1 package Text::Tradition::Collation;
4 use Encode qw( decode_utf8 );
8 use IPC::Run qw( run binary );
10 use Text::Tradition::Collation::Data;
11 use Text::Tradition::Collation::Reading;
12 use Text::Tradition::Collation::RelationshipStore;
13 use Text::Tradition::Error;
14 use XML::Easy::Syntax qw( $xml10_namestartchar_rx $xml10_namechar_rx );
16 use XML::LibXML::XPathContext;
20 isa => 'Text::Tradition::Collation::Data',
58 isa => 'Text::Tradition',
59 writer => '_set_tradition',
65 Text::Tradition::Collation - a software model for a text collation
70 my $t = Text::Tradition->new(
71 'name' => 'this is a text',
73 'file' => '/path/to/tei_parallel_seg_file.xml' );
75 my $c = $t->collation;
76 my @readings = $c->readings;
77 my @paths = $c->paths;
78 my @relationships = $c->relationships;
80 my $svg_variant_graph = $t->collation->as_svg();
84 Text::Tradition is a library for representation and analysis of collated
85 texts, particularly medieval ones. The Collation is the central feature of
86 a Tradition, where the text, its sequence of readings, and its relationships
87 between readings are actually kept.
93 The constructor. Takes a hash or hashref of the following arguments:
97 =item * tradition - The Text::Tradition object to which the collation
100 =item * linear - Whether the collation should be linear; that is, whether
101 transposed readings should be treated as two linked readings rather than one,
102 and therefore whether the collation graph is acyclic. Defaults to true.
104 =item * baselabel - The default label for the path taken by a base text
105 (if any). Defaults to 'base text'.
107 =item * wit_list_separator - The string to join a list of witnesses for
108 purposes of making labels in display graphs. Defaults to ', '.
110 =item * ac_label - The extra label to tack onto a witness sigil when
111 representing another layer of path for the given witness - that is, when
112 a text has more than one possible reading due to scribal corrections or
113 the like. Defaults to ' (a.c.)'.
115 =item * wordsep - The string used to separate words in the original text.
126 =head2 wit_list_separator
134 Simple accessors for collation attributes.
138 The meta-reading at the start of every witness path.
142 The meta-reading at the end of every witness path.
146 Returns all Reading objects in the graph.
148 =head2 reading( $id )
150 Returns the Reading object corresponding to the given ID.
152 =head2 add_reading( $reading_args )
154 Adds a new reading object to the collation.
155 See L<Text::Tradition::Collation::Reading> for the available arguments.
157 =head2 del_reading( $object_or_id )
159 Removes the given reading from the collation, implicitly removing its
160 paths and relationships.
162 =head2 merge_readings( $main, $second, $concatenate, $with_str )
164 Merges the $second reading into the $main one. If $concatenate is true, then
165 the merged node will carry the text of both readings, concatenated with either
166 $with_str (if specified) or a sensible default (the empty string if the
167 appropriate 'join_*' flag is set on either reading, or else $self->wordsep.)
169 The first two arguments may be either readings or reading IDs.
171 =head2 has_reading( $id )
173 Predicate to see whether a given reading ID is in the graph.
175 =head2 reading_witnesses( $object_or_id )
177 Returns a list of sigils whose witnesses contain the reading.
181 Returns all reading paths within the document - that is, all edges in the
182 collation graph. Each path is an arrayref of [ $source, $target ] reading IDs.
184 =head2 add_path( $source, $target, $sigil )
186 Links the given readings in the collation in sequence, under the given witness
187 sigil. The readings may be specified by object or ID.
189 =head2 del_path( $source, $target, $sigil )
191 Links the given readings in the collation in sequence, under the given witness
192 sigil. The readings may be specified by object or ID.
194 =head2 has_path( $source, $target );
196 Returns true if the two readings are linked in sequence in any witness.
197 The readings may be specified by object or ID.
201 Returns all Relationship objects in the collation.
203 =head2 add_relationship( $reading, $other_reading, $options )
205 Adds a new relationship of the type given in $options between the two readings,
206 which may be specified by object or ID. Returns a value of ( $status, @vectors)
207 where $status is true on success, and @vectors is a list of relationship edges
208 that were ultimately added.
209 See L<Text::Tradition::Collation::Relationship> for the available options.
214 my ( $class, @args ) = @_;
215 my %args = @args == 1 ? %{ $args[0] } : @args;
216 # TODO determine these from the Moose::Meta object
217 my @delegate_attrs = qw(sequence relations readings wit_list_separator baselabel
218 linear wordsep start end cached_table _graphcalc_done);
220 for my $attr (@delegate_attrs) {
221 $data_args{$attr} = delete $args{$attr} if exists $args{$attr};
223 $args{_data} = Text::Tradition::Collation::Data->new(%data_args);
229 $self->_set_relations( Text::Tradition::Collation::RelationshipStore->new( 'collation' => $self ) );
230 $self->_set_start( $self->add_reading(
231 { 'collation' => $self, 'is_start' => 1, 'init' => 1 } ) );
232 $self->_set_end( $self->add_reading(
233 { 'collation' => $self, 'is_end' => 1, 'init' => 1 } ) );
236 ### Reading construct/destruct functions
239 my( $self, $reading ) = @_;
240 unless( ref( $reading ) eq 'Text::Tradition::Collation::Reading' ) {
241 my %args = %$reading;
242 if( $args{'init'} ) {
243 # If we are initializing an empty collation, don't assume that we
244 # have set a tradition.
245 delete $args{'init'};
246 } elsif( $self->tradition->has_language && !exists $args{'language'} ) {
247 $args{'language'} = $self->tradition->language;
249 $reading = Text::Tradition::Collation::Reading->new(
250 'collation' => $self,
253 # First check to see if a reading with this ID exists.
254 if( $self->reading( $reading->id ) ) {
255 throw( "Collation already has a reading with id " . $reading->id );
257 $self->_graphcalc_done(0);
258 $self->_add_reading( $reading->id => $reading );
259 # Once the reading has been added, put it in both graphs.
260 $self->sequence->add_vertex( $reading->id );
261 $self->relations->add_reading( $reading->id );
265 around del_reading => sub {
270 if( ref( $arg ) eq 'Text::Tradition::Collation::Reading' ) {
273 # Remove the reading from the graphs.
274 $self->_graphcalc_done(0);
275 $self->_clear_cache; # Explicitly clear caches to GC the reading
276 $self->sequence->delete_vertex( $arg );
277 $self->relations->delete_reading( $arg );
280 $self->$orig( $arg );
287 my $cxfile = 't/data/Collatex-16.xml';
288 my $t = Text::Tradition->new(
290 'input' => 'CollateX',
293 my $c = $t->collation;
295 my $rno = scalar $c->readings;
296 # Split n21 for testing purposes
297 my $new_r = $c->add_reading( { 'id' => 'n21p0', 'text' => 'un', 'join_next' => 1 } );
298 my $old_r = $c->reading( 'n21' );
299 $old_r->alter_text( 'to' );
300 $c->del_path( 'n20', 'n21', 'A' );
301 $c->add_path( 'n20', 'n21p0', 'A' );
302 $c->add_path( 'n21p0', 'n21', 'A' );
304 ok( $c->reading( 'n21p0' ), "New reading exists" );
305 is( scalar $c->readings, $rno, "Reading add offset by flatten_ranks" );
307 # Combine n3 and n4 ( with his )
308 $c->merge_readings( 'n3', 'n4', 1 );
309 ok( !$c->reading('n4'), "Reading n4 is gone" );
310 is( $c->reading('n3')->text, 'with his', "Reading n3 has both words" );
312 # Collapse n9 and n10 ( rood / root )
313 $c->merge_readings( 'n9', 'n10' );
314 ok( !$c->reading('n10'), "Reading n10 is gone" );
315 is( $c->reading('n9')->text, 'rood', "Reading n9 has an unchanged word" );
317 # Combine n21 and n21p0
318 my $remaining = $c->reading('n21');
319 $remaining ||= $c->reading('n22'); # one of these should still exist
320 $c->merge_readings( 'n21p0', $remaining, 1 );
321 ok( !$c->reading('n21'), "Reading $remaining is gone" );
322 is( $c->reading('n21p0')->text, 'unto', "Reading n21p0 merged correctly" );
332 my( $kept_obj, $del_obj, $combine, $combine_char ) = $self->_objectify_args( @_ );
333 my $mergemeta = $kept_obj->is_meta;
334 throw( "Cannot merge meta and non-meta reading" )
335 unless ( $mergemeta && $del_obj->is_meta )
336 || ( !$mergemeta && !$del_obj->is_meta );
338 throw( "Cannot merge with start or end node" )
339 if( $kept_obj eq $self->start || $kept_obj eq $self->end
340 || $del_obj eq $self->start || $del_obj eq $self->end );
342 # We only need the IDs for adding paths to the graph, not the reading
343 # objects themselves.
