package BibTeX::Parser::Author;
{
  $BibTeX::Parser::Author::VERSION = '1.02';
}

use warnings;
use strict;

use BibTeX::Parser;


use overload
	'""' => \&to_string;



sub new {
	my $class = shift;

	if (@_) {
		my $self = [ $class->split(@_) ];
		return bless $self, $class;
	} else {
		return bless [], $class;
	}
}

sub _get_or_set_field {
	my ($self, $field, $value) = @_;
	if (defined $value) {
		$self->[$field] = $value;
	} else {
		return $self->[$field];
	}
}


sub first {
	shift->_get_or_set_field(0, @_);
}


sub von {
	shift->_get_or_set_field(1, @_);
}


sub last {
	shift->_get_or_set_field(2, @_);
}


sub jr {
	shift->_get_or_set_field(3, @_);
}


# Take a string and create an array [first, von, last, jr]
sub split {
    my ($self_or_class, $name) = @_;
    
    # remove whitespace at start and end of string
    $name =~ s/^\s*(.*)\s*$/$1/s;



    if (!length($name)) {
	return (undef, undef, undef, undef);
    }
    
    my @comma_separated = 
	BibTeX::Parser::_split_braced_string($name, 
					     '\s*,\s*');
    if (scalar(@comma_separated) == 0) {
	# Error?
	return (undef, undef, undef, undef);
    }

    my $first=undef;
    my $von=undef;
    my $last=undef;
    my $jr=undef;
    
    if (scalar(@comma_separated) == 1) {
	# First von Last form
	my @tokens = 	
	    BibTeX::Parser::_split_braced_string($name, '\s+');
	if (!scalar (@tokens)) {
	    return (undef, undef, undef, undef);
	}
	my ($start_von, $start_last) = _getStartVonLast (@tokens);
	if ($start_von >0) {
	    $first = join(' ', splice(@tokens,0,$start_von));
	}
	if (($start_last-$start_von) >0) {
	    $von = join(' ', splice(@tokens,0,$start_last-$start_von));
	}
	$last = join(' ',@tokens);
	return ($first, $von, $last, $jr);
    }
    # Now we work with von Last, [Jr,] First form
    if (scalar @comma_separated == 2) { # no jr
	my @tokens=  
	    BibTeX::Parser::_split_braced_string($comma_separated[1], '\s+');
	$first = join(' ', @tokens);
    } else { # jr is present
	my @tokens=  
	    BibTeX::Parser::_split_braced_string($comma_separated[1], '\s+');
	$jr = join(' ', @tokens);
	@tokens=  
	    BibTeX::Parser::_split_braced_string($comma_separated[2], '\s+');
	$first = join(' ', @tokens);
    }
    my @tokens = 	
	BibTeX::Parser::_split_braced_string($comma_separated[0], '\s+');
    my $start_last = _getStartLast(@tokens);
    if ($start_last > 0) {
	$von=join(' ', splice(@tokens,0,$start_last));
    }
    $last = join(' ',@tokens);
    return ($first, $von, $last, $jr);

}

# Return the index of the first von element and the first lastname 
# element.  If no von element, von=last

sub _getStartVonLast {
    my $length=scalar(@_);
    if ($length==1) {
	return (0,0);
    }
    my $start_von=-1;
    my $start_last=$length-1;
    for (my $i=0; $i<$length; $i++) {
	if  (_is_von_token($_[$i])) {
	    $start_von=$i;
	    last;
	}
    }
    if ($start_von== -1) { # no von part
	return($length-1, $length-1);
    }
    if ($start_von== $length-1) { # all parts but last are upper case?
	return($length-1, $length-1);
    }
    for (my $i=$start_von+1; $i<$length; $i++) {
	if (!_is_von_token($_[$i])) {
	    $start_last=$i;
	    last;
	}
    }
    return($start_von, $start_last);
}


# Return the index of the first lastname 
# element provided no first name elements are present

sub _getStartLast {
    my $length=scalar(@_);
    if ($length==1) {
	return 0;
    }
    my $start_last=$length-1;
    for (my $i=0; $i<$length; $i++) {
	if  (!_is_von_token($_[$i])) {
	    $start_last=$i;
	    last;
	}
    }
    return $start_last;
}


sub _split_name_parts {
    my $name = shift;

    if ( $name !~ /\{/ ) {
        return split /\s+/, $name;
    } else {
        my @parts;
        my $cur_token = '';
        while ( scalar( $name =~ /\G ( [^\s\{]* ) ( \s+ | \{ | \s* $ ) /xgc ) ) {
            $cur_token .= $1;
            if ( $2 =~ /\{/ ) {
                if ( scalar( $name =~ /\G([^\}]*)\}/gc ) ) {
                    $cur_token .= "{$1}";
                } else {
                    die "Unmatched brace in name '$name'";
                }
            } else {
                if ( $cur_token =~ /^{(.*)}$/ ) {
                    $cur_token = $1;
                }
                push @parts, $cur_token;
                $cur_token = '';
            }
        }
        return @parts;
    }

