File Coverage

blib/lib/XML/Grammar/FictionBase/FromProto/Parser/XmlIterator.pm

Criterion	Covered	Total	%
statement	70	323	21.6
branch	5	80	6.2
condition	0	9	0.0
subroutine	20	78	25.6
pod	1	1	100.0
total	96	491	19.5

package XML::Grammar::FictionBase::FromProto::Parser::XmlIterator;

use strict;
use warnings;

use Carp ();

use MooX 'late';

use XML::Grammar::Fiction::Err;
use XML::Grammar::Fiction::Struct::Tag;
use XML::Grammar::FictionBase::Event;

use XML::Grammar::Fiction::FromProto::Node::WithContent;
use XML::Grammar::Fiction::FromProto::Node::Element;
use XML::Grammar::Fiction::FromProto::Node::List;
use XML::Grammar::Fiction::FromProto::Node::Text;
use XML::Grammar::Fiction::FromProto::Node::Saying;
use XML::Grammar::Fiction::FromProto::Node::Description;
use XML::Grammar::Fiction::FromProto::Node::Paragraph;
use XML::Grammar::Fiction::FromProto::Node::InnerDesc;
use XML::Grammar::Fiction::FromProto::Node::Comment;

extends("XML::Grammar::FictionBase::FromProto::Parser::LineIterator");

has "_tags_stack" =>
(
    isa => "ArrayRef",
    is => "rw",
    default => sub { [] },
);


sub _get_tag
{
    my ($self, $idx) = @_;

    return $self->_tags_stack->[$idx];
}

sub _tags_stack_is_empty
{
    my $self = shift;

    return (! @{$self->_tags_stack});
}

sub _grep_tags_stack
{
    my $self = shift;
    my $cb = shift;

    return grep { $cb->($_) } @{$self->_tags_stack};
}

sub _push_tag
{
    my $self = shift;

    push( @{$self->_tags_stack}, @_);

    return;
}

sub _pop_tag
{
    my $self = shift;

    return pop(@{$self->_tags_stack});
}

has "_events_queue" =>
(
    isa => "ArrayRef",
    # isa => "ArrayRef",
    is => "rw",
    default => sub { []; },
);


sub _clear_events
{
    my $self = shift;

    $self->_events_queue([]);

    return;
}

sub _no_events
{
    my $self = shift;

    return (! @{$self->_events_queue});
}

sub _enqueue_event
{
    my $self = shift;
    my $event = shift;

    if (@_) {
        Carp::confess("More than one argument.");
    }

    push( @{$self->_events_queue}, $event);

    return;
}

sub _extract_event
{
    my $self = shift;

    return shift(@{$self->_events_queue});
}

has '_ret_tag' =>
(
    is => "rw",
    # TODO : add isa.
    predicate => "_has_ret_tag",
    clearer => "_clear_ret_tag",
);

# Whether we are inside a paragraph or not.
has "_in_para" => (isa => "Bool", is => "rw", default => 0,);

has '_tag_names_to_be_handled' =>
(
    is => 'ro',
    isa => 'HashRef[Bool]',
    lazy => 1,
    builder => '_build_tag_names_to_be_handled',
);

sub _build_tag_names_to_be_handled
{
    my $self = shift;

    return { map { $_ => 1 } @{$self->_list_valid_tag_events} };
}

sub _get_id_regex
{
    return qr{[a-zA-Z_\-]+};
}

sub _top_tag
{
    my $self = shift;
    return $self->_get_tag(-1);
}

sub _add_to_top_tag
{
    my ($self, $child) = @_;

    $self->_top_tag->append_child($child);

    return;
}

# TODO : Maybe move to a different sub-class or role.
sub _new_empty_list
{
    my $self = shift;
    return $self->_new_list([]);
}

sub _new_node
{
    my $self = shift;
    my $args = shift;