344 my $kept = $kept_obj->id;
345 my $deleted = $del_obj->id;
346 $self->_graphcalc_done(0);
348 # The kept reading should inherit the paths and the relationships
349 # of the deleted reading.
350 foreach my $path ( $self->sequence->edges_at( $deleted ) ) {
351 my @vector = ( $kept );
352 push( @vector, $path->[1] ) if $path->[0] eq $deleted;
353 unshift( @vector, $path->[0] ) if $path->[1] eq $deleted;
354 next if $vector[0] eq $vector[1]; # Don't add a self loop
355 my %wits = %{$self->sequence->get_edge_attributes( @$path )};
356 $self->sequence->add_edge( @vector );
357 my $fwits = $self->sequence->get_edge_attributes( @vector );
358 @wits{keys %$fwits} = values %$fwits;
359 $self->sequence->set_edge_attributes( @vector, \%wits );
361 $self->relations->merge_readings( $kept, $deleted, $combine );
363 # Do the deletion deed.
365 # Combine the text of the readings
366 my $joinstr = $combine_char;
367 unless( defined $joinstr ) {
368 $joinstr = '' if $kept_obj->join_next || $del_obj->join_prior;
369 $joinstr = $self->wordsep unless defined $joinstr;
371 $kept_obj->alter_text( join( $joinstr, $kept_obj->text, $del_obj->text ) );
372 # Change this reading to a joining one if necessary
373 $kept_obj->_set_join_next( $del_obj->join_next );
374 $kept_obj->normal_form(
375 join( $joinstr, $kept_obj->normal_form, $del_obj->normal_form ) );
376 # Combine the lexemes present in the readings
377 if( $kept_obj->has_lexemes && $del_obj->has_lexemes ) {
378 $kept_obj->add_lexeme( $del_obj->lexemes );
381 $self->del_reading( $deleted );
384 =head2 compress_readings
386 Where possible in the graph, compresses plain sequences of readings into a
387 single reading. The sequences must consist of readings with no
388 relationships to other readings, with only a single witness path between
389 them and no other witness paths from either that would skip the other. The
390 readings must also not be marked as nonsense or bad grammar.
392 WARNING: This operation cannot be undone.
396 sub compress_readings {
398 # Anywhere in the graph that there is a reading that joins only to a single
399 # successor, and neither of these have any relationships, just join the two
402 foreach my $rdg ( sort { $a->rank <=> $b->rank } $self->readings ) {
403 # While we are here, get rid of any extra wordforms from a disambiguated
405 if( $rdg->disambiguated ) {
406 foreach my $lex ( $rdg->lexemes ) {
407 $lex->clear_matching_forms();
408 $lex->add_matching_form( $lex->form );
411 # Now look for readings that can be joined to their successors.
412 next if $rdg->is_meta;
413 next if $gobbled{$rdg->id};
414 next if $rdg->grammar_invalid || $rdg->is_nonsense;
415 next if $rdg->related_readings();
417 while( $self->sequence->successors( $rdg ) == 1 ) {
418 my( $next ) = $self->reading( $self->sequence->successors( $rdg ) );
419 throw( "Infinite loop" ) if $seen{$next->id};
420 $seen{$next->id} = 1;
421 last if $self->sequence->predecessors( $next ) > 1;
422 last if $next->is_meta;
423 last if $next->grammar_invalid || $next->is_nonsense;
424 last if $next->related_readings();
425 say "Joining readings $rdg and $next";
426 $self->merge_readings( $rdg, $next, 1 );
429 # Make sure we haven't screwed anything up
430 foreach my $wit ( $self->tradition->witnesses ) {
431 my $pathtext = $self->path_text( $wit->sigil );
432 my $origtext = join( ' ', @{$wit->text} );
433 throw( "Text differs for witness " . $wit->sigil )
434 unless $pathtext eq $origtext;
435 if( $wit->is_layered ) {
436 $pathtext = $self->path_text( $wit->sigil.$self->ac_label );
437 $origtext = join( ' ', @{$wit->layertext} );
438 throw( "Ante-corr text differs for witness " . $wit->sigil )
439 unless $pathtext eq $origtext;
443 $self->relations->rebuild_equivalence();
444 $self->calculate_ranks();
447 # Helper function for manipulating the graph.
448 sub _stringify_args {
449 my( $self, $first, $second, @args ) = @_;
451 if ref( $first ) eq 'Text::Tradition::Collation::Reading';
452 $second = $second->id
453 if ref( $second ) eq 'Text::Tradition::Collation::Reading';
454 return( $first, $second, @args );
457 # Helper function for manipulating the graph.
458 sub _objectify_args {
459 my( $self, $first, $second, $arg ) = @_;
460 $first = $self->reading( $first )
461 unless ref( $first ) eq 'Text::Tradition::Collation::Reading';
462 $second = $self->reading( $second )
463 unless ref( $second ) eq 'Text::Tradition::Collation::Reading';
464 return( $first, $second, $arg );
471 # We only need the IDs for adding paths to the graph, not the reading
472 # objects themselves.
473 my( $source, $target, $wit ) = $self->_stringify_args( @_ );
475 $self->_graphcalc_done(0);
476 # Connect the readings
477 unless( $self->sequence->has_edge( $source, $target ) ) {
478 $self->sequence->add_edge( $source, $target );
479 $self->relations->add_equivalence_edge( $source, $target );
481 # Note the witness in question
482 $self->sequence->set_edge_attribute( $source, $target, $wit, 1 );
488 if( ref( $_[0] ) eq 'ARRAY' ) {
495 # We only need the IDs for adding paths to the graph, not the reading
496 # objects themselves.
497 my( $source, $target, $wit ) = $self->_stringify_args( @args );
499 $self->_graphcalc_done(0);
500 if( $self->sequence->has_edge_attribute( $source, $target, $wit ) ) {
501 $self->sequence->delete_edge_attribute( $source, $target, $wit );
503 unless( keys %{$self->sequence->get_edge_attributes( $source, $target )} ) {
504 $self->sequence->delete_edge( $source, $target );
505 $self->relations->delete_equivalence_edge( $source, $target );
510 # Extra graph-alike utility
513 my( $source, $target, $wit ) = $self->_stringify_args( @_ );
514 return undef unless $self->sequence->has_edge( $source, $target );
515 return $self->sequence->has_edge_attribute( $source, $target, $wit );
518 =head2 clear_witness( @sigil_list )
520 Clear the given witnesses out of the collation entirely, removing references
521 to them in paths, and removing readings that belong only to them. Should only
522 be called via $tradition->del_witness.
