1 package Text::Tradition::Collation;
4 use Encode qw( decode_utf8 );
8 use IPC::Run qw( run binary );
10 use Text::Tradition::Collation::Reading;
11 use Text::Tradition::Collation::RelationshipStore;
12 use Text::Tradition::Error;
13 use XML::Easy::Syntax qw( $xml10_namestartchar_rx $xml10_namechar_rx );
19 default => sub { Graph->new() },
27 isa => 'Text::Tradition::Collation::RelationshipStore',
29 relationships => 'relationships',
30 related_readings => 'related_readings',
31 get_relationship => 'get_relationship',
32 del_relationship => 'del_relationship',
33 equivalence => 'equivalence',
34 equivalence_graph => 'equivalence_graph',
36 writer => '_set_relations',
41 isa => 'Text::Tradition',
42 writer => '_set_tradition',
47 isa => 'HashRef[Text::Tradition::Collation::Reading]',
51 _add_reading => 'set',
52 del_reading => 'delete',
53 has_reading => 'exists',
56 default => sub { {} },
59 has 'wit_list_separator' => (
68 default => 'base text',
91 isa => 'Text::Tradition::Collation::Reading',
92 writer => '_set_start',
98 isa => 'Text::Tradition::Collation::Reading',
103 has 'cached_svg' => (
106 predicate => 'has_cached_svg',
107 clearer => 'wipe_svg',
110 has 'cached_table' => (
113 predicate => 'has_cached_table',
114 clearer => 'wipe_table',
117 has '_graphcalc_done' => (
125 Text::Tradition::Collation - a software model for a text collation
130 my $t = Text::Tradition->new(
131 'name' => 'this is a text',
133 'file' => '/path/to/tei_parallel_seg_file.xml' );
135 my $c = $t->collation;
136 my @readings = $c->readings;
137 my @paths = $c->paths;
138 my @relationships = $c->relationships;
140 my $svg_variant_graph = $t->collation->as_svg();
144 Text::Tradition is a library for representation and analysis of collated
145 texts, particularly medieval ones. The Collation is the central feature of
146 a Tradition, where the text, its sequence of readings, and its relationships
147 between readings are actually kept.
153 The constructor. Takes a hash or hashref of the following arguments:
157 =item * tradition - The Text::Tradition object to which the collation
160 =item * linear - Whether the collation should be linear; that is, whether
161 transposed readings should be treated as two linked readings rather than one,
162 and therefore whether the collation graph is acyclic. Defaults to true.
164 =item * baselabel - The default label for the path taken by a base text
165 (if any). Defaults to 'base text'.
167 =item * wit_list_separator - The string to join a list of witnesses for
168 purposes of making labels in display graphs. Defaults to ', '.
170 =item * ac_label - The extra label to tack onto a witness sigil when
171 representing another layer of path for the given witness - that is, when
172 a text has more than one possible reading due to scribal corrections or
173 the like. Defaults to ' (a.c.)'.
175 =item * wordsep - The string used to separate words in the original text.
186 =head2 wit_list_separator
194 Simple accessors for collation attributes.
198 The meta-reading at the start of every witness path.
202 The meta-reading at the end of every witness path.
206 Returns all Reading objects in the graph.
208 =head2 reading( $id )
210 Returns the Reading object corresponding to the given ID.
212 =head2 add_reading( $reading_args )
214 Adds a new reading object to the collation.
215 See L<Text::Tradition::Collation::Reading> for the available arguments.
217 =head2 del_reading( $object_or_id )
219 Removes the given reading from the collation, implicitly removing its
220 paths and relationships.
222 =head2 merge_readings( $main, $second, $concatenate, $with_str )
224 Merges the $second reading into the $main one. If $concatenate is true, then
225 the merged node will carry the text of both readings, concatenated with either
226 $with_str (if specified) or a sensible default (the empty string if the
227 appropriate 'join_*' flag is set on either reading, or else $self->wordsep.)
229 The first two arguments may be either readings or reading IDs.
231 =head2 has_reading( $id )
233 Predicate to see whether a given reading ID is in the graph.
235 =head2 reading_witnesses( $object_or_id )
237 Returns a list of sigils whose witnesses contain the reading.
241 Returns all reading paths within the document - that is, all edges in the
242 collation graph. Each path is an arrayref of [ $source, $target ] reading IDs.
244 =head2 add_path( $source, $target, $sigil )
246 Links the given readings in the collation in sequence, under the given witness
247 sigil. The readings may be specified by object or ID.
249 =head2 del_path( $source, $target, $sigil )
251 Links the given readings in the collation in sequence, under the given witness
252 sigil. The readings may be specified by object or ID.
254 =head2 has_path( $source, $target );
256 Returns true if the two readings are linked in sequence in any witness.
257 The readings may be specified by object or ID.
261 Returns all Relationship objects in the collation.
263 =head2 add_relationship( $reading, $other_reading, $options )
265 Adds a new relationship of the type given in $options between the two readings,
266 which may be specified by object or ID. Returns a value of ( $status, @vectors)
267 where $status is true on success, and @vectors is a list of relationship edges
268 that were ultimately added.
269 See L<Text::Tradition::Collation::Relationship> for the available options.
275 $self->_set_relations( Text::Tradition::Collation::RelationshipStore->new( 'collation' => $self ) );
276 $self->_set_start( $self->add_reading(
277 { 'collation' => $self, 'is_start' => 1, 'init' => 1 } ) );
278 $self->_set_end( $self->add_reading(
279 { 'collation' => $self, 'is_end' => 1, 'init' => 1 } ) );
282 ### Reading construct/destruct functions
285 my( $self, $reading ) = @_;
286 unless( ref( $reading ) eq 'Text::Tradition::Collation::Reading' ) {
287 my %args = %$reading;
288 if( $args{'init'} ) {
289 # If we are initializing an empty collation, don't assume that we
290 # have set a tradition.
291 delete $args{'init'};
292 } elsif( $self->tradition->has_language && !exists $args{'language'} ) {
293 $args{'language'} = $self->tradition->language;
295 $reading = Text::Tradition::Collation::Reading->new(
296 'collation' => $self,
299 # First check to see if a reading with this ID exists.
300 if( $self->reading( $reading->id ) ) {
301 throw( "Collation already has a reading with id " . $reading->id );
303 $self->_graphcalc_done(0);
304 $self->_add_reading( $reading->id => $reading );
305 # Once the reading has been added, put it in both graphs.
306 $self->sequence->add_vertex( $reading->id );
307 $self->relations->add_reading( $reading->id );
311 around del_reading => sub {
316 if( ref( $arg ) eq 'Text::Tradition::Collation::Reading' ) {
319 # Remove the reading from the graphs.
