1 package Text::Tradition::Collation;
3 use Encode qw( decode_utf8 );
7 use IPC::Run qw( run binary );
9 use Text::Tradition::Collation::Reading;
10 use Text::Tradition::Collation::RelationshipStore;
11 use Text::Tradition::Error;
13 use XML::LibXML::XPathContext;
19 default => sub { Graph->new() },
27 isa => 'Text::Tradition::Collation::RelationshipStore',
29 relationships => 'relationships',
30 related_readings => 'related_readings',
31 del_relationship => 'del_relationship',
33 writer => '_set_relations',
38 isa => 'Text::Tradition',
43 isa => 'HashRef[Text::Tradition::Collation::Reading]',
47 _add_reading => 'set',
48 del_reading => 'delete',
49 has_reading => 'exists',
52 default => sub { {} },
55 has 'wit_list_separator' => (
64 default => 'base text',
81 isa => 'Text::Tradition::Collation::Reading',
82 writer => '_set_start',
88 isa => 'Text::Tradition::Collation::Reading',
96 predicate => 'has_cached_svg',
97 clearer => 'wipe_svg',
100 has 'cached_table' => (
103 predicate => 'has_cached_table',
104 clearer => 'wipe_table',
107 has '_graphcalc_done' => (
115 Text::Tradition::Collation - a software model for a text collation
120 my $t = Text::Tradition->new(
121 'name' => 'this is a text',
123 'file' => '/path/to/tei_parallel_seg_file.xml' );
125 my $c = $t->collation;
126 my @readings = $c->readings;
127 my @paths = $c->paths;
128 my @relationships = $c->relationships;
130 my $svg_variant_graph = $t->collation->as_svg();
134 Text::Tradition is a library for representation and analysis of collated
135 texts, particularly medieval ones. The Collation is the central feature of
136 a Tradition, where the text, its sequence of readings, and its relationships
137 between readings are actually kept.
143 The constructor. Takes a hash or hashref of the following arguments:
147 =item * tradition - The Text::Tradition object to which the collation
150 =item * linear - Whether the collation should be linear; that is, whether
151 transposed readings should be treated as two linked readings rather than one,
152 and therefore whether the collation graph is acyclic. Defaults to true.
154 =item * baselabel - The default label for the path taken by a base text
155 (if any). Defaults to 'base text'.
157 =item * wit_list_separator - The string to join a list of witnesses for
158 purposes of making labels in display graphs. Defaults to ', '.
160 =item * ac_label - The extra label to tack onto a witness sigil when
161 representing another layer of path for the given witness - that is, when
162 a text has more than one possible reading due to scribal corrections or
163 the like. Defaults to ' (a.c.)'.
173 =head2 wit_list_separator
179 Simple accessors for collation attributes.
183 The meta-reading at the start of every witness path.
187 The meta-reading at the end of every witness path.
191 Returns all Reading objects in the graph.
193 =head2 reading( $id )
195 Returns the Reading object corresponding to the given ID.
197 =head2 add_reading( $reading_args )
199 Adds a new reading object to the collation.
200 See L<Text::Tradition::Collation::Reading> for the available arguments.
202 =head2 del_reading( $object_or_id )
204 Removes the given reading from the collation, implicitly removing its
205 paths and relationships.
207 =head2 merge_readings( $main, $second )
209 Merges the $second reading into the $main one.
210 The arguments may be either readings or reading IDs.
212 =head2 has_reading( $id )
214 Predicate to see whether a given reading ID is in the graph.
216 =head2 reading_witnesses( $object_or_id )
218 Returns a list of sigils whose witnesses contain the reading.
222 Returns all reading paths within the document - that is, all edges in the
223 collation graph. Each path is an arrayref of [ $source, $target ] reading IDs.
225 =head2 add_path( $source, $target, $sigil )
227 Links the given readings in the collation in sequence, under the given witness
228 sigil. The readings may be specified by object or ID.
230 =head2 del_path( $source, $target, $sigil )
232 Links the given readings in the collation in sequence, under the given witness
233 sigil. The readings may be specified by object or ID.
235 =head2 has_path( $source, $target );
237 Returns true if the two readings are linked in sequence in any witness.
238 The readings may be specified by object or ID.
242 Returns all Relationship objects in the collation.
244 =head2 add_relationship( $reading, $other_reading, $options )
246 Adds a new relationship of the type given in $options between the two readings,
247 which may be specified by object or ID. Returns a value of ( $status, @vectors)
248 where $status is true on success, and @vectors is a list of relationship edges
249 that were ultimately added.