527 my( $self, @sigils ) = @_;
529 $self->_graphcalc_done(0);
530 # Clear the witness(es) out of the paths
531 foreach my $e ( $self->paths ) {
532 foreach my $sig ( @sigils ) {
533 $self->del_path( $e, $sig );
537 # Clear out the newly unused readings
538 foreach my $r ( $self->readings ) {
539 unless( $self->reading_witnesses( $r ) ) {
540 $self->del_reading( $r );
545 sub add_relationship {
547 my( $source, $target, $opts ) = $self->_stringify_args( @_ );
548 my( @vectors ) = $self->relations->add_relationship( $source, $target, $opts );
549 $self->_graphcalc_done(0);
553 around qw/ get_relationship del_relationship / => sub {
557 if( @args == 1 && ref( $args[0] ) eq 'ARRAY' ) {
560 my( $source, $target ) = $self->_stringify_args( @args );
561 $self->$orig( $source, $target );
564 =head2 reading_witnesses( $reading )
566 Return a list of sigils corresponding to the witnesses in which the reading appears.
570 sub reading_witnesses {
571 my( $self, $reading ) = @_;
572 # We need only check either the incoming or the outgoing edges; I have
573 # arbitrarily chosen "incoming". Thus, special-case the start node.
574 if( $reading eq $self->start ) {
575 return map { $_->sigil } $self->tradition->witnesses;
578 foreach my $e ( $self->sequence->edges_to( $reading ) ) {
579 my $wits = $self->sequence->get_edge_attributes( @$e );
580 @all_witnesses{ keys %$wits } = 1;
582 my $acstr = $self->ac_label;
583 foreach my $acwit ( grep { $_ =~ s/^(.*)\Q$acstr\E$/$1/ } keys %all_witnesses ) {
584 delete $all_witnesses{$acwit.$acstr} if exists $all_witnesses{$acwit};
586 return keys %all_witnesses;
589 =head1 OUTPUT METHODS
591 =head2 as_svg( \%options )
593 Returns an SVG string that represents the graph, via as_dot and graphviz.
594 See as_dot for a list of options. Must have GraphViz (dot) installed to run.
599 my( $self, $opts ) = @_;
600 throw( "Need GraphViz installed to output SVG" )
601 unless File::Which::which( 'dot' );
602 my $want_subgraph = exists $opts->{'from'} || exists $opts->{'to'};
603 $self->calculate_ranks()
604 unless( $self->_graphcalc_done || $opts->{'nocalc'} || !$self->linear );
605 my @cmd = qw/dot -Tsvg/;
607 my $dotfile = File::Temp->new();
609 # $dotfile->unlink_on_destroy(0);
610 binmode $dotfile, ':utf8';
611 print $dotfile $self->as_dot( $opts );
612 push( @cmd, $dotfile->filename );
613 run( \@cmd, ">", binary(), \$svg );
614 $svg = decode_utf8( $svg );
619 =head2 as_dot( \%options )
621 Returns a string that is the collation graph expressed in dot
622 (i.e. GraphViz) format. Options include:
637 my( $self, $opts ) = @_;
638 my $startrank = $opts->{'from'} if $opts;
639 my $endrank = $opts->{'to'} if $opts;
640 my $color_common = $opts->{'color_common'} if $opts;
641 my $STRAIGHTENHACK = !$startrank && !$endrank && $self->end->rank
642 && $self->end->rank > 100;
643 $STRAIGHTENHACK = 1 if $opts->{'straight'}; # even for subgraphs or small graphs
645 # Check the arguments
647 return if $endrank && $startrank > $endrank;
648 return if $startrank > $self->end->rank;
650 if( defined $endrank ) {
651 return if $endrank < 0;
652 $endrank = undef if $endrank == $self->end->rank;
655 my $graph_name = $self->tradition->name;
656 $graph_name =~ s/[^\w\s]//g;
657 $graph_name = join( '_', split( /\s+/, $graph_name ) );
665 'fillcolor' => 'white',
670 'arrowhead' => 'open',
671 'color' => '#000000',
672 'fontcolor' => '#000000',
675 my $dot = sprintf( "digraph %s {\n", $graph_name );
676 $dot .= "\tgraph " . _dot_attr_string( \%graph_attrs ) . ";\n";
677 $dot .= "\tnode " . _dot_attr_string( \%node_attrs ) . ";\n";
679 # Output substitute start/end readings if necessary
681 $dot .= "\t\"__SUBSTART__\" [ label=\"...\",id=\"__START__\" ];\n";
684 $dot .= "\t\"__SUBEND__\" [ label=\"...\",id=\"__END__\" ];\n";
686 if( $STRAIGHTENHACK ) {
688 my $startlabel = $startrank ? '__SUBSTART__' : '__START__';
689 $dot .= "\tsubgraph { rank=same \"$startlabel\" \"#SILENT#\" }\n";
690 $dot .= "\t\"#SILENT#\" [ shape=diamond,color=white,penwidth=0,label=\"\" ];"
692 my %used; # Keep track of the readings that actually appear in the graph
693 # Sort the readings by rank if we have ranks; this speeds layout.
694 my @all_readings = $self->end->has_rank
695 ? sort { $a->rank <=> $b->rank } $self->readings
697 # TODO Refrain from outputting lacuna nodes - just grey out the edges.
698 foreach my $reading ( @all_readings ) {
699 # Only output readings within our rank range.
700 next if $startrank && $reading->rank < $startrank;
701 next if $endrank && $reading->rank > $endrank;
702 $used{$reading->id} = 1;
703 # Need not output nodes without separate labels
704 next if $reading->id eq $reading->text;
706 my $label = $reading->text;
707 $label .= '-' if $reading->join_next;
708 $label = "-$label" if $reading->join_prior;
709 $label =~ s/\"/\\\"/g;
710 $rattrs->{'label'} = $label;
711 $rattrs->{'id'} = $reading->id;
712 $rattrs->{'fillcolor'} = '#b3f36d' if $reading->is_common && $color_common;
713 $dot .= sprintf( "\t\"%s\" %s;\n", $reading->id, _dot_attr_string( $rattrs ) );
716 # Add the real edges. Need to weight one edge per rank jump, in a
718 # my $weighted = $self->_add_edge_weights;
719 my @edges = $self->paths;
720 my( %substart, %subend );
721 foreach my $edge ( @edges ) {
722 # Do we need to output this edge?
723 if( $used{$edge->[0]} && $used{$edge->[1]} ) {
724 my $label = $self->_path_display_label( $self->path_witnesses( $edge ) );
725 my $variables = { %edge_attrs, 'label' => $label };
727 # Account for the rank gap if necessary
728 my $rank0 = $self->reading( $edge->[0] )->rank
729 if $self->reading( $edge->[0] )->has_rank;
730 my $rank1 = $self->reading( $edge->[1] )->rank
731 if $self->reading( $edge->[1] )->has_rank;
732 if( defined $rank0 && defined $rank1 && $rank1 - $rank0 > 1 ) {
733 $variables->{'minlen'} = $rank1 - $rank0;
736 # Add the calculated edge weights
737 # if( exists $weighted->{$edge->[0]}
738 # && $weighted->{$edge->[0]} eq $edge->[1] ) {
739 # # $variables->{'color'} = 'red';
740 # $variables->{'weight'} = 3.0;
743 # EXPERIMENTAL: make edge width reflect no. of witnesses
744 my $extrawidth = scalar( $self->path_witnesses( $edge ) ) * 0.2;
745 $variables->{'penwidth'} = $extrawidth + 0.8; # gives 1 for a single wit
747 my $varopts = _dot_attr_string( $variables );
748 $dot .= sprintf( "\t\"%s\" -> \"%s\" %s;\n",
749 $edge->[0], $edge->[1], $varopts );
750 } elsif( $used{$edge->[0]} ) {
751 $subend{$edge->[0]} = $edge->[1];
752 } elsif( $used{$edge->[1]} ) {
753 $substart{$edge->[1]} = $edge->[0];
756 # Add substitute start and end edges if necessary
757 foreach my $node ( keys %substart ) {
758 my $witstr = $self->_path_display_label ( $self->path_witnesses( $substart{$node}, $node ) );
759 my $variables = { %edge_attrs, 'label' => $witstr };
760 my $nrdg = $self->reading( $node );
761 if( $nrdg->has_rank && $nrdg->rank > $startrank ) {
762 # Substart is actually one lower than $startrank
763 $variables->{'minlen'} = $nrdg->rank - ( $startrank - 1 );
765 my $varopts = _dot_attr_string( $variables );
766 $dot .= "\t\"__SUBSTART__\" -> \"$node\" $varopts;\n";
768 foreach my $node ( keys %subend ) {
769 my $witstr = $self->_path_display_label ( $self->path_witnesses( $node, $subend{$node} ) );
770 my $variables = { %edge_attrs, 'label' => $witstr };
771 my $varopts = _dot_attr_string( $variables );
772 $dot .= "\t\"$node\" -> \"__SUBEND__\" $varopts;\n";
775 if( $STRAIGHTENHACK ) {
776 my $endlabel = $endrank ? '__SUBEND__' : '__END__';
777 $dot .= "\t\"$endlabel\" -> \"#SILENT#\" [ color=white,penwidth=0 ];\n";
784 sub _dot_attr_string {
787 foreach my $k ( sort keys %$hash ) {
789 push( @attrs, $k.'="'.$v.'"' );
791 return( '[ ' . join( ', ', @attrs ) . ' ]' );
794 sub _add_edge_weights {
796 # Walk the graph from START to END, choosing the successor node with
797 # the largest number of witness paths each time.