320 $self->_graphcalc_done(0);
321 $self->_clear_cache; # Explicitly clear caches to GC the reading
322 $self->sequence->delete_vertex( $arg );
323 $self->relations->delete_reading( $arg );
326 $self->$orig( $arg );
333 my $cxfile = 't/data/Collatex-16.xml';
334 my $t = Text::Tradition->new(
336 'input' => 'CollateX',
339 my $c = $t->collation;
341 my $rno = scalar $c->readings;
342 # Split n21 for testing purposes
343 my $new_r = $c->add_reading( { 'id' => 'n21p0', 'text' => 'un', 'join_next' => 1 } );
344 my $old_r = $c->reading( 'n21' );
345 $old_r->alter_text( 'to' );
346 $c->del_path( 'n20', 'n21', 'A' );
347 $c->add_path( 'n20', 'n21p0', 'A' );
348 $c->add_path( 'n21p0', 'n21', 'A' );
350 ok( $c->reading( 'n21p0' ), "New reading exists" );
351 is( scalar $c->readings, $rno, "Reading add offset by flatten_ranks" );
353 # Combine n3 and n4 ( with his )
354 $c->merge_readings( 'n3', 'n4', 1 );
355 ok( !$c->reading('n4'), "Reading n4 is gone" );
356 is( $c->reading('n3')->text, 'with his', "Reading n3 has both words" );
358 # Collapse n9 and n10 ( rood / root )
359 $c->merge_readings( 'n9', 'n10' );
360 ok( !$c->reading('n10'), "Reading n10 is gone" );
361 is( $c->reading('n9')->text, 'rood', "Reading n9 has an unchanged word" );
363 # Combine n21 and n21p0
364 my $remaining = $c->reading('n21');
365 $remaining ||= $c->reading('n22'); # one of these should still exist
366 $c->merge_readings( 'n21p0', $remaining, 1 );
367 ok( !$c->reading('n21'), "Reading $remaining is gone" );
368 is( $c->reading('n21p0')->text, 'unto', "Reading n21p0 merged correctly" );
378 my( $kept_obj, $del_obj, $combine, $combine_char ) = $self->_objectify_args( @_ );
379 my $mergemeta = $kept_obj->is_meta;
380 throw( "Cannot merge meta and non-meta reading" )
381 unless ( $mergemeta && $del_obj->is_meta )
382 || ( !$mergemeta && !$del_obj->is_meta );
384 throw( "Cannot merge with start or end node" )
385 if( $kept_obj eq $self->start || $kept_obj eq $self->end
386 || $del_obj eq $self->start || $del_obj eq $self->end );
388 # We only need the IDs for adding paths to the graph, not the reading
389 # objects themselves.
390 my $kept = $kept_obj->id;
391 my $deleted = $del_obj->id;
392 $self->_graphcalc_done(0);
394 # The kept reading should inherit the paths and the relationships
395 # of the deleted reading.
396 foreach my $path ( $self->sequence->edges_at( $deleted ) ) {
397 my @vector = ( $kept );
398 push( @vector, $path->[1] ) if $path->[0] eq $deleted;
399 unshift( @vector, $path->[0] ) if $path->[1] eq $deleted;
400 next if $vector[0] eq $vector[1]; # Don't add a self loop
401 my %wits = %{$self->sequence->get_edge_attributes( @$path )};
402 $self->sequence->add_edge( @vector );
403 my $fwits = $self->sequence->get_edge_attributes( @vector );
404 @wits{keys %$fwits} = values %$fwits;
405 $self->sequence->set_edge_attributes( @vector, \%wits );
407 $self->relations->merge_readings( $kept, $deleted, $combine );
409 # Do the deletion deed.
411 # Combine the text of the readings
412 my $joinstr = $combine_char;
413 unless( defined $joinstr ) {
414 $joinstr = '' if $kept_obj->join_next || $del_obj->join_prior;
415 $joinstr = $self->wordsep unless defined $joinstr;
417 $kept_obj->alter_text( join( $joinstr, $kept_obj->text, $del_obj->text ) );
418 # Change this reading to a joining one if necessary
419 $kept_obj->_set_join_next( $del_obj->join_next );
420 $kept_obj->normal_form(
421 join( $joinstr, $kept_obj->normal_form, $del_obj->normal_form ) );
422 # Combine the lexemes present in the readings
423 if( $kept_obj->has_lexemes && $del_obj->has_lexemes ) {
424 $kept_obj->add_lexeme( $del_obj->lexemes );
427 $self->del_reading( $deleted );
430 =head2 compress_readings
432 Where possible in the graph, compresses plain sequences of readings into a
433 single reading. The sequences must consist of readings with no
434 relationships to other readings, with only a single witness path between
435 them and no other witness paths from either that would skip the other. The
436 readings must also not be marked as nonsense or bad grammar.
438 WARNING: This operation cannot be undone.
442 sub compress_readings {
444 # Anywhere in the graph that there is a reading that joins only to a single
445 # successor, and neither of these have any relationships, just join the two
448 foreach my $rdg ( sort { $a->rank <=> $b->rank } $self->readings ) {
449 # While we are here, get rid of any extra wordforms from a disambiguated
451 if( $rdg->disambiguated ) {
452 foreach my $lex ( $rdg->lexemes ) {
453 $lex->clear_matching_forms();
454 $lex->add_matching_form( $lex->form );
457 # Now look for readings that can be joined to their successors.
458 next if $rdg->is_meta;
459 next if $gobbled{$rdg->id};
460 next if $rdg->grammar_invalid || $rdg->is_nonsense;
461 next if $rdg->related_readings();
463 while( $self->sequence->successors( $rdg ) == 1 ) {
464 my( $next ) = $self->reading( $self->sequence->successors( $rdg ) );
465 throw( "Infinite loop" ) if $seen{$next->id};
466 $seen{$next->id} = 1;
467 last if $self->sequence->predecessors( $next ) > 1;
468 last if $next->is_meta;
469 last if $next->grammar_invalid || $next->is_nonsense;
470 last if $next->related_readings();
471 say "Joining readings $rdg and $next";
472 $self->merge_readings( $rdg, $next, 1 );
475 # Make sure we haven't screwed anything up
476 foreach my $wit ( $self->tradition->witnesses ) {
477 my $pathtext = $self->path_text( $wit->sigil );
478 my $origtext = join( ' ', @{$wit->text} );
479 throw( "Text differs for witness " . $wit->sigil )
480 unless $pathtext eq $origtext;
481 if( $wit->is_layered ) {
482 $pathtext = $self->path_text( $wit->sigil.$self->ac_label );
483 $origtext = join( ' ', @{$wit->layertext} );
484 throw( "Ante-corr text differs for witness " . $wit->sigil )
485 unless $pathtext eq $origtext;
489 $self->relations->rebuild_equivalence();
490 $self->calculate_ranks();
493 # Helper function for manipulating the graph.
494 sub _stringify_args {
495 my( $self, $first, $second, @args ) = @_;
497 if ref( $first ) eq 'Text::Tradition::Collation::Reading';
498 $second = $second->id
499 if ref( $second ) eq 'Text::Tradition::Collation::Reading';
500 return( $first, $second, @args );
503 # Helper function for manipulating the graph.
504 sub _objectify_args {
505 my( $self, $first, $second, $arg ) = @_;
506 $first = $self->reading( $first )
507 unless ref( $first ) eq 'Text::Tradition::Collation::Reading';
508 $second = $self->reading( $second )
509 unless ref( $second ) eq 'Text::Tradition::Collation::Reading';
510 return( $first, $second, $arg );
517 # We only need the IDs for adding paths to the graph, not the reading
518 # objects themselves.
519 my( $source, $target, $wit ) = $self->_stringify_args( @_ );
521 $self->_graphcalc_done(0);
522 # Connect the readings
523 unless( $self->sequence->has_edge( $source, $target ) ) {
524 $self->sequence->add_edge( $source, $target );
525 $self->relations->add_equivalence_edge( $source, $target );
527 # Note the witness in question
528 $self->sequence->set_edge_attribute( $source, $target, $wit, 1 );
534 if( ref( $_[0] ) eq 'ARRAY' ) {
541 # We only need the IDs for adding paths to the graph, not the reading
542 # objects themselves.
543 my( $source, $target, $wit ) = $self->_stringify_args( @args );
545 $self->_graphcalc_done(0);
546 if( $self->sequence->has_edge_attribute( $source, $target, $wit ) ) {
547 $self->sequence->delete_edge_attribute( $source, $target, $wit );
549 unless( keys %{$self->sequence->get_edge_attributes( $source, $target )} ) {
550 $self->sequence->delete_edge( $source, $target );
551 $self->relations->delete_equivalence_edge( $source, $target );
556 # Extra graph-alike utility
559 my( $source, $target, $wit ) = $self->_stringify_args( @_ );
560 return undef unless $self->sequence->has_edge( $source, $target );
561 return $self->sequence->has_edge_attribute( $source, $target, $wit );
564 =head2 clear_witness( @sigil_list )
566 Clear the given witnesses out of the collation entirely, removing references
567 to them in paths, and removing readings that belong only to them. Should only
568 be called via $tradition->del_witness.