250 See L<Text::Tradition::Collation::Relationship> for the available options.
256 $self->_set_relations( Text::Tradition::Collation::RelationshipStore->new( 'collation' => $self ) );
257 $self->_set_start( $self->add_reading( { 'collation' => $self, 'is_start' => 1 } ) );
258 $self->_set_end( $self->add_reading( { 'collation' => $self, 'is_end' => 1 } ) );
261 ### Reading construct/destruct functions
264 my( $self, $reading ) = @_;
265 unless( ref( $reading ) eq 'Text::Tradition::Collation::Reading' ) {
266 my %args = %$reading;
267 $reading = Text::Tradition::Collation::Reading->new(
268 'collation' => $self,
271 # First check to see if a reading with this ID exists.
272 if( $self->reading( $reading->id ) ) {
273 throw( "Collation already has a reading with id " . $reading->id );
275 $self->_graphcalc_done(0);
276 $self->_add_reading( $reading->id => $reading );
277 # Once the reading has been added, put it in both graphs.
278 $self->sequence->add_vertex( $reading->id );
279 $self->relations->add_reading( $reading->id );
283 around del_reading => sub {
288 if( ref( $arg ) eq 'Text::Tradition::Collation::Reading' ) {
291 # Remove the reading from the graphs.
292 $self->_graphcalc_done(0);
293 $self->sequence->delete_vertex( $arg );
294 $self->relations->delete_reading( $arg );
297 $self->$orig( $arg );
300 # merge_readings( $main, $to_be_deleted );
305 # We only need the IDs for adding paths to the graph, not the reading
306 # objects themselves.
307 my( $kept, $deleted, $combine_char ) = $self->_stringify_args( @_ );
308 $self->_graphcalc_done(0);
310 # The kept reading should inherit the paths and the relationships
311 # of the deleted reading.
312 foreach my $path ( $self->sequence->edges_at( $deleted ) ) {
313 my @vector = ( $kept );
314 push( @vector, $path->[1] ) if $path->[0] eq $deleted;
315 unshift( @vector, $path->[0] ) if $path->[1] eq $deleted;
316 next if $vector[0] eq $vector[1]; # Don't add a self loop
317 my %wits = %{$self->sequence->get_edge_attributes( @$path )};
318 $self->sequence->add_edge( @vector );
319 my $fwits = $self->sequence->get_edge_attributes( @vector );
320 @wits{keys %$fwits} = values %$fwits;
321 $self->sequence->set_edge_attributes( @vector, \%wits );
323 $self->relations->merge_readings( $kept, $deleted, $combine_char );
325 # Do the deletion deed.
326 if( $combine_char ) {
327 my $kept_obj = $self->reading( $kept );
328 my $new_text = join( $combine_char, $kept_obj->text,
329 $self->reading( $deleted )->text );
330 $kept_obj->alter_text( $new_text );
332 $self->del_reading( $deleted );
336 # Helper function for manipulating the graph.
337 sub _stringify_args {
338 my( $self, $first, $second, $arg ) = @_;
340 if ref( $first ) eq 'Text::Tradition::Collation::Reading';
341 $second = $second->id
342 if ref( $second ) eq 'Text::Tradition::Collation::Reading';
343 return( $first, $second, $arg );
346 # Helper function for manipulating the graph.
347 sub _objectify_args {
348 my( $self, $first, $second, $arg ) = @_;
349 $first = $self->reading( $first )
350 unless ref( $first ) eq 'Text::Tradition::Collation::Reading';
351 $second = $self->reading( $second )
352 unless ref( $second ) eq 'Text::Tradition::Collation::Reading';
353 return( $first, $second, $arg );
360 # We only need the IDs for adding paths to the graph, not the reading
361 # objects themselves.
362 my( $source, $target, $wit ) = $self->_stringify_args( @_ );
364 $self->_graphcalc_done(0);
365 # Connect the readings
366 $self->sequence->add_edge( $source, $target );
367 # Note the witness in question
368 $self->sequence->set_edge_attribute( $source, $target, $wit, 1 );
374 if( ref( $_[0] ) eq 'ARRAY' ) {
381 # We only need the IDs for adding paths to the graph, not the reading
382 # objects themselves.