799 my $curr = $self->start->id;
800 my $ranked = $self->end->has_rank;
801 while( $curr ne $self->end->id ) {
802 my $rank = $ranked ? $self->reading( $curr )->rank : 0;
803 my @succ = sort { $self->path_witnesses( $curr, $a )
804 <=> $self->path_witnesses( $curr, $b ) }
805 $self->sequence->successors( $curr );
806 my $next = pop @succ;
807 my $nextrank = $ranked ? $self->reading( $next )->rank : 0;
808 # Try to avoid lacunae in the weighted path.
810 ( $self->reading( $next )->is_lacuna ||
811 $nextrank - $rank > 1 ) ){
814 $weighted->{$curr} = $next;
820 =head2 path_witnesses( $edge )
822 Returns the list of sigils whose witnesses are associated with the given edge.
823 The edge can be passed as either an array or an arrayref of ( $source, $target ).
828 my( $self, @edge ) = @_;
829 # If edge is an arrayref, cope.
830 if( @edge == 1 && ref( $edge[0] ) eq 'ARRAY' ) {
834 my @wits = keys %{$self->sequence->get_edge_attributes( @edge )};
838 # Helper function. Make a display label for the given witnesses, showing a.c.
839 # witnesses only where the main witness is not also in the list.
840 sub _path_display_label {
843 map { $wits{$_} = 1 } @_;
845 # If an a.c. wit is listed, remove it if the main wit is also listed.
846 # Otherwise keep it for explicit listing.
847 my $aclabel = $self->ac_label;
849 foreach my $w ( sort keys %wits ) {
850 if( $w =~ /^(.*)\Q$aclabel\E$/ ) {
851 if( exists $wits{$1} ) {
854 push( @disp_ac, $w );
859 # See if we are in a majority situation.
860 my $maj = scalar( $self->tradition->witnesses ) * 0.6;
861 $maj = $maj > 5 ? $maj : 5;
862 if( scalar keys %wits > $maj ) {
863 unshift( @disp_ac, 'majority' );
864 return join( ', ', @disp_ac );
866 return join( ', ', sort keys %wits );
870 =head2 readings_at_rank( $rank )
872 Returns a list of readings at a given rank, taken from the alignment table.
876 sub readings_at_rank {
877 my( $self, $rank ) = @_;
878 my $table = $self->alignment_table;
879 # Table rank is real rank - 1.
880 my @elements = map { $_->{'tokens'}->[$rank-1] } @{$table->{'alignment'}};
882 foreach my $e ( @elements ) {
883 next unless ref( $e ) eq 'HASH';
884 next unless exists $e->{'t'};
885 $readings{$e->{'t'}->id} = $e->{'t'};
887 return values %readings;
892 Returns a GraphML representation of the collation. The GraphML will contain
893 two graphs. The first expresses the attributes of the readings and the witness
894 paths that link them; the second expresses the relationships that link the
895 readings. This is the native transfer format for a tradition.
905 my $datafile = 't/data/florilegium_tei_ps.xml';
906 my $tradition = Text::Tradition->new( 'input' => 'TEI',
911 ok( $tradition, "Got a tradition object" );
912 is( scalar $tradition->witnesses, 13, "Found all witnesses" );
913 ok( $tradition->collation, "Tradition has a collation" );
915 my $c = $tradition->collation;
916 is( scalar $c->readings, $READINGS, "Collation has all readings" );
917 is( scalar $c->paths, $PATHS, "Collation has all paths" );
918 is( scalar $c->relationships, 0, "Collation has all relationships" );
920 # Add a few relationships
921 $c->add_relationship( 'w123', 'w125', { 'type' => 'collated' } );
922 $c->add_relationship( 'w193', 'w196', { 'type' => 'collated' } );
923 $c->add_relationship( 'w257', 'w262', { 'type' => 'transposition' } );
925 # Now write it to GraphML and parse it again.
927 my $graphml = $c->as_graphml;
928 my $st = Text::Tradition->new( 'input' => 'Self', 'string' => $graphml );
929 is( scalar $st->collation->readings, $READINGS, "Reparsed collation has all readings" );
930 is( scalar $st->collation->paths, $PATHS, "Reparsed collation has all paths" );
931 is( scalar $st->collation->relationships, 3, "Reparsed collation has new relationships" );
933 # Now add a stemma, write to GraphML, and look at the output.
936 $tradition->enable_stemmata;
941 skip "Analysis module not present", 3 if $SKIP_STEMMA;
942 my $stemma = $tradition->add_stemma( 'dotfile' => 't/data/florilegium.dot' );
943 is( ref( $stemma ), 'Text::Tradition::Stemma', "Parsed dotfile into stemma" );
944 is( $tradition->stemmata, 1, "Tradition now has the stemma" );
945 $graphml = $c->as_graphml;
946 like( $graphml, qr/digraph/, "Digraph declaration exists in GraphML" );
949 # Now add a user, write to GraphML, and look at the output.
950 unlike( $graphml, qr/testuser/, "Test user name does not exist in GraphML yet" );
951 my $testuser = Text::Tradition::User->new(
952 id => 'testuser', password => 'testpass' );
953 is( ref( $testuser ), 'Text::Tradition::User', "Created test user object" );
954 $testuser->add_tradition( $tradition );
955 is( $tradition->user->id, $testuser->id, "Tradition assigned to test user" );
956 $graphml = $c->as_graphml;
957 like( $graphml, qr/testuser/, "Test user name now exists in GraphML" );
963 ## TODO MOVE this to Tradition.pm
965 my( $self, $options ) = @_;
966 $self->calculate_ranks unless $self->_graphcalc_done;
968 my $start = $options->{'from'}
969 ? $self->reading( $options->{'from'} ) : $self->start;
970 my $end = $options->{'to'}
971 ? $self->reading( $options->{'to'} ) : $self->end;
972 if( $start->has_rank && $end->has_rank && $end->rank < $start->rank ) {
973 throw( 'Start node must be before end node' );
975 # The readings need to be ranked for this to work.
976 $start = $self->start unless $start->has_rank;
977 $end = $self->end unless $end->has_rank;
979 unless( $start eq $self->start ) {
980 $rankoffset = $start->rank - 1;
985 my $graphml_ns = 'http://graphml.graphdrawing.org/xmlns';
986 my $xsi_ns = 'http://www.w3.org/2001/XMLSchema-instance';
987 my $graphml_schema = 'http://graphml.graphdrawing.org/xmlns ' .