573 my( $self, @sigils ) = @_;
575 $self->_graphcalc_done(0);
576 # Clear the witness(es) out of the paths
577 foreach my $e ( $self->paths ) {
578 foreach my $sig ( @sigils ) {
579 $self->del_path( $e, $sig );
583 # Clear out the newly unused readings
584 foreach my $r ( $self->readings ) {
585 unless( $self->reading_witnesses( $r ) ) {
586 $self->del_reading( $r );
591 sub add_relationship {
593 my( $source, $target, $opts ) = $self->_stringify_args( @_ );
594 my( @vectors ) = $self->relations->add_relationship( $source, $target, $opts );
595 $self->_graphcalc_done(0);
599 around qw/ get_relationship del_relationship / => sub {
603 if( @args == 1 && ref( $args[0] ) eq 'ARRAY' ) {
606 my( $source, $target ) = $self->_stringify_args( @args );
607 $self->$orig( $source, $target );
610 =head2 reading_witnesses( $reading )
612 Return a list of sigils corresponding to the witnesses in which the reading appears.
616 sub reading_witnesses {
617 my( $self, $reading ) = @_;
618 # We need only check either the incoming or the outgoing edges; I have
619 # arbitrarily chosen "incoming". Thus, special-case the start node.
620 if( $reading eq $self->start ) {
621 return map { $_->sigil } $self->tradition->witnesses;
624 foreach my $e ( $self->sequence->edges_to( $reading ) ) {
625 my $wits = $self->sequence->get_edge_attributes( @$e );
626 @all_witnesses{ keys %$wits } = 1;
628 my $acstr = $self->ac_label;
629 foreach my $acwit ( grep { $_ =~ s/^(.*)\Q$acstr\E$/$1/ } keys %all_witnesses ) {
630 delete $all_witnesses{$acwit.$acstr} if exists $all_witnesses{$acwit};
632 return keys %all_witnesses;
635 =head1 OUTPUT METHODS
637 =head2 as_svg( \%options )
639 Returns an SVG string that represents the graph, via as_dot and graphviz.
640 See as_dot for a list of options. Must have GraphViz (dot) installed to run.
645 my( $self, $opts ) = @_;
646 throw( "Need GraphViz installed to output SVG" )
647 unless File::Which::which( 'dot' );
648 my $want_subgraph = exists $opts->{'from'} || exists $opts->{'to'};
649 $self->calculate_ranks()
650 unless( $self->_graphcalc_done || $opts->{'nocalc'} || !$self->linear );
651 if( !$self->has_cached_svg || $opts->{'recalc'} || $want_subgraph ) {
652 my @cmd = qw/dot -Tsvg/;
654 my $dotfile = File::Temp->new();
656 # $dotfile->unlink_on_destroy(0);
657 binmode $dotfile, ':utf8';
658 print $dotfile $self->as_dot( $opts );
659 push( @cmd, $dotfile->filename );
660 run( \@cmd, ">", binary(), \$svg );
661 $svg = decode_utf8( $svg );
662 $self->cached_svg( $svg ) unless $want_subgraph;
665 return $self->cached_svg;
670 =head2 as_dot( \%options )
672 Returns a string that is the collation graph expressed in dot
673 (i.e. GraphViz) format. Options include:
688 my( $self, $opts ) = @_;
689 my $startrank = $opts->{'from'} if $opts;
690 my $endrank = $opts->{'to'} if $opts;
691 my $color_common = $opts->{'color_common'} if $opts;
692 my $STRAIGHTENHACK = !$startrank && !$endrank && $self->end->rank
693 && $self->end->rank > 100;
694 $STRAIGHTENHACK = 1 if $opts->{'straight'}; # even for subgraphs or small graphs
696 # Check the arguments
698 return if $endrank && $startrank > $endrank;
699 return if $startrank > $self->end->rank;
701 if( defined $endrank ) {
702 return if $endrank < 0;
703 $endrank = undef if $endrank == $self->end->rank;
706 my $graph_name = $self->tradition->name;
707 $graph_name =~ s/[^\w\s]//g;
708 $graph_name = join( '_', split( /\s+/, $graph_name ) );
716 'fillcolor' => 'white',
721 'arrowhead' => 'open',
722 'color' => '#000000',
723 'fontcolor' => '#000000',
726 my $dot = sprintf( "digraph %s {\n", $graph_name );
727 $dot .= "\tgraph " . _dot_attr_string( \%graph_attrs ) . ";\n";
728 $dot .= "\tnode " . _dot_attr_string( \%node_attrs ) . ";\n";
730 # Output substitute start/end readings if necessary
732 $dot .= "\t\"__SUBSTART__\" [ label=\"...\",id=\"__START__\" ];\n";
735 $dot .= "\t\"__SUBEND__\" [ label=\"...\",id=\"__END__\" ];\n";
737 if( $STRAIGHTENHACK ) {
739 my $startlabel = $startrank ? '__SUBSTART__' : '__START__';
740 $dot .= "\tsubgraph { rank=same \"$startlabel\" \"#SILENT#\" }\n";
741 $dot .= "\t\"#SILENT#\" [ shape=diamond,color=white,penwidth=0,label=\"\" ];"
743 my %used; # Keep track of the readings that actually appear in the graph
744 # Sort the readings by rank if we have ranks; this speeds layout.
745 my @all_readings = $self->end->has_rank
746 ? sort { $a->rank <=> $b->rank } $self->readings
748 # TODO Refrain from outputting lacuna nodes - just grey out the edges.
749 foreach my $reading ( @all_readings ) {
750 # Only output readings within our rank range.
751 next if $startrank && $reading->rank < $startrank;
752 next if $endrank && $reading->rank > $endrank;
753 $used{$reading->id} = 1;
754 # Need not output nodes without separate labels
755 next if $reading->id eq $reading->text;
757 my $label = $reading->text;
758 $label .= '-' if $reading->join_next;
759 $label = "-$label" if $reading->join_prior;
760 $label =~ s/\"/\\\"/g;
761 $rattrs->{'label'} = $label;
762 $rattrs->{'id'} = $reading->id;
763 $rattrs->{'fillcolor'} = '#b3f36d' if $reading->is_common && $color_common;
764 $dot .= sprintf( "\t\"%s\" %s;\n", $reading->id, _dot_attr_string( $rattrs ) );
767 # Add the real edges. Need to weight one edge per rank jump, in a
769 # my $weighted = $self->_add_edge_weights;
770 my @edges = $self->paths;
771 my( %substart, %subend );
772 foreach my $edge ( @edges ) {
773 # Do we need to output this edge?