383 my( $source, $target, $wit ) = $self->_stringify_args( @args );
385 $self->_graphcalc_done(0);
386 if( $self->sequence->has_edge_attribute( $source, $target, $wit ) ) {
387 $self->sequence->delete_edge_attribute( $source, $target, $wit );
389 unless( keys %{$self->sequence->get_edge_attributes( $source, $target )} ) {
390 $self->sequence->delete_edge( $source, $target );
395 # Extra graph-alike utility
398 my( $source, $target, $wit ) = $self->_stringify_args( @_ );
399 return undef unless $self->sequence->has_edge( $source, $target );
400 return $self->sequence->has_edge_attribute( $source, $target, $wit );
403 =head2 clear_witness( @sigil_list )
405 Clear the given witnesses out of the collation entirely, removing references
406 to them in paths, and removing readings that belong only to them. Should only
407 be called via $tradition->del_witness.
412 my( $self, @sigils ) = @_;
414 $self->_graphcalc_done(0);
415 # Clear the witness(es) out of the paths
416 foreach my $e ( $self->paths ) {
417 foreach my $sig ( @sigils ) {
418 $self->del_path( $e, $sig );
422 # Clear out the newly unused readings
423 foreach my $r ( $self->readings ) {
424 unless( $self->reading_witnesses( $r ) ) {
425 $self->del_reading( $r );
430 sub add_relationship {
432 my( $source, $target, $opts ) = $self->_stringify_args( @_ );
433 my( @vectors ) = $self->relations->add_relationship( $source,
434 $self->reading( $source ), $target, $self->reading( $target ), $opts );
435 $self->_graphcalc_done(0);
439 =head2 reading_witnesses( $reading )
441 Return a list of sigils corresponding to the witnesses in which the reading appears.
445 sub reading_witnesses {
446 my( $self, $reading ) = @_;
447 # We need only check either the incoming or the outgoing edges; I have
448 # arbitrarily chosen "incoming". Thus, special-case the start node.
449 if( $reading eq $self->start ) {
450 return map { $_->sigil } $self->tradition->witnesses;
453 foreach my $e ( $self->sequence->edges_to( $reading ) ) {
454 my $wits = $self->sequence->get_edge_attributes( @$e );
455 @all_witnesses{ keys %$wits } = 1;
457 my $acstr = $self->ac_label;
458 foreach my $acwit ( grep { $_ =~ s/^(.*)\Q$acstr\E$/$1/ } keys %all_witnesses ) {
459 delete $all_witnesses{$acwit.$acstr} if exists $all_witnesses{$acwit};
461 return keys %all_witnesses;
464 =head1 OUTPUT METHODS
466 =head2 as_svg( \%options )
468 Returns an SVG string that represents the graph, via as_dot and graphviz.
469 See as_dot for a list of options. Must have GraphViz (dot) installed to run.
474 my( $self, $opts ) = @_;
475 throw( "Need GraphViz installed to output SVG" )
476 unless File::Which::which( 'dot' );
477 my $want_subgraph = exists $opts->{'from'} || exists $opts->{'to'};
478 $self->calculate_ranks() unless $self->_graphcalc_done;
479 if( !$self->has_cached_svg || $opts->{'recalc'} || $want_subgraph ) {
480 my @cmd = qw/dot -Tsvg/;
482 my $dotfile = File::Temp->new();
484 # $dotfile->unlink_on_destroy(0);
485 binmode $dotfile, ':utf8';
486 print $dotfile $self->as_dot( $opts );
487 push( @cmd, $dotfile->filename );
488 run( \@cmd, ">", binary(), \$svg );
489 $svg = decode_utf8( $svg );
490 $self->cached_svg( $svg ) unless $want_subgraph;
493 return $self->cached_svg;
498 =head2 as_dot( \%options )
500 Returns a string that is the collation graph expressed in dot
501 (i.e. GraphViz) format. Options include:
516 my( $self, $opts ) = @_;
517 my $startrank = $opts->{'from'} if $opts;
518 my $endrank = $opts->{'to'} if $opts;
519 my $color_common = $opts->{'color_common'} if $opts;
520 my $STRAIGHTENHACK = !$startrank && !$endrank && $self->end->rank
521 && $self->end->rank > 100;
523 # Check the arguments
525 return if $endrank && $startrank > $endrank;
526 return if $startrank > $self->end->rank;
528 if( defined $endrank ) {
529 return if $endrank < 0;
530 $endrank = undef if $endrank == $self->end->rank;
533 my $graph_name = $self->tradition->name;
534 $graph_name =~ s/[^\w\s]//g;
535 $graph_name = join( '_', split( /\s+/, $graph_name ) );
543 'fillcolor' => 'white',
548 'arrowhead' => 'open',
549 'color' => '#000000',
550 'fontcolor' => '#000000',
553 my $dot = sprintf( "digraph %s {\n", $graph_name );
554 $dot .= "\tgraph " . _dot_attr_string( \%graph_attrs ) . ";\n";
555 $dot .= "\tnode " . _dot_attr_string( \%node_attrs ) . ";\n";
557 # Output substitute start/end readings if necessary
559 $dot .= "\t\"#SUBSTART#\" [ label=\"...\" ];\n";
562 $dot .= "\t\"#SUBEND#\" [ label=\"...\" ];\n";
564 if( $STRAIGHTENHACK ) {
566 $dot .= "\tsubgraph { rank=same \"#START#\" \"#SILENT#\" }\n";
567 $dot .= "\t\"#SILENT#\" [ shape=diamond,color=white,penwidth=0,label=\"\" ];"
569 my %used; # Keep track of the readings that actually appear in the graph
570 # Sort the readings by rank if we have ranks; this speeds layout.