988 'http://graphml.graphdrawing.org/xmlns/1.0/graphml.xsd';
990 # Create the document and root node
992 my $graphml = XML::LibXML->createDocument( "1.0", "UTF-8" );
993 my $root = $graphml->createElementNS( $graphml_ns, 'graphml' );
994 $graphml->setDocumentElement( $root );
995 $root->setNamespace( $xsi_ns, 'xsi', 0 );
996 $root->setAttributeNS( $xsi_ns, 'schemaLocation', $graphml_schema );
998 # List of attribute types to save on our objects and their corresponding
1003 'Bool' => 'boolean',
1004 'ReadingID' => 'string',
1005 'RelationshipType' => 'string',
1006 'RelationshipScope' => 'string',
1009 # Add the data keys for the graph. Include an extra key 'version' for the
1010 # GraphML output version.
1011 my %graph_data_keys;
1013 my %graph_attributes = ( 'version' => 'string' );
1014 # Graph attributes include those of Tradition and those of Collation.
1016 my $tmeta = $self->tradition->meta;
1017 my $cmeta = $self->meta;
1018 map { $gattr_from{$_->name} = 'Tradition' } $tmeta->get_all_attributes;
1019 map { $gattr_from{$_->name} = 'Collation' } $cmeta->get_all_attributes;
1020 foreach my $attr ( ( $tmeta->get_all_attributes, $cmeta->get_all_attributes ) ) {
1021 next if $attr->name =~ /^_/;
1022 next unless $save_types{$attr->type_constraint->name};
1023 $graph_attributes{$attr->name} = $save_types{$attr->type_constraint->name};
1025 # Extra custom keys for complex objects that should be saved in some form.
1026 # The subroutine should return a string, or undef/empty.
1027 if( $tmeta->has_method('stemmata') ) {
1028 $graph_attributes{'stemmata'} = sub {
1030 map { push( @stemstrs, $_->editable( {linesep => ''} ) ) }
1031 $self->tradition->stemmata;
1032 join( "\n", @stemstrs );
1036 $graph_attributes{'user'} = sub {
1037 $self->tradition->user ? $self->tradition->user->id : undef
1040 foreach my $datum ( sort keys %graph_attributes ) {
1041 $graph_data_keys{$datum} = 'dg'.$gdi++;
1042 my $key = $root->addNewChild( $graphml_ns, 'key' );
1043 my $dtype = ref( $graph_attributes{$datum} ) ? 'string'
1044 : $graph_attributes{$datum};
1045 $key->setAttribute( 'attr.name', $datum );
1046 $key->setAttribute( 'attr.type', $dtype );
1047 $key->setAttribute( 'for', 'graph' );
1048 $key->setAttribute( 'id', $graph_data_keys{$datum} );
1051 # Add the data keys for reading nodes
1052 my %reading_attributes;
1053 my $rmeta = Text::Tradition::Collation::Reading->meta;
1054 foreach my $attr( $rmeta->get_all_attributes ) {
1055 next if $attr->name =~ /^_/;
1056 next unless $save_types{$attr->type_constraint->name};
1057 $reading_attributes{$attr->name} = $save_types{$attr->type_constraint->name};
1059 # Extra custom key for the reading morphology
1060 $reading_attributes{'lexemes'} = 'string';
1064 foreach my $datum ( sort keys %reading_attributes ) {
1065 $node_data_keys{$datum} = 'dn'.$ndi++;
1066 my $key = $root->addNewChild( $graphml_ns, 'key' );
1067 $key->setAttribute( 'attr.name', $datum );
1068 $key->setAttribute( 'attr.type', $reading_attributes{$datum} );
1069 $key->setAttribute( 'for', 'node' );
1070 $key->setAttribute( 'id', $node_data_keys{$datum} );
1073 # Add the data keys for edges, that is, paths and relationships. Path
1074 # data does not come from a Moose class so is here manually.
1077 my %edge_attributes = (
1078 witness => 'string', # ID/label for a path
1079 extra => 'boolean', # Path key
1081 my @path_attributes = keys %edge_attributes; # track our manual additions
1082 my $pmeta = Text::Tradition::Collation::Relationship->meta;
1083 foreach my $attr( $pmeta->get_all_attributes ) {
1084 next if $attr->name =~ /^_/;
1085 next unless $save_types{$attr->type_constraint->name};
1086 $edge_attributes{$attr->name} = $save_types{$attr->type_constraint->name};
1088 foreach my $datum ( sort keys %edge_attributes ) {
1089 $edge_data_keys{$datum} = 'de'.$edi++;
1090 my $key = $root->addNewChild( $graphml_ns, 'key' );
1091 $key->setAttribute( 'attr.name', $datum );
1092 $key->setAttribute( 'attr.type', $edge_attributes{$datum} );
1093 $key->setAttribute( 'for', 'edge' );
1094 $key->setAttribute( 'id', $edge_data_keys{$datum} );
1097 # Add the collation graph itself. First, sanitize the name to a valid XML ID.
1098 my $xmlidname = $self->tradition->name;
1099 $xmlidname =~ s/(?!$xml10_namechar_rx)./_/g;
1100 if( $xmlidname !~ /^$xml10_namestartchar_rx/ ) {
1101 $xmlidname = '_'.$xmlidname;
1103 my $sgraph = $root->addNewChild( $graphml_ns, 'graph' );
1104 $sgraph->setAttribute( 'edgedefault', 'directed' );
1105 $sgraph->setAttribute( 'id', $xmlidname );
1106 $sgraph->setAttribute( 'parse.edgeids', 'canonical' );
1107 $sgraph->setAttribute( 'parse.edges', 0 ); # fill in later
1108 $sgraph->setAttribute( 'parse.nodeids', 'canonical' );
1109 $sgraph->setAttribute( 'parse.nodes', 0 ); # fill in later
1110 $sgraph->setAttribute( 'parse.order', 'nodesfirst' );
1112 # Tradition/collation attribute data
1113 foreach my $datum ( keys %graph_attributes ) {
1115 if( $datum eq 'version' ) {
1117 } elsif( ref( $graph_attributes{$datum} ) ) {
1118 my $sub = $graph_attributes{$datum};
1120 } elsif( $gattr_from{$datum} eq 'Tradition' ) {
1121 $value = $self->tradition->$datum;
1123 $value = $self->$datum;
1125 _add_graphml_data( $sgraph, $graph_data_keys{$datum}, $value );
1130 # Add our readings to the graph
1131 foreach my $n ( sort { $a->id cmp $b->id } $self->readings ) {
1132 next if $n->has_rank && $n ne $self->start && $n ne $self->end &&
1133 ( $n->rank < $start->rank || $n->rank > $end->rank );
1134 $use_readings{$n->id} = 1;
1135 # Add to the main graph
1136 my $node_el = $sgraph->addNewChild( $graphml_ns, 'node' );
1137 my $node_xmlid = 'n' . $node_ctr++;
1138 $node_hash{ $n->id } = $node_xmlid;
1139 $node_el->setAttribute( 'id', $node_xmlid );
1140 foreach my $d ( keys %reading_attributes ) {
1142 # Custom serialization
1143 if( $d eq 'lexemes' ) {
1144 # If nval is a true value, we have lexemes so we need to
1145 # serialize them. Otherwise set nval to undef so that the
1146 # key is excluded from this reading.
1147 $nval = $nval ? $n->_serialize_lexemes : undef;
1148 } elsif( $d eq 'normal_form' && $n->normal_form eq $n->text ) {
1151 if( $rankoffset && $d eq 'rank' && $n ne $self->start ) {
1152 # Adjust the ranks within the subgraph.