774 if( $used{$edge->[0]} && $used{$edge->[1]} ) {
775 my $label = $self->_path_display_label( $self->path_witnesses( $edge ) );
776 my $variables = { %edge_attrs, 'label' => $label };
778 # Account for the rank gap if necessary
779 my $rank0 = $self->reading( $edge->[0] )->rank
780 if $self->reading( $edge->[0] )->has_rank;
781 my $rank1 = $self->reading( $edge->[1] )->rank
782 if $self->reading( $edge->[1] )->has_rank;
783 if( defined $rank0 && defined $rank1 && $rank1 - $rank0 > 1 ) {
784 $variables->{'minlen'} = $rank1 - $rank0;
787 # Add the calculated edge weights
788 # if( exists $weighted->{$edge->[0]}
789 # && $weighted->{$edge->[0]} eq $edge->[1] ) {
790 # # $variables->{'color'} = 'red';
791 # $variables->{'weight'} = 3.0;
794 # EXPERIMENTAL: make edge width reflect no. of witnesses
795 my $extrawidth = scalar( $self->path_witnesses( $edge ) ) * 0.2;
796 $variables->{'penwidth'} = $extrawidth + 0.8; # gives 1 for a single wit
798 my $varopts = _dot_attr_string( $variables );
799 $dot .= sprintf( "\t\"%s\" -> \"%s\" %s;\n",
800 $edge->[0], $edge->[1], $varopts );
801 } elsif( $used{$edge->[0]} ) {
802 $subend{$edge->[0]} = $edge->[1];
803 } elsif( $used{$edge->[1]} ) {
804 $substart{$edge->[1]} = $edge->[0];
807 # Add substitute start and end edges if necessary
808 foreach my $node ( keys %substart ) {
809 my $witstr = $self->_path_display_label ( $self->path_witnesses( $substart{$node}, $node ) );
810 my $variables = { %edge_attrs, 'label' => $witstr };
811 my $nrdg = $self->reading( $node );
812 if( $nrdg->has_rank && $nrdg->rank > $startrank ) {
813 # Substart is actually one lower than $startrank
814 $variables->{'minlen'} = $nrdg->rank - ( $startrank - 1 );
816 my $varopts = _dot_attr_string( $variables );
817 $dot .= "\t\"__SUBSTART__\" -> \"$node\" $varopts;\n";
819 foreach my $node ( keys %subend ) {
820 my $witstr = $self->_path_display_label ( $self->path_witnesses( $node, $subend{$node} ) );
821 my $variables = { %edge_attrs, 'label' => $witstr };
822 my $varopts = _dot_attr_string( $variables );
823 $dot .= "\t\"$node\" -> \"__SUBEND__\" $varopts;\n";
826 if( $STRAIGHTENHACK ) {
827 my $endlabel = $endrank ? '__SUBEND__' : '__END__';
828 $dot .= "\t\"$endlabel\" -> \"#SILENT#\" [ color=white,penwidth=0 ];\n";
835 sub _dot_attr_string {
838 foreach my $k ( sort keys %$hash ) {
840 push( @attrs, $k.'="'.$v.'"' );
842 return( '[ ' . join( ', ', @attrs ) . ' ]' );
845 sub _add_edge_weights {
847 # Walk the graph from START to END, choosing the successor node with
848 # the largest number of witness paths each time.
850 my $curr = $self->start->id;
851 my $ranked = $self->end->has_rank;
852 while( $curr ne $self->end->id ) {
853 my $rank = $ranked ? $self->reading( $curr )->rank : 0;
854 my @succ = sort { $self->path_witnesses( $curr, $a )
855 <=> $self->path_witnesses( $curr, $b ) }
856 $self->sequence->successors( $curr );
857 my $next = pop @succ;
858 my $nextrank = $ranked ? $self->reading( $next )->rank : 0;
859 # Try to avoid lacunae in the weighted path.
861 ( $self->reading( $next )->is_lacuna ||
862 $nextrank - $rank > 1 ) ){
865 $weighted->{$curr} = $next;
871 =head2 path_witnesses( $edge )
873 Returns the list of sigils whose witnesses are associated with the given edge.
874 The edge can be passed as either an array or an arrayref of ( $source, $target ).
879 my( $self, @edge ) = @_;
880 # If edge is an arrayref, cope.
881 if( @edge == 1 && ref( $edge[0] ) eq 'ARRAY' ) {
885 my @wits = keys %{$self->sequence->get_edge_attributes( @edge )};
889 # Helper function. Make a display label for the given witnesses, showing a.c.
890 # witnesses only where the main witness is not also in the list.
891 sub _path_display_label {
894 map { $wits{$_} = 1 } @_;
896 # If an a.c. wit is listed, remove it if the main wit is also listed.
897 # Otherwise keep it for explicit listing.
898 my $aclabel = $self->ac_label;
900 foreach my $w ( sort keys %wits ) {
901 if( $w =~ /^(.*)\Q$aclabel\E$/ ) {
902 if( exists $wits{$1} ) {
905 push( @disp_ac, $w );
910 # See if we are in a majority situation.
911 my $maj = scalar( $self->tradition->witnesses ) * 0.6;
912 $maj = $maj > 5 ? $maj : 5;
913 if( scalar keys %wits > $maj ) {
914 unshift( @disp_ac, 'majority' );
915 return join( ', ', @disp_ac );
917 return join( ', ', sort keys %wits );
921 =head2 readings_at_rank( $rank )
923 Returns a list of readings at a given rank, taken from the alignment table.
927 sub readings_at_rank {
928 my( $self, $rank ) = @_;
929 my $table = $self->alignment_table;
930 # Table rank is real rank - 1.
931 my @elements = map { $_->{'tokens'}->[$rank-1] } @{$table->{'alignment'}};
933 foreach my $e ( @elements ) {
934 next unless ref( $e ) eq 'HASH';
935 next unless exists $e->{'t'};
936 $readings{$e->{'t'}->id} = $e->{'t'};
938 return values %readings;
943 Returns a GraphML representation of the collation. The GraphML will contain
944 two graphs. The first expresses the attributes of the readings and the witness
945 paths that link them; the second expresses the relationships that link the
946 readings. This is the native transfer format for a tradition.
955 my $datafile = 't/data/florilegium_tei_ps.xml';
956 my $tradition = Text::Tradition->new( 'input' => 'TEI',
961 ok( $tradition, "Got a tradition object" );
962 is( scalar $tradition->witnesses, 13, "Found all witnesses" );
963 ok( $tradition->collation, "Tradition has a collation" );
965 my $c = $tradition->collation;
966 is( scalar $c->readings, $READINGS, "Collation has all readings" );
967 is( scalar $c->paths, $PATHS, "Collation has all paths" );
968 is( scalar $c->relationships, 0, "Collation has all relationships" );
970 # Add a few relationships
971 $c->add_relationship( 'w123', 'w125', { 'type' => 'collated' } );
972 $c->add_relationship( 'w193', 'w196', { 'type' => 'collated' } );
973 $c->add_relationship( 'w257', 'w262', { 'type' => 'transposition' } );
975 # Now write it to GraphML and parse it again.
977 my $graphml = $c->as_graphml;
978 my $st = Text::Tradition->new( 'input' => 'Self', 'string' => $graphml );
979 is( scalar $st->collation->readings, $READINGS, "Reparsed collation has all readings" );
980 is( scalar $st->collation->paths, $PATHS, "Reparsed collation has all paths" );
981 is( scalar $st->collation->relationships, 3, "Reparsed collation has new relationships" );
983 # Now add a stemma, write to GraphML, and look at the output.
984 my $stemma = $tradition->add_stemma( 'dotfile' => 't/data/florilegium.dot' );
985 is( ref( $stemma ), 'Text::Tradition::Stemma', "Parsed dotfile into stemma" );
986 is( $tradition->stemmata, 1, "Tradition now has the stemma" );
987 $graphml = $c->as_graphml;
988 like( $graphml, qr/digraph/, "Digraph declaration exists in GraphML" );
990 # Now add a user, write to GraphML, and look at the output.
991 unlike( $graphml, qr/testuser/, "Test user name does not exist in GraphML yet" );
992 my $testuser = Text::Tradition::User->new(
993 id => 'testuser', password => 'testpass' );
994 is( ref( $testuser ), 'Text::Tradition::User', "Created test user object" );
995 $testuser->add_tradition( $tradition );
996 is( $tradition->user->id, $testuser->id, "Tradition assigned to test user" );
997 $graphml = $c->as_graphml;
998 like( $graphml, qr/testuser/, "Test user name now exists in GraphML" );
1005 my( $self, $options ) = @_;
1006 $self->calculate_ranks unless $self->_graphcalc_done;
1008 my $start = $options->{'from'}
1009 ? $self->reading( $options->{'from'} ) : $self->start;
1010 my $end = $options->{'to'}
1011 ? $self->reading( $options->{'to'} ) : $self->end;
1012 if( $start->has_rank && $end->has_rank && $end->rank < $start->rank ) {
1013 throw( 'Start node must be before end node' );
1015 # The readings need to be ranked for this to work.
1016 $start = $self->start unless $start->has_rank;
1017 $end = $self->end unless $end->has_rank;
1019 unless( $start eq $self->start ) {
1020 $rankoffset = $start->rank - 1;
1025 my $graphml_ns = 'http://graphml.graphdrawing.org/xmlns';
1026 my $xsi_ns = 'http://www.w3.org/2001/XMLSchema-instance';
1027 my $graphml_schema = 'http://graphml.graphdrawing.org/xmlns ' .