571 my @all_readings = $self->end->has_rank
572 ? sort { $a->rank <=> $b->rank } $self->readings
574 # TODO Refrain from outputting lacuna nodes - just grey out the edges.
575 foreach my $reading ( @all_readings ) {
576 # Only output readings within our rank range.
577 next if $startrank && $reading->rank < $startrank;
578 next if $endrank && $reading->rank > $endrank;
579 $used{$reading->id} = 1;
580 # Need not output nodes without separate labels
581 next if $reading->id eq $reading->text;
583 my $label = $reading->text;
584 $label =~ s/\"/\\\"/g;
585 $rattrs->{'label'} = $label;
586 $rattrs->{'fillcolor'} = '#b3f36d' if $reading->is_common && $color_common;
587 $dot .= sprintf( "\t\"%s\" %s;\n", $reading->id, _dot_attr_string( $rattrs ) );
590 # Add the real edges. Need to weight one edge per rank jump, in a
592 # my $weighted = $self->_add_edge_weights;
593 my @edges = $self->paths;
594 my( %substart, %subend );
595 foreach my $edge ( @edges ) {
596 # Do we need to output this edge?
597 if( $used{$edge->[0]} && $used{$edge->[1]} ) {
598 my $label = $self->_path_display_label( $self->path_witnesses( $edge ) );
599 my $variables = { %edge_attrs, 'label' => $label };
601 # Account for the rank gap if necessary
602 my $rank0 = $self->reading( $edge->[0] )->rank
603 if $self->reading( $edge->[0] )->has_rank;
604 my $rank1 = $self->reading( $edge->[1] )->rank
605 if $self->reading( $edge->[1] )->has_rank;
606 if( defined $rank0 && defined $rank1 && $rank1 - $rank0 > 1 ) {
607 $variables->{'minlen'} = $rank1 - $rank0;
610 # Add the calculated edge weights
611 # if( exists $weighted->{$edge->[0]}
612 # && $weighted->{$edge->[0]} eq $edge->[1] ) {
613 # # $variables->{'color'} = 'red';
614 # $variables->{'weight'} = 3.0;
617 # EXPERIMENTAL: make edge width reflect no. of witnesses
618 my $extrawidth = scalar( $self->path_witnesses( $edge ) ) * 0.2;
619 $variables->{'penwidth'} = $extrawidth + 0.8; # gives 1 for a single wit
621 my $varopts = _dot_attr_string( $variables );
622 $dot .= sprintf( "\t\"%s\" -> \"%s\" %s;\n",
623 $edge->[0], $edge->[1], $varopts );
624 } elsif( $used{$edge->[0]} ) {
625 $subend{$edge->[0]} = 1;
626 } elsif( $used{$edge->[1]} ) {
627 $substart{$edge->[1]} = 1;
630 # Add substitute start and end edges if necessary
631 foreach my $node ( keys %substart ) {
632 my $witstr = $self->_path_display_label ( $self->reading_witnesses( $self->reading( $node ) ) );
633 my $variables = { %edge_attrs, 'label' => $witstr };
634 my $varopts = _dot_attr_string( $variables );
635 $dot .= "\t\"#SUBSTART#\" -> \"$node\" $varopts;";
637 foreach my $node ( keys %subend ) {
638 my $witstr = $self->_path_display_label ( $self->reading_witnesses( $self->reading( $node ) ) );
639 my $variables = { %edge_attrs, 'label' => $witstr };
640 my $varopts = _dot_attr_string( $variables );
641 $dot .= "\t\"$node\" -> \"#SUBEND#\" $varopts;";
644 if( $STRAIGHTENHACK ) {
645 $dot .= "\t\"#END#\" -> \"#SILENT#\" [ color=white,penwidth=0 ];\n";
652 sub _dot_attr_string {
655 foreach my $k ( sort keys %$hash ) {
657 push( @attrs, $k.'="'.$v.'"' );
659 return( '[ ' . join( ', ', @attrs ) . ' ]' );
662 sub _add_edge_weights {
664 # Walk the graph from START to END, choosing the successor node with
665 # the largest number of witness paths each time.