    # t == type
    my $class =
        "XML::Grammar::Fiction::FromProto::Node::"
        . delete($args->{'t'})
        ;

    return $class->new(%$args);
}


sub _create_elem
{
    my $self = shift;
    my $open = shift;

    my $children = @_ ? shift(@_) : $self->_new_empty_list();

    return
        $self->_new_node(
            {
                t => (
                    $open->name() eq "desc" ? "Description"
                    : $open->name() eq "innerdesc" ? "InnerDesc"
                    : "Element"
                ),
                name => $open->name(),
                children => $children,
                attrs => $open->attrs(),
                open_line => $open->line(),
            }
        );
}

sub _new_list
{
    my $self = shift;
    my $contents = shift;

    return $self->_new_node(
        {
            t => "List",
            contents => $contents,
        }
    );
}

sub _generic_para_contents_assert
{
    my ($self, $predicate, $message, $contents) = @_;

    if (List::MoreUtils::any { $predicate->($_) } @{$contents || []})
    {
        Carp::confess ($message);
    }

    return;
}

sub _assert_not_contains_saying
{
    my ($self, $contents) = @_;

    return $self->_generic_para_contents_assert(
        sub { ref($_) ne "" && $_->isa("XML::Grammar::Fiction::FromProto::Node::Saying") },
        qq{Para contains a saying.},
        $contents
    );
}

sub _assert_not_contains_undef
{
    my ($self, $contents) = @_;

    return $self->_generic_para_contents_assert(
        sub { !defined($_) },
        qq{Para contains an undef member.},
        $contents
    );
}

sub _new_para
{
    my ($self, $contents) = @_;

    $self->_assert_not_contains_saying($contents);
    $self->_assert_not_contains_undef($contents);

    return $self->_new_node(
        {
            t => "Paragraph",
            children => $self->_new_list($contents),
        }
    );
}

sub _new_text
{
    my $self = shift;
    my $contents = shift;

    return $self->_new_node(
        {
            t => "Text",
            children => $self->_new_list($contents),
        }
    );
}

sub _new_comment
{
    my $self = shift;
    my $text = shift;

    return $self->_new_node(
        {
            t => "Comment",
            text => $text,
        }
    );
}

sub _parse_opening_tag_attrs
{
    my $self = shift;

    my $l = $self->curr_line_ref();

    my @attrs;

    my $id_regex = $self->_get_id_regex();

    while ($$l =~ m{\G\s*($id_regex)="([^"]+)"\s*}cg)
    {
        push @attrs, { 'key' => $1, 'value' => $2, };
    }

    return \@attrs;
}

sub _opening_tag_asserts
{
    my $self = shift;

    if ($self->eof)
    {
        Carp::confess (qq{Reached EOF in _parse_opening_tag.});
    }

    if (!defined($self->curr_pos()))
    {
        Carp::confess (qq{curr_pos is not defined in _parse_opening_tag.});
    }

    return;
}

sub _parse_opening_tag
{
    my $self = shift;

    $self->_opening_tag_asserts;

    my $l = $self->curr_line_ref();

    my $id_regex = $self->_get_id_regex();

    if ($$l !~ m{\G<($id_regex)}cg)
    {
        $self->throw_text_error(
            'XML::Grammar::Fiction::Err::Parse::CannotMatchOpeningTag',
            "Cannot match opening tag.",
        );
    }

    my $id = $1;

    my $attrs = $self->_parse_opening_tag_attrs();

    my $is_standalone = 0;
    if ($$l =~ m{\G\s*/\s*>}cg)
    {
        $is_standalone = 1;
    }
    elsif ($$l !~ m{\G>}g)
    {
        $self->throw_text_error(
            'XML::Grammar::Fiction::Err::Parse::NoRightAngleBracket',
            "Cannot match the \">\" of the opening tag",
        );
    }

    return XML::Grammar::Fiction::Struct::Tag->new(
        name => $id,
        is_standalone => $is_standalone,
        line => $self->line_num(),
        attrs => $attrs,
    );
}

sub _parse_closing_tag
{
    my $self = shift;

    my $l = $self->curr_line_ref();

    my $id_regex = $self->_get_id_regex();

    if ($$l !~ m{\G}g)
    {
        $self->throw_text_error(
            'XML::Grammar::Fiction::Err::Parse::WrongClosingTagSyntax',
            "Cannot match closing tag",
        );
    }

    return XML::Grammar::Fiction::Struct::Tag->new(
        name => $1,
        line => $self->line_num(),
    );
}

sub _check_for_open_tag
{
    my $self = shift;

    if ($self->_tags_stack_is_empty())
    {
        $self->throw_text_error(
            'XML::Grammar::Fiction::Err::Parse::CannotMatchOpeningTag',
            "Cannot match opening tag.",
        );
    }

    return;
}

sub _is_event_a_saying
{
    my ($self, $event) = @_;