1028 'http://graphml.graphdrawing.org/xmlns/1.0/graphml.xsd';
1030 # Create the document and root node
1031 require XML::LibXML;
1032 my $graphml = XML::LibXML->createDocument( "1.0", "UTF-8" );
1033 my $root = $graphml->createElementNS( $graphml_ns, 'graphml' );
1034 $graphml->setDocumentElement( $root );
1035 $root->setNamespace( $xsi_ns, 'xsi', 0 );
1036 $root->setAttributeNS( $xsi_ns, 'schemaLocation', $graphml_schema );
1038 # List of attribute types to save on our objects and their corresponding
1043 'Bool' => 'boolean',
1044 'ReadingID' => 'string',
1045 'RelationshipType' => 'string',
1046 'RelationshipScope' => 'string',
1049 # List of attribute names *not* to save on our objects.
1050 # We will also not save any attribute beginning with _.
1052 map { $skipsave{$_} = 1 } qw/ cached_svg /;
1054 # Add the data keys for the graph. Include an extra key 'version' for the
1055 # GraphML output version.
1056 my %graph_data_keys;
1058 my %graph_attributes = ( 'version' => 'string' );
1059 # Graph attributes include those of Tradition and those of Collation.
1061 my $tmeta = $self->tradition->meta;
1062 my $cmeta = $self->meta;
1063 map { $gattr_from{$_->name} = 'Tradition' } $tmeta->get_all_attributes;
1064 map { $gattr_from{$_->name} = 'Collation' } $cmeta->get_all_attributes;
1065 foreach my $attr ( ( $tmeta->get_all_attributes, $cmeta->get_all_attributes ) ) {
1066 next if $attr->name =~ /^_/;
1067 next if $skipsave{$attr->name};
1068 next unless $save_types{$attr->type_constraint->name};
1069 $graph_attributes{$attr->name} = $save_types{$attr->type_constraint->name};
1071 # Extra custom keys for complex objects that should be saved in some form.
1072 # The subroutine should return a string, or undef/empty.
1073 $graph_attributes{'stemmata'} = sub {
1075 map { push( @stemstrs, $_->editable( {linesep => ''} ) ) }
1076 $self->tradition->stemmata;
1077 join( "\n", @stemstrs );
1079 $graph_attributes{'user'} = sub {
1080 $self->tradition->user ? $self->tradition->user->id : undef
1083 foreach my $datum ( sort keys %graph_attributes ) {
1084 $graph_data_keys{$datum} = 'dg'.$gdi++;
1085 my $key = $root->addNewChild( $graphml_ns, 'key' );
1086 my $dtype = ref( $graph_attributes{$datum} ) ? 'string'
1087 : $graph_attributes{$datum};
1088 $key->setAttribute( 'attr.name', $datum );
1089 $key->setAttribute( 'attr.type', $dtype );
1090 $key->setAttribute( 'for', 'graph' );
1091 $key->setAttribute( 'id', $graph_data_keys{$datum} );
1094 # Add the data keys for reading nodes
1095 my %reading_attributes;
1096 my $rmeta = Text::Tradition::Collation::Reading->meta;
1097 foreach my $attr( $rmeta->get_all_attributes ) {
1098 next if $attr->name =~ /^_/;
1099 next if $skipsave{$attr->name};
1100 next unless $save_types{$attr->type_constraint->name};
1101 $reading_attributes{$attr->name} = $save_types{$attr->type_constraint->name};
1103 # Extra custom key for the reading morphology
1104 $reading_attributes{'lexemes'} = 'string';
1108 foreach my $datum ( sort keys %reading_attributes ) {
1109 $node_data_keys{$datum} = 'dn'.$ndi++;
1110 my $key = $root->addNewChild( $graphml_ns, 'key' );
1111 $key->setAttribute( 'attr.name', $datum );
1112 $key->setAttribute( 'attr.type', $reading_attributes{$datum} );
1113 $key->setAttribute( 'for', 'node' );
1114 $key->setAttribute( 'id', $node_data_keys{$datum} );
1117 # Add the data keys for edges, that is, paths and relationships. Path
1118 # data does not come from a Moose class so is here manually.
1121 my %edge_attributes = (
1122 witness => 'string', # ID/label for a path
1123 extra => 'boolean', # Path key
1125 my @path_attributes = keys %edge_attributes; # track our manual additions
1126 my $pmeta = Text::Tradition::Collation::Relationship->meta;
1127 foreach my $attr( $pmeta->get_all_attributes ) {
1128 next if $attr->name =~ /^_/;
1129 next if $skipsave{$attr->name};
1130 next unless $save_types{$attr->type_constraint->name};
1131 $edge_attributes{$attr->name} = $save_types{$attr->type_constraint->name};
1133 foreach my $datum ( sort keys %edge_attributes ) {
1134 $edge_data_keys{$datum} = 'de'.$edi++;
1135 my $key = $root->addNewChild( $graphml_ns, 'key' );
1136 $key->setAttribute( 'attr.name', $datum );
1137 $key->setAttribute( 'attr.type', $edge_attributes{$datum} );
1138 $key->setAttribute( 'for', 'edge' );
1139 $key->setAttribute( 'id', $edge_data_keys{$datum} );
1142 # Add the collation graph itself. First, sanitize the name to a valid XML ID.
1143 my $xmlidname = $self->tradition->name;
1144 $xmlidname =~ s/(?!$xml10_namechar_rx)./_/g;
1145 if( $xmlidname !~ /^$xml10_namestartchar_rx/ ) {
1146 $xmlidname = '_'.$xmlidname;
1148 my $sgraph = $root->addNewChild( $graphml_ns, 'graph' );
1149 $sgraph->setAttribute( 'edgedefault', 'directed' );
1150 $sgraph->setAttribute( 'id', $xmlidname );
1151 $sgraph->setAttribute( 'parse.edgeids', 'canonical' );
1152 $sgraph->setAttribute( 'parse.edges', 0 ); # fill in later
1153 $sgraph->setAttribute( 'parse.nodeids', 'canonical' );
1154 $sgraph->setAttribute( 'parse.nodes', 0 ); # fill in later
1155 $sgraph->setAttribute( 'parse.order', 'nodesfirst' );
1157 # Tradition/collation attribute data
1158 foreach my $datum ( keys %graph_attributes ) {
1160 if( $datum eq 'version' ) {
1162 } elsif( ref( $graph_attributes{$datum} ) ) {
1163 my $sub = $graph_attributes{$datum};
1165 } elsif( $gattr_from{$datum} eq 'Tradition' ) {
1166 $value = $self->tradition->$datum;
1168 $value = $self->$datum;
1170 _add_graphml_data( $sgraph, $graph_data_keys{$datum}, $value );
1175 # Add our readings to the graph
1176 foreach my $n ( sort { $a->id cmp $b->id } $self->readings ) {
1177 next if $n->has_rank && $n ne $self->start && $n ne $self->end &&
1178 ( $n->rank < $start->rank || $n->rank > $end->rank );
1179 $use_readings{$n->id} = 1;
1180 # Add to the main graph
1181 my $node_el = $sgraph->addNewChild( $graphml_ns, 'node' );
1182 my $node_xmlid = 'n' . $node_ctr++;
1183 $node_hash{ $n->id } = $node_xmlid;
1184 $node_el->setAttribute( 'id', $node_xmlid );
1185 foreach my $d ( keys %reading_attributes ) {
1187 # Custom serialization
1188 if( $d eq 'lexemes' ) {
1189 # If nval is a true value, we have lexemes so we need to
1190 # serialize them. Otherwise set nval to undef so that the
1191 # key is excluded from this reading.
1192 $nval = $nval ? $n->_serialize_lexemes : undef;
1193 } elsif( $d eq 'normal_form' && $n->normal_form eq $n->text ) {
1196 if( $rankoffset && $d eq 'rank' && $n ne $self->start ) {
1197 # Adjust the ranks within the subgraph.