1153 $nval = $n eq $self->end ? $end->rank - $rankoffset + 1
1154 : $nval - $rankoffset;
1156 _add_graphml_data( $node_el, $node_data_keys{$d}, $nval )
1161 # Add the path edges to the sequence graph
1163 foreach my $e ( sort { $a->[0] cmp $b->[0] } $self->sequence->edges() ) {
1164 # We add an edge in the graphml for every witness in $e.
1165 next unless( $use_readings{$e->[0]} || $use_readings{$e->[1]} );
1166 my @edge_wits = sort $self->path_witnesses( $e );
1167 $e->[0] = $self->start->id unless $use_readings{$e->[0]};
1168 $e->[1] = $self->end->id unless $use_readings{$e->[1]};
1169 # Skip any path from start to end; that witness is not in the subgraph.
1170 next if ( $e->[0] eq $self->start->id && $e->[1] eq $self->end->id );
1171 foreach my $wit ( @edge_wits ) {
1172 my( $id, $from, $to ) = ( 'e'.$edge_ctr++,
1173 $node_hash{ $e->[0] },
1174 $node_hash{ $e->[1] } );
1175 my $edge_el = $sgraph->addNewChild( $graphml_ns, 'edge' );
1176 $edge_el->setAttribute( 'source', $from );
1177 $edge_el->setAttribute( 'target', $to );
1178 $edge_el->setAttribute( 'id', $id );
1180 # It's a witness path, so add the witness
1182 my $key = $edge_data_keys{'witness'};
1183 # Is this an ante-corr witness?
1184 my $aclabel = $self->ac_label;
1185 if( $wit =~ /^(.*)\Q$aclabel\E$/ ) {
1186 # Keep the base witness
1188 # ...and record that this is an 'extra' reading path
1189 _add_graphml_data( $edge_el, $edge_data_keys{'extra'}, $aclabel );
1191 _add_graphml_data( $edge_el, $edge_data_keys{'witness'}, $base );
1195 # Report the actual number of nodes and edges that went in
1196 $sgraph->setAttribute( 'parse.edges', $edge_ctr );
1197 $sgraph->setAttribute( 'parse.nodes', $node_ctr );
1199 # Add the relationship graph to the XML
1200 map { delete $edge_data_keys{$_} } @path_attributes;
1201 $self->relations->_as_graphml( $graphml_ns, $root, \%node_hash,
1202 $node_data_keys{'id'}, \%edge_data_keys );
1204 # Save and return the thing
1205 my $result = decode_utf8( $graphml->toString(1) );
1209 sub _add_graphml_data {
1210 my( $el, $key, $value ) = @_;
1211 return unless defined $value;
1212 my $data_el = $el->addNewChild( $el->namespaceURI, 'data' );
1213 $data_el->setAttribute( 'key', $key );
1214 $data_el->appendText( $value );
1219 Returns a CSV alignment table representation of the collation graph, one
1220 row per witness (or witness uncorrected.)
1226 my $table = $self->alignment_table;
1227 my $csv = Text::CSV->new( { binary => 1, quote_null => 0 } );
1229 # Make the header row
1230 $csv->combine( map { $_->{'witness'} } @{$table->{'alignment'}} );
1231 push( @result, decode_utf8( $csv->string ) );
1232 # Make the rest of the rows
1233 foreach my $idx ( 0 .. $table->{'length'} - 1 ) {
1234 my @rowobjs = map { $_->{'tokens'}->[$idx] } @{$table->{'alignment'}};
1235 my @row = map { $_ ? $_->{'t'}->text : $_ } @rowobjs;
1236 $csv->combine( @row );
1237 push( @result, decode_utf8( $csv->string ) );
1239 return join( "\n", @result );
1242 =head2 alignment_table( $use_refs, $include_witnesses )
1244 Return a reference to an alignment table, in a slightly enhanced CollateX
1245 format which looks like this:
1247 $table = { alignment => [ { witness => "SIGIL",
1248 tokens => [ { t => "TEXT" }, ... ] },
1249 { witness => "SIG2",
1250 tokens => [ { t => "TEXT" }, ... ] },
1252 length => TEXTLEN };
1254 If $use_refs is set to 1, the reading object is returned in the table
1255 instead of READINGTEXT; if not, the text of the reading is returned.
1257 If $include_witnesses is set to a hashref, only the witnesses whose sigil
1258 keys have a true hash value will be included.
1262 sub alignment_table {
1264 $self->calculate_ranks() unless $self->_graphcalc_done;
1265 return $self->cached_table if $self->has_cached_table;
1267 # Make sure we can do this
1268 throw( "Need a linear graph in order to make an alignment table" )
1269 unless $self->linear;
1270 $self->calculate_ranks unless $self->end->has_rank;
1272 my $table = { 'alignment' => [], 'length' => $self->end->rank - 1 };
1273 my @all_pos = ( 1 .. $self->end->rank - 1 );
1274 foreach my $wit ( sort { $a->sigil cmp $b->sigil } $self->tradition->witnesses ) {
1275 # say STDERR "Making witness row(s) for " . $wit->sigil;
1276 my @wit_path = $self->reading_sequence( $self->start, $self->end, $wit->sigil );
1277 my @row = _make_witness_row( \@wit_path, \@all_pos );
1278 push( @{$table->{'alignment'}},
1279 { 'witness' => $wit->sigil, 'tokens' => \@row } );
1280 if( $wit->is_layered ) {
1281 my @wit_ac_path = $self->reading_sequence( $self->start, $self->end,
1282 $wit->sigil.$self->ac_label );
1283 my @ac_row = _make_witness_row( \@wit_ac_path, \@all_pos );
1284 push( @{$table->{'alignment'}},
1285 { 'witness' => $wit->sigil.$self->ac_label, 'tokens' => \@ac_row } );
1288 $self->cached_table( $table );
1292 sub _make_witness_row {
1293 my( $path, $positions ) = @_;
1295 map { $char_hash{$_} = undef } @$positions;
1297 foreach my $rdg ( @$path ) {
1298 say STDERR "rank " . $rdg->rank if $debug;
1299 # say STDERR "No rank for " . $rdg->id unless defined $rdg->rank;
1300 $char_hash{$rdg->rank} = { 't' => $rdg };
1302 my @row = map { $char_hash{$_} } @$positions;
1303 # Fill in lacuna markers for undef spots in the row
1304 my $last_el = shift @row;
1305 my @filled_row = ( $last_el );
1306 foreach my $el ( @row ) {
1307 # If we are using node reference, make the lacuna node appear many times
1308 # in the table. If not, use the lacuna tag.
1309 if( $last_el && $last_el->{'t'}->is_lacuna && !defined $el ) {
1312 push( @filled_row, $el );
1318 =head1 NAVIGATION METHODS
1320 =head2 reading_sequence( $first, $last, $sigil, $backup )
1322 Returns the ordered list of readings, starting with $first and ending
1323 with $last, for the witness given in $sigil. If a $backup sigil is
1324 specified (e.g. when walking a layered witness), it will be used wherever
1325 no $sigil path exists. If there is a base text reading, that will be
1326 used wherever no path exists for $sigil or $backup.
1330 # TODO Think about returning some lazy-eval iterator.
1331 # TODO Get rid of backup; we should know from what witness is whether we need it.
1333 sub reading_sequence {
1334 my( $self, $start, $end, $witness ) = @_;
1336 $witness = $self->baselabel unless $witness;
1337 my @readings = ( $start );
1340 while( $n && $n->id ne $end->id ) {
1341 if( exists( $seen{$n->id} ) ) {
1342 throw( "Detected loop for $witness at " . $n->id );
1346 my $next = $self->next_reading( $n, $witness );
1348 throw( "Did not find any path for $witness from reading " . $n->id );
1350 push( @readings, $next );
1353 # Check that the last reading is our end reading.
1354 my $last = $readings[$#readings];
1355 throw( "Last reading found from " . $start->text .