    return $event->is_tag_of_name("saying");
}

sub _is_event_a_para
{
    my ($self, $event) = @_;

    return $event->is_tag_of_name("para");
}

sub _is_event_elem
{
    my ($self, $event) = @_;

    return $event->type() eq "elem";
}

sub _handle_event
{
    my ($self, $event) = @_;

    if ((! $self->_check_and_handle_tag_event($event))
        && $self->_is_event_elem($event)
    )
    {
        $self->_handle_elem_event($event);
    }

    return;
}

sub _handle_specific_tag_event
{
    my ($self, $event) = @_;

    my $tag_name = $event->tag();
    my $type = $event->is_open() ? "open" : "close";

    my $method = "_handle_${type}_${tag_name}";

    $self->$method($event);

    return 1;
}

sub _check_and_handle_tag_event
{
    my ($self, $event) = @_;

    if ($event->tag && exists($self->_tag_names_to_be_handled->{$event->tag}))
    {
        return $self->_handle_specific_tag_event($event);
    }
    else
    {
        return;
    }
}

sub _handle_para_event
{
    my ($self, $event) = @_;

    return
          $event->is_open()
        ? $self->_handle_open_para($event)
        : $self->_handle_close_para($event)
        ;
}

sub _handle_elem_event
{
    my ($self, $event) = @_;

    $self->_add_to_top_tag( $event->elem());

    return;
}

sub _handle_non_tag_text
{
    my $self = shift;

    $self->_check_for_open_tag();

    my $contents = $self->_parse_text();

    foreach my $event (@$contents)
    {
        $self->_handle_event($event);
    }

    return;
}


sub _look_for_and_handle_tag
{
    my $self = shift;

    my ($is_tag_cond, $is_close) = $self->_look_ahead_for_tag();

    # Check if it's a closing tag.
    if ($is_close)
    {
        return $self->_handle_close_tag();
    }
    elsif ($is_tag_cond)
    {
        $self->_handle_open_tag();
    }
    else
    {
        $self->_handle_non_tag_text();
    }
    return;
}

sub _merge_tag
{
    my $self = shift;
    my $open_tag = shift;

    my $new_elem =
        $self->_create_elem(
            $open_tag,
            $self->_new_list($open_tag->detach_children()),
        );

    if (! $self->_tags_stack_is_empty())
    {
        $self->_add_to_top_tag($new_elem);
        return;
    }
    else
    {
        return $new_elem;
    }
}

sub _handle_close_tag
{
    my $self = shift;

    my $close = $self->_parse_closing_tag();

    my $open = $self->_pop_tag();

    if ($open->name() ne $close->name())
    {
        XML::Grammar::Fiction::Err::Parse::TagsMismatch->throw(
            error => "Tags do not match",
            opening_tag => $open,
            closing_tag => $close,
        );
    }

    return $self->_merge_tag($open);
}

sub _look_ahead_for_comment
{
    my $self = shift;

    if ($self->curr_line_continues_with(qr{});

        $self->_add_to_top_tag(
            $self->_new_comment($text),
        );

        return 1;
    }
    else
    {
        return;
    }
}

sub _decode_entities_in_text
{
    my ($self, $orig_text) = @_;

    my $ret = '';

    # Incrementally parse $text for entities.
    pos($orig_text) = 0;

    while ($orig_text =~ m{\G(.*?)(\&|\z)}msg)
    {
        my ($before, $indicator) = ($1, $2);

        $ret .= $before;

        if ($indicator eq '&')
        {
            if ($orig_text =~ m{\G(\#?\w+;)}cg)
            {
                $ret .= HTML::Entities::decode_entities("&$1");
            }
            else
            {
                Carp::confess(
                    sprintf(
                        "Cannot match entity '%s' at line %d",
                        substr($orig_text, pos($orig_text)-1, 10),
                        $self->line_num(),
                    )
                );
            }
        }
    }

    return $ret;
}

sub _parse_non_tag_text_unit
{
    my $self = shift;

    my $orig_text = $self->consume_up_to(
        $self->_non_tag_text_unit_consume_regex
    );

    my $text = $self->_decode_entities_in_text($orig_text);

    my $l = $self->curr_line_ref();

    my $ret_elem = $self->_new_text([$text]);
    my $is_para_end = 0;