667 my $curr = $self->start->id;
668 my $ranked = $self->end->has_rank;
669 while( $curr ne $self->end->id ) {
670 my $rank = $ranked ? $self->reading( $curr )->rank : 0;
671 my @succ = sort { $self->path_witnesses( $curr, $a )
672 <=> $self->path_witnesses( $curr, $b ) }
673 $self->sequence->successors( $curr );
674 my $next = pop @succ;
675 my $nextrank = $ranked ? $self->reading( $next )->rank : 0;
676 # Try to avoid lacunae in the weighted path.
678 ( $self->reading( $next )->is_lacuna ||
679 $nextrank - $rank > 1 ) ){
682 $weighted->{$curr} = $next;
688 =head2 path_witnesses( $edge )
690 Returns the list of sigils whose witnesses are associated with the given edge.
691 The edge can be passed as either an array or an arrayref of ( $source, $target ).
696 my( $self, @edge ) = @_;
697 # If edge is an arrayref, cope.
698 if( @edge == 1 && ref( $edge[0] ) eq 'ARRAY' ) {
702 my @wits = keys %{$self->sequence->get_edge_attributes( @edge )};
706 sub _path_display_label {
709 my $maj = scalar( $self->tradition->witnesses ) * 0.6;
710 if( scalar @wits > $maj ) {
711 # TODO break out a.c. wits
714 return join( ', ', @wits );
718 =head2 witnesses_at_rank
720 Returns a list of witnesses that are not lacunose, for a given rank.
724 sub witnesses_at_rank {
725 my( $self, $rank ) = @_;
730 Returns a GraphML representation of the collation. The GraphML will contain
731 two graphs. The first expresses the attributes of the readings and the witness
732 paths that link them; the second expresses the relationships that link the
733 readings. This is the native transfer format for a tradition.
742 my $datafile = 't/data/florilegium_tei_ps.xml';
743 my $tradition = Text::Tradition->new( 'input' => 'TEI',
748 ok( $tradition, "Got a tradition object" );
749 is( scalar $tradition->witnesses, 13, "Found all witnesses" );
750 ok( $tradition->collation, "Tradition has a collation" );
752 my $c = $tradition->collation;
753 is( scalar $c->readings, $READINGS, "Collation has all readings" );
754 is( scalar $c->paths, $PATHS, "Collation has all paths" );
755 is( scalar $c->relationships, 0, "Collation has all relationships" );
757 # Add a few relationships
758 $c->add_relationship( 'w123', 'w125', { 'type' => 'collated' } );
759 $c->add_relationship( 'w193', 'w196', { 'type' => 'collated' } );
760 $c->add_relationship( 'w257', 'w262', { 'type' => 'transposition' } );
762 # Now write it to GraphML and parse it again.
764 my $graphml = $c->as_graphml;
765 my $st = Text::Tradition->new( 'input' => 'Self', 'string' => $graphml );
766 is( scalar $st->collation->readings, $READINGS, "Reparsed collation has all readings" );
767 is( scalar $st->collation->paths, $PATHS, "Reparsed collation has all paths" );
768 is( scalar $st->collation->relationships, 3, "Reparsed collation has new relationships" );
776 $self->calculate_ranks unless $self->_graphcalc_done;
779 my $graphml_ns = 'http://graphml.graphdrawing.org/xmlns';
780 my $xsi_ns = 'http://www.w3.org/2001/XMLSchema-instance';
781 my $graphml_schema = 'http://graphml.graphdrawing.org/xmlns ' .