1356 " for witness $witness is not the end!" ) # TODO do we get this far?
1357 unless $last->id eq $end->id;
1362 =head2 next_reading( $reading, $sigil );
1364 Returns the reading that follows the given reading along the given witness
1370 # Return the successor via the corresponding path.
1372 my $answer = $self->_find_linked_reading( 'next', @_ );
1373 return undef unless $answer;
1374 return $self->reading( $answer );
1377 =head2 prior_reading( $reading, $sigil )
1379 Returns the reading that precedes the given reading along the given witness
1385 # Return the predecessor via the corresponding path.
1387 my $answer = $self->_find_linked_reading( 'prior', @_ );
1388 return $self->reading( $answer );
1391 sub _find_linked_reading {
1392 my( $self, $direction, $node, $path ) = @_;
1394 # Get a backup if we are dealing with a layered witness
1396 my $aclabel = $self->ac_label;
1397 if( $path && $path =~ /^(.*)\Q$aclabel\E$/ ) {
1401 my @linked_paths = $direction eq 'next'
1402 ? $self->sequence->edges_from( $node )
1403 : $self->sequence->edges_to( $node );
1404 return undef unless scalar( @linked_paths );
1406 # We have to find the linked path that contains all of the
1407 # witnesses supplied in $path.
1408 my( @path_wits, @alt_path_wits );
1409 @path_wits = sort( $self->_witnesses_of_label( $path ) ) if $path;
1410 @alt_path_wits = sort( $self->_witnesses_of_label( $alt_path ) ) if $alt_path;
1413 foreach my $le ( @linked_paths ) {
1414 if( $self->sequence->has_edge_attribute( @$le, $self->baselabel ) ) {
1417 my @le_wits = sort $self->path_witnesses( $le );
1418 if( _is_within( \@path_wits, \@le_wits ) ) {
1419 # This is the right path.
1420 return $direction eq 'next' ? $le->[1] : $le->[0];
1421 } elsif( _is_within( \@alt_path_wits, \@le_wits ) ) {
1425 # Got this far? Return the alternate path if it exists.
1426 return $direction eq 'next' ? $alt_le->[1] : $alt_le->[0]
1429 # Got this far? Return the base path if it exists.
1430 return $direction eq 'next' ? $base_le->[1] : $base_le->[0]
1433 # Got this far? We have no appropriate path.
1434 warn "Could not find $direction node from " . $node->id
1435 . " along path $path";
1441 my( $set1, $set2 ) = @_;
1442 my $ret = @$set1; # will be 0, i.e. false, if set1 is empty
1443 foreach my $el ( @$set1 ) {
1444 $ret = 0 unless grep { /^\Q$el\E$/ } @$set2;
1449 # Return the string that joins together a list of witnesses for
1450 # display on a single path.
1451 sub _witnesses_of_label {
1452 my( $self, $label ) = @_;
1453 my $regex = $self->wit_list_separator;
1454 my @answer = split( /\Q$regex\E/, $label );
1458 =head2 common_readings
1460 Returns the list of common readings in the graph (i.e. those readings that are
1461 shared by all non-lacunose witnesses.)
1465 sub common_readings {
1467 my @common = grep { $_->is_common } $self->readings;
1471 =head2 path_text( $sigil, [, $start, $end ] )
1473 Returns the text of a witness (plus its backup, if we are using a layer)
1474 as stored in the collation. The text is returned as a string, where the
1475 individual readings are joined with spaces and the meta-readings (e.g.
1476 lacunae) are omitted. Optional specification of $start and $end allows
1477 the generation of a subset of the witness text.
1482 my( $self, $wit, $start, $end ) = @_;
1483 $start = $self->start unless $start;
1484 $end = $self->end unless $end;
1485 my @path = grep { !$_->is_meta } $self->reading_sequence( $start, $end, $wit );
1488 foreach my $r ( @path ) {
1489 unless ( $r->join_prior || !$last || $last->join_next ) {
1492 $pathtext .= $r->text;
1498 =head1 INITIALIZATION METHODS
1500 These are mostly for use by parsers.
1502 =head2 make_witness_path( $witness )
1504 Link the array of readings contained in $witness->path (and in
1505 $witness->uncorrected_path if it exists) into collation paths.
1506 Clear out the arrays when finished.
1508 =head2 make_witness_paths
1510 Call make_witness_path for all witnesses in the tradition.
1514 # For use when a collation is constructed from a base text and an apparatus.
1515 # We have the sequences of readings and just need to add path edges.
1516 # When we are done, clear out the witness path attributes, as they are no
1518 # TODO Find a way to replace the witness path attributes with encapsulated functions?
1520 sub make_witness_paths {
1522 foreach my $wit ( $self->tradition->witnesses ) {
1523 # say STDERR "Making path for " . $wit->sigil;
1524 $self->make_witness_path( $wit );
1528 sub make_witness_path {
1529 my( $self, $wit ) = @_;
1530 my @chain = @{$wit->path};
1531 my $sig = $wit->sigil;
1532 # Add start and end if necessary
1533 unshift( @chain, $self->start ) unless $chain[0] eq $self->start;
1534 push( @chain, $self->end ) unless $chain[-1] eq $self->end;
1535 foreach my $idx ( 0 .. $#chain-1 ) {
1536 $self->add_path( $chain[$idx], $chain[$idx+1], $sig );
1538 if( $wit->is_layered ) {
1539 @chain = @{$wit->uncorrected_path};
1540 unshift( @chain, $self->start ) unless $chain[0] eq $self->start;
1541 push( @chain, $self->end ) unless $chain[-1] eq $self->end;
1542 foreach my $idx( 0 .. $#chain-1 ) {
1543 my $source = $chain[$idx];
1544 my $target = $chain[$idx+1];
1545 $self->add_path( $source, $target, $sig.$self->ac_label )
1546 unless $self->has_path( $source, $target, $sig );
1550 $wit->clear_uncorrected_path;
1553 =head2 calculate_ranks
1555 Calculate the reading ranks (that is, their aligned positions relative
1556 to each other) for the graph. This can only be called on linear collations.
1560 use Text::Tradition;
1562 my $cxfile = 't/data/Collatex-16.xml';
1563 my $t = Text::Tradition->new(
1565 'input' => 'CollateX',
1568 my $c = $t->collation;
1571 my $table = $c->alignment_table;
1572 ok( $c->has_cached_table, "Alignment table was cached" );
1573 is( $c->alignment_table, $table, "Cached table returned upon second call" );
1574 $c->calculate_ranks;
1575 is( $c->alignment_table, $table, "Cached table retained with no rank change" );
1576 $c->add_relationship( 'n24', 'n23', { 'type' => 'spelling' } );
1577 isnt( $c->alignment_table, $table, "Alignment table changed after relationship add" );
1583 sub calculate_ranks {
1585 # Save the existing ranks, in case we need to invalidate the cached SVG.
1587 map { $existing_ranks{$_} = $_->rank } $self->readings;
1589 # Do the rankings based on the relationship equivalence graph, starting
1590 # with the start node.
1591 my ( $node_ranks, $rank_nodes ) = $self->relations->equivalence_ranks();
1593 # Transfer our rankings from the topological graph to the real one.
1594 foreach my $r ( $self->readings ) {
1595 if( defined $node_ranks->{$self->equivalence( $r->id )} ) {
1596 $r->rank( $node_ranks->{$self->equivalence( $r->id )} );
1598 # Die. Find the last rank we calculated.
1599 my @all_defined = sort { ( $node_ranks->{$self->equivalence( $a->id )}||-1 )
1600 <=> ( $node_ranks->{$self->equivalence( $b->id )}||-1 ) }
1602 my $last = pop @all_defined;
1603 throw( "Ranks not calculated after $last - do you have a cycle in the graph?" );
1606 # Do we need to invalidate the cached data?
