escaped{$next}
	and return $tokenizer->make_token( 2, $make );

    return;
}

1;

__END__

=head1 SUPPORT

Support is by the author. Please file bug reports at
L<https://rt.cpan.org/Public/Dist/Display.html?Name=PPIx-Regexp>,
L<https://github.com/trwyant/perl-PPIx-Regexp/issues>, or in
electronic mail to the author.

=head1 AUTHOR

Thomas R. Wyant, III F<wyant at cpan dot org>

=head1 COPYRIGHT AND LICENSE

Copyright (C) 2009-2023 by Thomas R. Wyant, III

This program is free software; you can redistribute it and/or modify it
under the same terms as Perl 5.10.0. For more details, see the full text
of the licenses in the directory LICENSES.

This program is distributed in the hope that it will be useful, but
without any warranty; without even the implied warranty of
merchantability or fitness for a particular purpose.

=cut

# ex: set textwidth=72 :