782 'http://graphml.graphdrawing.org/xmlns/1.0/graphml.xsd';
784 # Create the document and root node
785 my $graphml = XML::LibXML->createDocument( "1.0", "UTF-8" );
786 my $root = $graphml->createElementNS( $graphml_ns, 'graphml' );
787 $graphml->setDocumentElement( $root );
788 $root->setNamespace( $xsi_ns, 'xsi', 0 );
789 $root->setAttributeNS( $xsi_ns, 'schemaLocation', $graphml_schema );
791 # Add the data keys for the graph
794 my @graph_attributes = qw/ version wit_list_separator baselabel linear ac_label /;
795 foreach my $datum ( @graph_attributes ) {
796 $graph_data_keys{$datum} = 'dg'.$gdi++;
797 my $key = $root->addNewChild( $graphml_ns, 'key' );
798 $key->setAttribute( 'attr.name', $datum );
799 $key->setAttribute( 'attr.type', $key eq 'linear' ? 'boolean' : 'string' );
800 $key->setAttribute( 'for', 'graph' );
801 $key->setAttribute( 'id', $graph_data_keys{$datum} );
804 # Add the data keys for nodes
811 is_start => 'boolean',
813 is_lacuna => 'boolean',
815 foreach my $datum ( keys %node_data ) {
816 $node_data_keys{$datum} = 'dn'.$ndi++;
817 my $key = $root->addNewChild( $graphml_ns, 'key' );
818 $key->setAttribute( 'attr.name', $datum );
819 $key->setAttribute( 'attr.type', $node_data{$datum} );
820 $key->setAttribute( 'for', 'node' );
821 $key->setAttribute( 'id', $node_data_keys{$datum} );
824 # Add the data keys for edges, i.e. witnesses
828 class => 'string', # Class, deprecated soon
829 witness => 'string', # ID/label for a path
830 relationship => 'string', # ID/label for a relationship
831 extra => 'boolean', # Path key
832 scope => 'string', # Relationship key
833 annotation => 'string', # Relationship key
834 non_correctable => 'boolean', # Relationship key
835 non_independent => 'boolean', # Relationship key
837 foreach my $datum ( keys %edge_data ) {
838 $edge_data_keys{$datum} = 'de'.$edi++;
839 my $key = $root->addNewChild( $graphml_ns, 'key' );
840 $key->setAttribute( 'attr.name', $datum );
841 $key->setAttribute( 'attr.type', $edge_data{$datum} );
842 $key->setAttribute( 'for', 'edge' );
843 $key->setAttribute( 'id', $edge_data_keys{$datum} );
846 # Add the collation graph itself
847 my $sgraph = $root->addNewChild( $graphml_ns, 'graph' );
848 $sgraph->setAttribute( 'edgedefault', 'directed' );
849 $sgraph->setAttribute( 'id', $self->tradition->name );
850 $sgraph->setAttribute( 'parse.edgeids', 'canonical' );
851 $sgraph->setAttribute( 'parse.edges', scalar($self->paths) );
852 $sgraph->setAttribute( 'parse.nodeids', 'canonical' );
853 $sgraph->setAttribute( 'parse.nodes', scalar($self->readings) );
854 $sgraph->setAttribute( 'parse.order', 'nodesfirst' );
856 # Collation attribute data
857 foreach my $datum ( @graph_attributes ) {
858 my $value = $datum eq 'version' ? '3.0' : $self->$datum;
859 _add_graphml_data( $sgraph, $graph_data_keys{$datum}, $value );
864 # Add our readings to the graph
865 foreach my $n ( sort { $a->id cmp $b->id } $self->readings ) {
866 # Add to the main graph
867 my $node_el = $sgraph->addNewChild( $graphml_ns, 'node' );
868 my $node_xmlid = 'n' . $node_ctr++;
869 $node_hash{ $n->id } = $node_xmlid;
870 $node_el->setAttribute( 'id', $node_xmlid );
871 foreach my $d ( keys %node_data ) {
873 _add_graphml_data( $node_el, $node_data_keys{$d}, $nval )
878 # Add the path edges to the sequence graph
880 foreach my $e ( sort { $a->[0] cmp $b->[0] } $self->sequence->edges() ) {
881 # We add an edge in the graphml for every witness in $e.
882 foreach my $wit ( sort $self->path_witnesses( $e ) ) {
883 my( $id, $from, $to ) = ( 'e'.$edge_ctr++,
884 $node_hash{ $e->[0] },
885 $node_hash{ $e->[1] } );
886 my $edge_el = $sgraph->addNewChild( $graphml_ns, 'edge' );
887 $edge_el->setAttribute( 'source', $from );
888 $edge_el->setAttribute( 'target', $to );
889 $edge_el->setAttribute( 'id', $id );
891 # It's a witness path, so add the witness
893 my $key = $edge_data_keys{'witness'};
894 # Is this an ante-corr witness?