1198 $nval = $n eq $self->end ? $end->rank - $rankoffset + 1
1199 : $nval - $rankoffset;
1201 _add_graphml_data( $node_el, $node_data_keys{$d}, $nval )
1206 # Add the path edges to the sequence graph
1208 foreach my $e ( sort { $a->[0] cmp $b->[0] } $self->sequence->edges() ) {
1209 # We add an edge in the graphml for every witness in $e.
1210 next unless( $use_readings{$e->[0]} || $use_readings{$e->[1]} );
1211 my @edge_wits = sort $self->path_witnesses( $e );
1212 $e->[0] = $self->start->id unless $use_readings{$e->[0]};
1213 $e->[1] = $self->end->id unless $use_readings{$e->[1]};
1214 # Skip any path from start to end; that witness is not in the subgraph.
1215 next if ( $e->[0] eq $self->start->id && $e->[1] eq $self->end->id );
1216 foreach my $wit ( @edge_wits ) {
1217 my( $id, $from, $to ) = ( 'e'.$edge_ctr++,
1218 $node_hash{ $e->[0] },
1219 $node_hash{ $e->[1] } );
1220 my $edge_el = $sgraph->addNewChild( $graphml_ns, 'edge' );
1221 $edge_el->setAttribute( 'source', $from );
1222 $edge_el->setAttribute( 'target', $to );
1223 $edge_el->setAttribute( 'id', $id );
1225 # It's a witness path, so add the witness
1227 my $key = $edge_data_keys{'witness'};
1228 # Is this an ante-corr witness?
1229 my $aclabel = $self->ac_label;
1230 if( $wit =~ /^(.*)\Q$aclabel\E$/ ) {
1231 # Keep the base witness
1233 # ...and record that this is an 'extra' reading path
1234 _add_graphml_data( $edge_el, $edge_data_keys{'extra'}, $aclabel );
1236 _add_graphml_data( $edge_el, $edge_data_keys{'witness'}, $base );
1240 # Report the actual number of nodes and edges that went in
1241 $sgraph->setAttribute( 'parse.edges', $edge_ctr );
1242 $sgraph->setAttribute( 'parse.nodes', $node_ctr );
1244 # Add the relationship graph to the XML
1245 map { delete $edge_data_keys{$_} } @path_attributes;
1246 $self->relations->_as_graphml( $graphml_ns, $root, \%node_hash,
1247 $node_data_keys{'id'}, \%edge_data_keys );
1249 # Save and return the thing
1250 my $result = decode_utf8( $graphml->toString(1) );
1254 sub _add_graphml_data {
1255 my( $el, $key, $value ) = @_;
1256 return unless defined $value;
1257 my $data_el = $el->addNewChild( $el->namespaceURI, 'data' );
1258 $data_el->setAttribute( 'key', $key );
1259 $data_el->appendText( $value );
1264 Returns a CSV alignment table representation of the collation graph, one
1265 row per witness (or witness uncorrected.)
1271 my $table = $self->alignment_table;
1272 my $csv = Text::CSV->new( { binary => 1, quote_null => 0 } );
1274 # Make the header row
1275 $csv->combine( map { $_->{'witness'} } @{$table->{'alignment'}} );
1276 push( @result, decode_utf8( $csv->string ) );
1277 # Make the rest of the rows
1278 foreach my $idx ( 0 .. $table->{'length'} - 1 ) {
1279 my @rowobjs = map { $_->{'tokens'}->[$idx] } @{$table->{'alignment'}};
1280 my @row = map { $_ ? $_->{'t'}->text : $_ } @rowobjs;
1281 $csv->combine( @row );
1282 push( @result, decode_utf8( $csv->string ) );
1284 return join( "\n", @result );
1287 =head2 alignment_table( $use_refs, $include_witnesses )
1289 Return a reference to an alignment table, in a slightly enhanced CollateX
1290 format which looks like this:
1292 $table = { alignment => [ { witness => "SIGIL",
1293 tokens => [ { t => "TEXT" }, ... ] },
1294 { witness => "SIG2",
1295 tokens => [ { t => "TEXT" }, ... ] },
1297 length => TEXTLEN };
1299 If $use_refs is set to 1, the reading object is returned in the table
1300 instead of READINGTEXT; if not, the text of the reading is returned.
1302 If $include_witnesses is set to a hashref, only the witnesses whose sigil
1303 keys have a true hash value will be included.
1307 sub alignment_table {
1309 $self->calculate_ranks() unless $self->_graphcalc_done;
1310 return $self->cached_table if $self->has_cached_table;
1312 # Make sure we can do this
1313 throw( "Need a linear graph in order to make an alignment table" )
1314 unless $self->linear;
1315 $self->calculate_ranks unless $self->end->has_rank;
1317 my $table = { 'alignment' => [], 'length' => $self->end->rank - 1 };
1318 my @all_pos = ( 1 .. $self->end->rank - 1 );
1319 foreach my $wit ( sort { $a->sigil cmp $b->sigil } $self->tradition->witnesses ) {
1320 # say STDERR "Making witness row(s) for " . $wit->sigil;
1321 my @wit_path = $self->reading_sequence( $self->start, $self->end, $wit->sigil );
1322 my @row = _make_witness_row( \@wit_path, \@all_pos );
1323 push( @{$table->{'alignment'}},
1324 { 'witness' => $wit->sigil, 'tokens' => \@row } );
1325 if( $wit->is_layered ) {
1326 my @wit_ac_path = $self->reading_sequence( $self->start, $self->end,
1327 $wit->sigil.$self->ac_label );
1328 my @ac_row = _make_witness_row( \@wit_ac_path, \@all_pos );
1329 push( @{$table->{'alignment'}},
1330 { 'witness' => $wit->sigil.$self->ac_label, 'tokens' => \@ac_row } );
1333 $self->cached_table( $table );
1337 sub _make_witness_row {
1338 my( $path, $positions ) = @_;
1340 map { $char_hash{$_} = undef } @$positions;
1342 foreach my $rdg ( @$path ) {
1343 my $rtext = $rdg->text;
1344 $rtext = '#LACUNA#' if $rdg->is_lacuna;
1345 say STDERR "rank " . $rdg->rank if $debug;
1346 # say STDERR "No rank for " . $rdg->id unless defined $rdg->rank;
1347 $char_hash{$rdg->rank} = { 't' => $rdg };
1349 my @row = map { $char_hash{$_} } @$positions;
1350 # Fill in lacuna markers for undef spots in the row
1351 my $last_el = shift @row;
1352 my @filled_row = ( $last_el );
1353 foreach my $el ( @row ) {
1354 # If we are using node reference, make the lacuna node appear many times
1355 # in the table. If not, use the lacuna tag.
1356 if( $last_el && $last_el->{'t'}->is_lacuna && !defined $el ) {
1359 push( @filled_row, $el );
1365 =head1 NAVIGATION METHODS
1367 =head2 reading_sequence( $first, $last, $sigil, $backup )
1369 Returns the ordered list of readings, starting with $first and ending
1370 with $last, for the witness given in $sigil. If a $backup sigil is
1371 specified (e.g. when walking a layered witness), it will be used wherever
1372 no $sigil path exists. If there is a base text reading, that will be
1373 used wherever no path exists for $sigil or $backup.
1377 # TODO Think about returning some lazy-eval iterator.
1378 # TODO Get rid of backup; we should know from what witness is whether we need it.
1380 sub reading_sequence {
1381 my( $self, $start, $end, $witness ) = @_;
1383 $witness = $self->baselabel unless $witness;
1384 my @readings = ( $start );
1387 while( $n && $n->id ne $end->id ) {
1388 if( exists( $seen{$n->id} ) ) {
1389 throw( "Detected loop for $witness at " . $n->id );
1393 my $next = $self->next_reading( $n, $witness );
1395 throw( "Did not find any path for $witness from reading " . $n->id );
1397 push( @readings, $next );
1400 # Check that the last reading is our end reading.