    # Demote the cursor to before the < of the tag.
    #
    if ($self->at_line_start)
    {
        $is_para_end = 1;
    }
    else
    {
        pos($$l)--;
        if (substr($$l, pos($$l), 1) eq "\n")
        {
            $is_para_end = 1;
        }
    }

    if ($text !~ /\S/)
    {
        return;
    }
    else
    {
        return
        {
            elem => $ret_elem,
            para_end => $is_para_end,
        };
    }
}

sub _parse_text_unit
{
    my $self = shift;

    if (defined(my $event = $self->_extract_event()))
    {
        return $event;
    }
    else
    {
        $self->_generate_text_unit_events();
        return $self->_extract_event();
    }
}

sub _flush_events
{
    my $self = shift;

    my @ret = @{$self->_events_queue()};

    $self->_clear_events;

    return \@ret;
}

sub _parse_text
{
    my $self = shift;

    my @ret;

    while (my $unit = $self->_parse_text_unit())
    {
        push @ret, $unit;

        if ($unit->is_open_or_close)
        {
            return [@ret, @{$self->_flush_events()}];
        }
    }

    return \@ret;
}

sub _look_for_tag_opener
{
    my $self = shift;

    my $l = $self->curr_line_ref();

    if ($$l =~ m{\G(<(?:/)?)}cg)
    {
        return $1;
    }
    else
    {
        return;
    }
}

sub _is_closing_tag {
    my $self = shift;
    my $tag_start = shift;

    return $tag_start =~ m{/};
}

sub _generate_tag_event
{
    my $self = shift;

    my $l = $self->curr_line_ref();
    my $orig_pos = pos($$l);

    if (defined(my $tag_start = $self->_look_for_tag_opener()))
    {
        # If it's a tag.

        # TODO : implement the comment handling.
        # We have a tag.

        pos($$l) = $orig_pos;

        $self->_enqueue_event(
            XML::Grammar::FictionBase::Event->new(
                {'type' => ($self->_is_closing_tag($tag_start) ? "close" : "open")}
            ),
        );

        return 1;
    }
    else
    {
        return;
    }
}

sub _handle_open_tag
{
    my $self = shift;

    my $open = $self->_parse_opening_tag();

    $open->children([]);

    # TODO : add the check for is_standalone in XML-Grammar-Fiction
    # too.
    if ($open->is_standalone())
    {
        if (defined($self->_merge_tag($open)))
        {
            Carp::confess ("Top element/tag cannot be standalone.");
        }
        else
        {
            return;
        }
    }
    else
    {
        $self->_push_tag($open);

        return;
    }
}

sub _generate_text_unit_events
{
    my $self = shift;

    # $self->skip_multiline_space();

    if (! $self->_generate_tag_event())
    {
        $self->_generate_non_tag_text_event();
    }

    return;
}

sub _flush_ret_tag
{
    my $self = shift;

    my $ret = $self->_ret_tag();

    $self->_clear_ret_tag();

    return $ret;
}

sub _main_loop
{
    my $self = shift;

    while (! defined($self->_ret_tag()))
    {
        $self->_main_loop_iter();
    }

    return;
}

sub _parse_all
{
    my $self = shift;

    $self->_main_loop();

    return $self->_flush_ret_tag();
}

sub _assert_not_eof
{
    my $self = shift;

    if ($self->eof() && $self->_no_events())
    {
        if (! $self->_tags_stack_is_empty() )
        {
            XML::Grammar::Fiction::Err::Parse::TagNotClosedAtEOF->throw(
                error => "Tag not closed at EOF.",
                opening_tag => $self->_top_tag(),
            );
        }
        else
        {
            Carp::confess (qq{Reached EOF.});
        }
    }

    return;
}

sub _main_loop_iter
{
    my $self = shift;

    $self->_assert_not_eof;

    if ($self->_look_ahead_for_comment)
    {
        return;
    }
    else
    {
        return $self->_main_loop_iter_body;
    }
}

sub _attempt_to_calc_new_ret_tag
{
    my $self = shift;

    $self->_ret_tag(scalar($self->_look_for_and_handle_tag()));

    return;
}

sub _main_loop_iter_body
{
    my $self = shift;

    if ($self->_main_loop_iter_body_prelude())
    {
        $self->_attempt_to_calc_new_ret_tag();
    }

    return;
}


our $VERSION = '0.14.11';


sub process_text
{
    my ($self, $text) = @_;

    $self->setup_text($text);

    return $self->_parse_all();
}


1;

__END__