895 my $aclabel = $self->ac_label;
896 if( $wit =~ /^(.*)\Q$aclabel\E$/ ) {
897 # Keep the base witness
899 # ...and record that this is an 'extra' reading path
900 _add_graphml_data( $edge_el, $edge_data_keys{'extra'}, $aclabel );
902 _add_graphml_data( $edge_el, $edge_data_keys{'witness'}, $base );
903 _add_graphml_data( $edge_el, $edge_data_keys{'class'}, 'path' );
907 # Add the relationship graph to the XML
908 $self->relations->_as_graphml( $graphml_ns, $root, \%node_hash,
909 $node_data_keys{'id'}, \%edge_data_keys );
911 # Save and return the thing
912 my $result = decode_utf8( $graphml->toString(1) );
916 sub _add_graphml_data {
917 my( $el, $key, $value ) = @_;
918 return unless defined $value;
919 my $data_el = $el->addNewChild( $el->namespaceURI, 'data' );
920 $data_el->setAttribute( 'key', $key );
921 $data_el->appendText( $value );
926 Returns a CSV alignment table representation of the collation graph, one
927 row per witness (or witness uncorrected.)
933 my $table = $self->alignment_table;
934 my $csv = Text::CSV_XS->new( { binary => 1, quote_null => 0 } );
936 # Make the header row
937 $csv->combine( map { $_->{'witness'} } @{$table->{'alignment'}} );
938 push( @result, decode_utf8( $csv->string ) );
939 # Make the rest of the rows
940 foreach my $idx ( 0 .. $table->{'length'} - 1 ) {
941 my @rowobjs = map { $_->{'tokens'}->[$idx] } @{$table->{'alignment'}};
942 my @row = map { $_ ? $_->{'t'}->text : $_ } @rowobjs;
943 $csv->combine( @row );
944 push( @result, decode_utf8( $csv->string ) );
946 return join( "\n", @result );
949 =head2 alignment_table( $use_refs, $include_witnesses )
951 Return a reference to an alignment table, in a slightly enhanced CollateX
952 format which looks like this:
954 $table = { alignment => [ { witness => "SIGIL",
955 tokens => [ { t => "TEXT" }, ... ] },
957 tokens => [ { t => "TEXT" }, ... ] },
961 If $use_refs is set to 1, the reading object is returned in the table
962 instead of READINGTEXT; if not, the text of the reading is returned.
964 If $include_witnesses is set to a hashref, only the witnesses whose sigil
965 keys have a true hash value will be included.
969 sub alignment_table {
971 $self->calculate_ranks() unless $self->_graphcalc_done;
972 return $self->cached_table if $self->has_cached_table;
974 # Make sure we can do this
975 throw( "Need a linear graph in order to make an alignment table" )
976 unless $self->linear;
977 $self->calculate_ranks unless $self->end->has_rank;
979 my $table = { 'alignment' => [], 'length' => $self->end->rank - 1 };
980 my @all_pos = ( 1 .. $self->end->rank - 1 );
981 foreach my $wit ( sort { $a->sigil cmp $b->sigil } $self->tradition->witnesses ) {
982 # print STDERR "Making witness row(s) for " . $wit->sigil . "\n";
983 my @wit_path = $self->reading_sequence( $self->start, $self->end, $wit->sigil );
984 my @row = _make_witness_row( \@wit_path, \@all_pos );
985 push( @{$table->{'alignment'}},
986 { 'witness' => $wit->sigil, 'tokens' => \@row } );
987 if( $wit->is_layered ) {
988 my @wit_ac_path = $self->reading_sequence( $self->start, $self->end,
989 $wit->sigil.$self->ac_label );
990 my @ac_row = _make_witness_row( \@wit_ac_path, \@all_pos );
991 push( @{$table->{'alignment'}},
992 { 'witness' => $wit->sigil.$self->ac_label, 'tokens' => \@ac_row } );
995 $self->cached_table( $table );
999 sub _make_witness_row {
1000 my( $path, $positions ) = @_;
1002 map { $char_hash{$_} = undef } @$positions;
1004 foreach my $rdg ( @$path ) {
1005 my $rtext = $rdg->text;
1006 $rtext = '#LACUNA#' if $rdg->is_lacuna;
1007 print STDERR "rank " . $rdg->rank . "\n" if $debug;
1008 # print STDERR "No rank for " . $rdg->id . "\n" unless defined $rdg->rank;
1009 $char_hash{$rdg->rank} = { 't' => $rdg };
1011 my @row = map { $char_hash{$_} } @$positions;
1012 # Fill in lacuna markers for undef spots in the row
1013 my $last_el = shift @row;
1014 my @filled_row = ( $last_el );
1015 foreach my $el ( @row ) {
1016 # If we are using node reference, make the lacuna node appear many times
1017 # in the table. If not, use the lacuna tag.