1401 my $last = $readings[$#readings];
1402 throw( "Last reading found from " . $start->text .
1403 " for witness $witness is not the end!" ) # TODO do we get this far?
1404 unless $last->id eq $end->id;
1409 =head2 next_reading( $reading, $sigil );
1411 Returns the reading that follows the given reading along the given witness
1417 # Return the successor via the corresponding path.
1419 my $answer = $self->_find_linked_reading( 'next', @_ );
1420 return undef unless $answer;
1421 return $self->reading( $answer );
1424 =head2 prior_reading( $reading, $sigil )
1426 Returns the reading that precedes the given reading along the given witness
1432 # Return the predecessor via the corresponding path.
1434 my $answer = $self->_find_linked_reading( 'prior', @_ );
1435 return $self->reading( $answer );
1438 sub _find_linked_reading {
1439 my( $self, $direction, $node, $path ) = @_;
1441 # Get a backup if we are dealing with a layered witness
1443 my $aclabel = $self->ac_label;
1444 if( $path && $path =~ /^(.*)\Q$aclabel\E$/ ) {
1448 my @linked_paths = $direction eq 'next'
1449 ? $self->sequence->edges_from( $node )
1450 : $self->sequence->edges_to( $node );
1451 return undef unless scalar( @linked_paths );
1453 # We have to find the linked path that contains all of the
1454 # witnesses supplied in $path.
1455 my( @path_wits, @alt_path_wits );
1456 @path_wits = sort( $self->_witnesses_of_label( $path ) ) if $path;
1457 @alt_path_wits = sort( $self->_witnesses_of_label( $alt_path ) ) if $alt_path;
1460 foreach my $le ( @linked_paths ) {
1461 if( $self->sequence->has_edge_attribute( @$le, $self->baselabel ) ) {
1464 my @le_wits = sort $self->path_witnesses( $le );
1465 if( _is_within( \@path_wits, \@le_wits ) ) {
1466 # This is the right path.
1467 return $direction eq 'next' ? $le->[1] : $le->[0];
1468 } elsif( _is_within( \@alt_path_wits, \@le_wits ) ) {
1472 # Got this far? Return the alternate path if it exists.
1473 return $direction eq 'next' ? $alt_le->[1] : $alt_le->[0]
1476 # Got this far? Return the base path if it exists.
1477 return $direction eq 'next' ? $base_le->[1] : $base_le->[0]
1480 # Got this far? We have no appropriate path.
1481 warn "Could not find $direction node from " . $node->id
1482 . " along path $path";
1488 my( $set1, $set2 ) = @_;
1489 my $ret = @$set1; # will be 0, i.e. false, if set1 is empty
1490 foreach my $el ( @$set1 ) {
1491 $ret = 0 unless grep { /^\Q$el\E$/ } @$set2;
1496 # Return the string that joins together a list of witnesses for
1497 # display on a single path.
1498 sub _witnesses_of_label {
1499 my( $self, $label ) = @_;
1500 my $regex = $self->wit_list_separator;
1501 my @answer = split( /\Q$regex\E/, $label );
1505 =head2 common_readings
1507 Returns the list of common readings in the graph (i.e. those readings that are
1508 shared by all non-lacunose witnesses.)
1512 sub common_readings {
1514 my @common = grep { $_->is_common } $self->readings;
1518 =head2 path_text( $sigil, [, $start, $end ] )
1520 Returns the text of a witness (plus its backup, if we are using a layer)
1521 as stored in the collation. The text is returned as a string, where the
1522 individual readings are joined with spaces and the meta-readings (e.g.
1523 lacunae) are omitted. Optional specification of $start and $end allows
1524 the generation of a subset of the witness text.
1529 my( $self, $wit, $start, $end ) = @_;
1530 $start = $self->start unless $start;
1531 $end = $self->end unless $end;
1532 my @path = grep { !$_->is_meta } $self->reading_sequence( $start, $end, $wit );
1535 foreach my $r ( @path ) {
1536 unless ( $r->join_prior || !$last || $last->join_next ) {
1539 $pathtext .= $r->text;
1545 =head1 INITIALIZATION METHODS
1547 These are mostly for use by parsers.
1549 =head2 make_witness_path( $witness )
1551 Link the array of readings contained in $witness->path (and in
1552 $witness->uncorrected_path if it exists) into collation paths.
1553 Clear out the arrays when finished.
1555 =head2 make_witness_paths
1557 Call make_witness_path for all witnesses in the tradition.
1561 # For use when a collation is constructed from a base text and an apparatus.
1562 # We have the sequences of readings and just need to add path edges.
1563 # When we are done, clear out the witness path attributes, as they are no
1565 # TODO Find a way to replace the witness path attributes with encapsulated functions?
1567 sub make_witness_paths {
1569 foreach my $wit ( $self->tradition->witnesses ) {
1570 # say STDERR "Making path for " . $wit->sigil;
1571 $self->make_witness_path( $wit );
1575 sub make_witness_path {
1576 my( $self, $wit ) = @_;
1577 my @chain = @{$wit->path};
1578 my $sig = $wit->sigil;
1579 # Add start and end if necessary
1580 unshift( @chain, $self->start ) unless $chain[0] eq $self->start;
1581 push( @chain, $self->end ) unless $chain[-1] eq $self->end;
1582 foreach my $idx ( 0 .. $#chain-1 ) {
1583 $self->add_path( $chain[$idx], $chain[$idx+1], $sig );
1585 if( $wit->is_layered ) {
1586 @chain = @{$wit->uncorrected_path};
1587 unshift( @chain, $self->start ) unless $chain[0] eq $self->start;
1588 push( @chain, $self->end ) unless $chain[-1] eq $self->end;
1589 foreach my $idx( 0 .. $#chain-1 ) {
1590 my $source = $chain[$idx];
1591 my $target = $chain[$idx+1];
1592 $self->add_path( $source, $target, $sig.$self->ac_label )
1593 unless $self->has_path( $source, $target, $sig );
1597 $wit->clear_uncorrected_path;
1600 =head2 calculate_ranks
1602 Calculate the reading ranks (that is, their aligned positions relative
1603 to each other) for the graph. This can only be called on linear collations.
1607 use Text::Tradition;
1609 my $cxfile = 't/data/Collatex-16.xml';
1610 my $t = Text::Tradition->new(
1612 'input' => 'CollateX',
1615 my $c = $t->collation;
1618 my $table = $c->alignment_table;
1619 ok( $c->has_cached_table, "Alignment table was cached" );
1620 is( $c->alignment_table, $table, "Cached table returned upon second call" );
1621 $c->calculate_ranks;
1622 is( $c->alignment_table, $table, "Cached table retained with no rank change" );
1623 $c->add_relationship( 'n24', 'n23', { 'type' => 'spelling' } );
1624 isnt( $c->alignment_table, $table, "Alignment table changed after relationship add" );
1630 sub calculate_ranks {
1632 # Save the existing ranks, in case we need to invalidate the cached SVG.
1634 map { $existing_ranks{$_} = $_->rank } $self->readings;
1636 # Do the rankings based on the relationship equivalence graph, starting
1637 # with the start node.
1638 my ( $node_ranks, $rank_nodes ) = $self->relations->equivalence_ranks();
1640 # Transfer our rankings from the topological graph to the real one.
1641 foreach my $r ( $self->readings ) {
1642 if( defined $node_ranks->{$self->equivalence( $r->id )} ) {
1643 $r->rank( $node_ranks->{$self->equivalence( $r->id )} );
1645 # Die. Find the last rank we calculated.
1646 my @all_defined = sort { ( $node_ranks->{$self->equivalence( $a->id )}||-1 )
1647 <=> ( $node_ranks->{$self->equivalence( $b->id )}||-1 ) }
1649 my $last = pop @all_defined;
1650 throw( "Ranks not calculated after $last - do you have a cycle in the graph?" );
1653 # Do we need to invalidate the cached data?