1607 if( $self->has_cached_table ) {
1608 foreach my $r ( $self->readings ) {
1609 next if defined( $existing_ranks{$r} )
1610 && $existing_ranks{$r} == $r->rank;
1611 # Something has changed, so clear the cache
1612 $self->_clear_cache;
1613 # ...and recalculate the common readings.
1614 $self->calculate_common_readings();
1618 # The graph calculation information is now up to date.
1619 $self->_graphcalc_done(1);
1624 $self->wipe_table if $self->has_cached_table;
1628 =head2 flatten_ranks
1630 A convenience method for parsing collation data. Searches the graph for readings
1631 with the same text at the same rank, and merges any that are found.
1637 my %unique_rank_rdg;
1639 foreach my $rdg ( $self->readings ) {
1640 next unless $rdg->has_rank;
1641 my $key = $rdg->rank . "||" . $rdg->text;
1642 if( exists $unique_rank_rdg{$key} ) {
1643 # Make sure they don't have different grammatical forms
1644 my $ur = $unique_rank_rdg{$key};
1645 if( $rdg->disambiguated && $ur->disambiguated ) {
1646 my $rform = join( '//', map { $_->form->to_string } $rdg->lexemes );
1647 my $uform = join( '//', map { $_->form->to_string } $ur->lexemes );
1648 next unless $rform eq $uform;
1649 } elsif( $rdg->disambiguated xor $ur->disambiguated ) {
1653 #say STDERR "Combining readings at same rank: $key";
1655 $self->merge_readings( $unique_rank_rdg{$key}, $rdg );
1656 # TODO see if this now makes a common point.
1658 $unique_rank_rdg{$key} = $rdg;
1661 # If we merged readings, the ranks are still fine but the alignment
1662 # table is wrong. Wipe it.
1663 $self->wipe_table() if $changed;
1667 =head2 calculate_common_readings
1669 Goes through the graph identifying the readings that appear in every witness
1670 (apart from those with lacunae at that spot.) Marks them as common and returns
1675 use Text::Tradition;
1677 my $cxfile = 't/data/Collatex-16.xml';
1678 my $t = Text::Tradition->new(
1680 'input' => 'CollateX',
1683 my $c = $t->collation;
1685 my @common = $c->calculate_common_readings();
1686 is( scalar @common, 8, "Found correct number of common readings" );
1687 my @marked = sort $c->common_readings();
1688 is( scalar @common, 8, "All common readings got marked as such" );
1689 my @expected = qw/ n1 n11 n16 n19 n20 n5 n6 n7 /;
1690 is_deeply( \@marked, \@expected, "Found correct list of common readings" );
1696 sub calculate_common_readings {
1699 map { $_->is_common( 0 ) } $self->readings;
1700 # Implicitly calls calculate_ranks
1701 my $table = $self->alignment_table;
1702 foreach my $idx ( 0 .. $table->{'length'} - 1 ) {
1703 my @row = map { $_->{'tokens'}->[$idx]
1704 ? $_->{'tokens'}->[$idx]->{'t'} : '' }
1705 @{$table->{'alignment'}};
1707 foreach my $r ( @row ) {
1709 $hash{$r->id} = $r unless $r->is_meta;
1711 $hash{'UNDEF'} = $r;
1714 if( keys %hash == 1 && !exists $hash{'UNDEF'} ) {
1715 my( $r ) = values %hash;
1717 push( @common, $r );
1723 =head2 text_from_paths
1725 Calculate the text array for all witnesses from the path, for later consistency
1726 checking. Only to be used if there is no non-graph-based way to know the
1731 sub text_from_paths {
1733 foreach my $wit ( $self->tradition->witnesses ) {
1734 my @readings = $self->reading_sequence( $self->start, $self->end, $wit->sigil );
1736 foreach my $r ( @readings ) {
1737 next if $r->is_meta;
1738 push( @text, $r->text );
1740 $wit->text( \@text );
1741 if( $wit->is_layered ) {
1742 my @ucrdgs = $self->reading_sequence( $self->start, $self->end,
1743 $wit->sigil.$self->ac_label );
1745 foreach my $r ( @ucrdgs ) {
1746 next if $r->is_meta;
1747 push( @uctext, $r->text );
1749 $wit->layertext( \@uctext );
1754 =head1 UTILITY FUNCTIONS
1756 =head2 common_predecessor( $reading_a, $reading_b )
1758 Find the last reading that occurs in sequence before both the given readings.
1759 At the very least this should be $self->start.
1761 =head2 common_successor( $reading_a, $reading_b )
1763 Find the first reading that occurs in sequence after both the given readings.
1764 At the very least this should be $self->end.
1768 use Text::Tradition;
1770 my $cxfile = 't/data/Collatex-16.xml';
1771 my $t = Text::Tradition->new(
1773 'input' => 'CollateX',
1776 my $c = $t->collation;
1778 is( $c->common_predecessor( 'n24', 'n23' )->id,
1779 'n20', "Found correct common predecessor" );
1780 is( $c->common_successor( 'n24', 'n23' )->id,
1781 '__END__', "Found correct common successor" );
1783 is( $c->common_predecessor( 'n19', 'n17' )->id,
1784 'n16', "Found correct common predecessor for readings on same path" );
1785 is( $c->common_successor( 'n21', 'n10' )->id,
1786 '__END__', "Found correct common successor for readings on same path" );
1792 ## Return the closest reading that is a predecessor of both the given readings.
1793 sub common_predecessor {
1795 my( $r1, $r2 ) = $self->_objectify_args( @_ );
1796 return $self->_common_in_path( $r1, $r2, 'predecessors' );
1799 sub common_successor {
1801 my( $r1, $r2 ) = $self->_objectify_args( @_ );
1802 return $self->_common_in_path( $r1, $r2, 'successors' );
1806 # TODO think about how to do this without ranks...
1807 sub _common_in_path {
1808 my( $self, $r1, $r2, $dir ) = @_;
1809 my $iter = $self->end->rank;
1811 my @last_r1 = ( $r1 );
1812 my @last_r2 = ( $r2 );
1813 # my %all_seen = ( $r1 => 'r1', $r2 => 'r2' );
1815 # say STDERR "Finding common $dir for $r1, $r2";
1816 while( !@candidates ) {
1817 last unless $iter--; # Avoid looping infinitely
1818 # Iterate separately down the graph from r1 and r2
1819 my( @new_lc1, @new_lc2 );
1820 foreach my $lc ( @last_r1 ) {
1821 foreach my $p ( $lc->$dir ) {
1822 if( $all_seen{$p->id} && $all_seen{$p->id} ne 'r1' ) {
1823 # say STDERR "Path candidate $p from $lc";
1824 push( @candidates, $p );
1825 } elsif( !$all_seen{$p->id} ) {
1826 $all_seen{$p->id} = 'r1';
1827 push( @new_lc1, $p );
1831 foreach my $lc ( @last_r2 ) {
1832 foreach my $p ( $lc->$dir ) {
1833 if( $all_seen{$p->id} && $all_seen{$p->id} ne 'r2' ) {
1834 # say STDERR "Path candidate $p from $lc";
1835 push( @candidates, $p );
1836 } elsif( !$all_seen{$p->id} ) {
1837 $all_seen{$p->id} = 'r2';
1838 push( @new_lc2, $p );
1842 @last_r1 = @new_lc1;
1843 @last_r2 = @new_lc2;
1845 my @answer = sort { $a->rank <=> $b->rank } @candidates;
1846 return $dir eq 'predecessors' ? pop( @answer ) : shift ( @answer );
1850 Text::Tradition::Error->throw(
1851 'ident' => 'Collation error',
1857 __PACKAGE__->meta->make_immutable;
1861 This package is free software and is provided "as is" without express
1862 or implied warranty. You can redistribute it and/or modify it under
1863 the same terms as Perl itself.
1867 Tara L Andrews E<lt>aurum@cpan.orgE<gt>