}


sub _get_single_author_from_tokens {
    my (@tokens) = @_;
    if (@tokens == 0) {
        return (undef, undef, undef, undef);
    } elsif (@tokens == 1) {	# name without comma
        if ( $tokens[0] =~ /(^|\s)[[:lower:]]/) { # name has von part or has only lowercase names
            my @name_parts = _split_name_parts $tokens[0];

            my $first;
            while (@name_parts && ucfirst($name_parts[0]) eq $name_parts[0] ) {
                $first .= $first ? ' ' . shift @name_parts : shift @name_parts;
            }

            my $von;
            # von part are lowercase words
            while ( @name_parts && lc($name_parts[0]) eq $name_parts[0] ) {
                $von .= $von ? ' ' . shift @name_parts : shift @name_parts;
            }

            if (@name_parts) {
                return ($first, $von, join(" ", @name_parts), undef);
            } else {
                return (undef, undef, $tokens[0], undef);
            }
        } else {
            if ( $tokens[0] !~ /\{/ && $tokens[0] =~ /^((.*)\s+)?\b(\S+)$/) {
                return ($2, undef, $3, undef);
            } else {
                my @name_parts = _split_name_parts $tokens[0];
                return ($name_parts[0], undef, $name_parts[1], undef);
            }
        }

    } elsif (@tokens == 2) {
        my @von_last_parts = _split_name_parts $tokens[0];
        my $von;
        # von part are lowercase words
        while ( @von_last_parts && lc($von_last_parts[0]) eq $von_last_parts[0] ) {
            $von .= $von ? ' ' . shift @von_last_parts : shift @von_last_parts;
        }
        return ($tokens[1], $von, join(" ", @von_last_parts), undef);
    } else {
        my @von_last_parts = _split_name_parts $tokens[0];
        my $von;
        # von part are lowercase words
        while ( @von_last_parts && lc($von_last_parts[0]) eq $von_last_parts[0] ) {
            $von .= $von ? ' ' . shift @von_last_parts : shift @von_last_parts;
        }
        return ($tokens[2], $von, join(" ", @von_last_parts), $tokens[1]);
    }

}




sub to_string {
	my $self = shift;

	if ($self->jr) {
		return ($self->von ? $self->von . " " : '') . " " . $self->last . ", " . $self->jr . ", " . $self->first;
	} else {
		return ($self->von ? $self->von . " " : '') . $self->last . ($self->first ? ", " . $self->first : '');
	}
}


# Return 1 if the first letter on brace level 0 is lowercase
sub _is_von_token {
    my $string = shift;
    while ($string =~ 
	   s/^(\\[[:alpha:]]+\{|\{|\\[[:^alpha:]]?|[[:^alpha:]])//) {
	if ($1 eq '{' ) {
	    my $numbraces=1;
	    while ($numbraces !=0 && length($string)) {
		my $symbol = substr($string, 0, 1);
		if ($symbol eq '{') {
		    $numbraces ++;
		} elsif ($symbol eq '}') {
		    $numbraces --;
		}
		$string = substr($string,1);
	    }
	}
    }

    if (length $string ) {
	my $symbol = substr($string, 0, 1);
	if (lc($symbol) eq $symbol) {
	    return 1;
	} else {
	    return 0;
	}
    } else {
	return 1;
    }

}

1; # End of BibTeX::Entry

__END__
=pod

=head1 NAME

BibTeX::Author - Contains a single author for a BibTeX document.

=head1 SYNOPSIS

This class ist a wrapper for a single BibTeX author. It is usually created
by a BibTeX::Parser.

    use BibTeX::Parser::Author;

    my $entry = BibTeX::Parser::Author->new($full_name);
    
    my $firstname = $author->first;
    my $von	  = $author->von;
    my $last      = $author->last;
    my $jr	  = $author->jr;

    # or ...
    
    my ($first, $von, $last, $jr) = BibTeX::Author->split($fullname);



=head1 FUNCTIONS

=head2 new

Create new author object. Expects full name as parameter.

=head2 first

Set or get first name(s).

=head2 von

Set or get 'von' part of name.

=head2 last

Set or get last name(s).

=head2 jr

Set or get 'jr' part of name.

=head2 split

Split name into (firstname, von part, last name, jr part). Returns array
with four strings, some of them possibly empty.

=head2 to_string

Return string representation of the name.

=head1 NOTES

BibTeX allows three representations of a person's name:

=over 4

=item 1.

First von Last

=item 2.

von Last, First

=item 3.

von Last, Jr, First

=back

The module always converts the first form to the second of third one
to allow simple string comparisons.  

The algorithm to determine the von part is the following: von part
consists of tokens where the first letter at brace level 0 is in lower case.
Anything in a "special characters" is on brace level 0.  Thus the following
tokens are considered von parts:  C<von>, C<\NOOP{von}Von>, and
the following token is not: C<{von}>

=head1 VERSION

version 1.02


=head1 AUTHOR

Gerhard Gossen <gerhard.gossen@googlemail.com> and
Boris Veytsman <boris@varphi.com>

=head1 COPYRIGHT AND LICENSE

This software is copyright (c) 2013--2016 by Gerhard Gossen and Boris Veytsman.

This is free software; you can redistribute it and/or modify it under
the same terms as the Perl 5 programming language system itself.

=cut