1654 if( $self->has_cached_svg || $self->has_cached_table ) {
1655 foreach my $r ( $self->readings ) {
1656 next if defined( $existing_ranks{$r} )
1657 && $existing_ranks{$r} == $r->rank;
1658 # Something has changed, so clear the cache
1659 $self->_clear_cache;
1660 # ...and recalculate the common readings.
1661 $self->calculate_common_readings();
1665 # The graph calculation information is now up to date.
1666 $self->_graphcalc_done(1);
1671 $self->wipe_svg if $self->has_cached_svg;
1672 $self->wipe_table if $self->has_cached_table;
1676 =head2 flatten_ranks
1678 A convenience method for parsing collation data. Searches the graph for readings
1679 with the same text at the same rank, and merges any that are found.
1685 my %unique_rank_rdg;
1687 foreach my $rdg ( $self->readings ) {
1688 next unless $rdg->has_rank;
1689 my $key = $rdg->rank . "||" . $rdg->text;
1690 if( exists $unique_rank_rdg{$key} ) {
1691 # Make sure they don't have different grammatical forms
1692 my $ur = $unique_rank_rdg{$key};
1693 if( $rdg->disambiguated && $ur->disambiguated ) {
1694 my $rform = join( '//', map { $_->form->to_string } $rdg->lexemes );
1695 my $uform = join( '//', map { $_->form->to_string } $ur->lexemes );
1696 next unless $rform eq $uform;
1697 } elsif( $rdg->disambiguated xor $ur->disambiguated ) {
1701 #say STDERR "Combining readings at same rank: $key";
1703 $self->merge_readings( $unique_rank_rdg{$key}, $rdg );
1704 # TODO see if this now makes a common point.
1706 $unique_rank_rdg{$key} = $rdg;
1709 # If we merged readings, the ranks are still fine but the alignment
1710 # table is wrong. Wipe it.
1711 $self->wipe_table() if $changed;
1715 =head2 calculate_common_readings
1717 Goes through the graph identifying the readings that appear in every witness
1718 (apart from those with lacunae at that spot.) Marks them as common and returns
1723 use Text::Tradition;
1725 my $cxfile = 't/data/Collatex-16.xml';
1726 my $t = Text::Tradition->new(
1728 'input' => 'CollateX',
1731 my $c = $t->collation;
1733 my @common = $c->calculate_common_readings();
1734 is( scalar @common, 8, "Found correct number of common readings" );
1735 my @marked = sort $c->common_readings();
1736 is( scalar @common, 8, "All common readings got marked as such" );
1737 my @expected = qw/ n1 n11 n16 n19 n20 n5 n6 n7 /;
1738 is_deeply( \@marked, \@expected, "Found correct list of common readings" );
1744 sub calculate_common_readings {
1747 map { $_->is_common( 0 ) } $self->readings;
1748 # Implicitly calls calculate_ranks
1749 my $table = $self->alignment_table;
1750 foreach my $idx ( 0 .. $table->{'length'} - 1 ) {
1751 my @row = map { $_->{'tokens'}->[$idx]
1752 ? $_->{'tokens'}->[$idx]->{'t'} : '' }
1753 @{$table->{'alignment'}};
1755 foreach my $r ( @row ) {
1757 $hash{$r->id} = $r unless $r->is_meta;
1759 $hash{'UNDEF'} = $r;
1762 if( keys %hash == 1 && !exists $hash{'UNDEF'} ) {
1763 my( $r ) = values %hash;
1765 push( @common, $r );
1771 =head2 text_from_paths
1773 Calculate the text array for all witnesses from the path, for later consistency
1774 checking. Only to be used if there is no non-graph-based way to know the
1779 sub text_from_paths {
1781 foreach my $wit ( $self->tradition->witnesses ) {
1782 my @readings = $self->reading_sequence( $self->start, $self->end, $wit->sigil );
1784 foreach my $r ( @readings ) {
1785 next if $r->is_meta;
1786 push( @text, $r->text );
1788 $wit->text( \@text );
1789 if( $wit->is_layered ) {
1790 my @ucrdgs = $self->reading_sequence( $self->start, $self->end,
1791 $wit->sigil.$self->ac_label );
1793 foreach my $r ( @ucrdgs ) {
1794 next if $r->is_meta;
1795 push( @uctext, $r->text );
1797 $wit->layertext( \@uctext );
1802 =head1 UTILITY FUNCTIONS
1804 =head2 common_predecessor( $reading_a, $reading_b )
1806 Find the last reading that occurs in sequence before both the given readings.
1807 At the very least this should be $self->start.
1809 =head2 common_successor( $reading_a, $reading_b )
1811 Find the first reading that occurs in sequence after both the given readings.
1812 At the very least this should be $self->end.
1816 use Text::Tradition;
1818 my $cxfile = 't/data/Collatex-16.xml';
1819 my $t = Text::Tradition->new(
1821 'input' => 'CollateX',
1824 my $c = $t->collation;
1826 is( $c->common_predecessor( 'n24', 'n23' )->id,
1827 'n20', "Found correct common predecessor" );
1828 is( $c->common_successor( 'n24', 'n23' )->id,
1829 '__END__', "Found correct common successor" );
1831 is( $c->common_predecessor( 'n19', 'n17' )->id,
1832 'n16', "Found correct common predecessor for readings on same path" );
1833 is( $c->common_successor( 'n21', 'n10' )->id,
1834 '__END__', "Found correct common successor for readings on same path" );
1840 ## Return the closest reading that is a predecessor of both the given readings.
1841 sub common_predecessor {
1843 my( $r1, $r2 ) = $self->_objectify_args( @_ );
1844 return $self->_common_in_path( $r1, $r2, 'predecessors' );
1847 sub common_successor {
1849 my( $r1, $r2 ) = $self->_objectify_args( @_ );
1850 return $self->_common_in_path( $r1, $r2, 'successors' );
1854 # TODO think about how to do this without ranks...
1855 sub _common_in_path {
1856 my( $self, $r1, $r2, $dir ) = @_;
1857 my $iter = $self->end->rank;
1859 my @last_r1 = ( $r1 );
1860 my @last_r2 = ( $r2 );
1861 # my %all_seen = ( $r1 => 'r1', $r2 => 'r2' );
1863 # say STDERR "Finding common $dir for $r1, $r2";
1864 while( !@candidates ) {
1865 last unless $iter--; # Avoid looping infinitely
1866 # Iterate separately down the graph from r1 and r2
1867 my( @new_lc1, @new_lc2 );
1868 foreach my $lc ( @last_r1 ) {
1869 foreach my $p ( $lc->$dir ) {
1870 if( $all_seen{$p->id} && $all_seen{$p->id} ne 'r1' ) {
1871 # say STDERR "Path candidate $p from $lc";
1872 push( @candidates, $p );
1873 } elsif( !$all_seen{$p->id} ) {
1874 $all_seen{$p->id} = 'r1';
1875 push( @new_lc1, $p );
1879 foreach my $lc ( @last_r2 ) {
1880 foreach my $p ( $lc->$dir ) {
1881 if( $all_seen{$p->id} && $all_seen{$p->id} ne 'r2' ) {
1882 # say STDERR "Path candidate $p from $lc";
1883 push( @candidates, $p );
1884 } elsif( !$all_seen{$p->id} ) {
1885 $all_seen{$p->id} = 'r2';
1886 push( @new_lc2, $p );
1890 @last_r1 = @new_lc1;
1891 @last_r2 = @new_lc2;
1893 my @answer = sort { $a->rank <=> $b->rank } @candidates;
1894 return $dir eq 'predecessors' ? pop( @answer ) : shift ( @answer );
1898 Text::Tradition::Error->throw(
1899 'ident' => 'Collation error',
1905 __PACKAGE__->meta->make_immutable;
1909 This package is free software and is provided "as is" without express
1910 or implied warranty. You can redistribute it and/or modify it under
1911 the same terms as Perl itself.
1915 Tara L Andrews E<lt>aurum@cpan.orgE<gt>