1018 if( $last_el && $last_el->{'t'}->is_lacuna && !defined $el ) {
1021 push( @filled_row, $el );
1027 =head1 NAVIGATION METHODS
1029 =head2 reading_sequence( $first, $last, $sigil, $backup )
1031 Returns the ordered list of readings, starting with $first and ending
1032 with $last, for the witness given in $sigil. If a $backup sigil is
1033 specified (e.g. when walking a layered witness), it will be used wherever
1034 no $sigil path exists. If there is a base text reading, that will be
1035 used wherever no path exists for $sigil or $backup.
1039 # TODO Think about returning some lazy-eval iterator.
1040 # TODO Get rid of backup; we should know from what witness is whether we need it.
1042 sub reading_sequence {
1043 my( $self, $start, $end, $witness ) = @_;
1045 $witness = $self->baselabel unless $witness;
1046 my @readings = ( $start );
1049 while( $n && $n->id ne $end->id ) {
1050 if( exists( $seen{$n->id} ) ) {
1051 throw( "Detected loop for $witness at " . $n->id );
1055 my $next = $self->next_reading( $n, $witness );
1057 throw( "Did not find any path for $witness from reading " . $n->id );
1059 push( @readings, $next );
1062 # Check that the last reading is our end reading.
1063 my $last = $readings[$#readings];
1064 throw( "Last reading found from " . $start->text .
1065 " for witness $witness is not the end!" ) # TODO do we get this far?
1066 unless $last->id eq $end->id;
1071 =head2 next_reading( $reading, $sigil );
1073 Returns the reading that follows the given reading along the given witness
1079 # Return the successor via the corresponding path.
1081 my $answer = $self->_find_linked_reading( 'next', @_ );
1082 return undef unless $answer;
1083 return $self->reading( $answer );
1086 =head2 prior_reading( $reading, $sigil )
1088 Returns the reading that precedes the given reading along the given witness
1094 # Return the predecessor via the corresponding path.
1096 my $answer = $self->_find_linked_reading( 'prior', @_ );
1097 return $self->reading( $answer );
1100 sub _find_linked_reading {
1101 my( $self, $direction, $node, $path ) = @_;
1103 # Get a backup if we are dealing with a layered witness
1105 my $aclabel = $self->ac_label;
1106 if( $path && $path =~ /^(.*)\Q$aclabel\E$/ ) {
1110 my @linked_paths = $direction eq 'next'
1111 ? $self->sequence->edges_from( $node )
1112 : $self->sequence->edges_to( $node );
1113 return undef unless scalar( @linked_paths );
1115 # We have to find the linked path that contains all of the
1116 # witnesses supplied in $path.
1117 my( @path_wits, @alt_path_wits );
1118 @path_wits = sort( $self->_witnesses_of_label( $path ) ) if $path;
1119 @alt_path_wits = sort( $self->_witnesses_of_label( $alt_path ) ) if $alt_path;
1122 foreach my $le ( @linked_paths ) {
1123 if( $self->sequence->has_edge_attribute( @$le, $self->baselabel ) ) {
1126 my @le_wits = sort $self->path_witnesses( $le );
1127 if( _is_within( \@path_wits, \@le_wits ) ) {
1128 # This is the right path.
1129 return $direction eq 'next' ? $le->[1] : $le->[0];
1130 } elsif( _is_within( \@alt_path_wits, \@le_wits ) ) {
1134 # Got this far? Return the alternate path if it exists.
1135 return $direction eq 'next' ? $alt_le->[1] : $alt_le->[0]
1138 # Got this far? Return the base path if it exists.
1139 return $direction eq 'next' ? $base_le->[1] : $base_le->[0]
1142 # Got this far? We have no appropriate path.
1143 warn "Could not find $direction node from " . $node->id
1144 . " along path $path";
1150 my( $set1, $set2 ) = @_;
1151 my $ret = @$set1; # will be 0, i.e. false, if set1 is empty
1152 foreach my $el ( @$set1 ) {
1153 $ret = 0 unless grep { /^\Q$el\E$/ } @$set2;
1158 # Return the string that joins together a list of witnesses for
1159 # display on a single path.
1160 sub _witnesses_of_label {
1161 my( $self, $label ) = @_;
1162 my $regex = $self->wit_list_separator;
1163 my @answer = split( /\Q$regex\E/, $label );
1167 =head2 common_readings
1169 Returns the list of common readings in the graph (i.e. those readings that are
1170 shared by all non-lacunose witnesses.)
1174 sub common_readings {
1176 my @common = grep { $_->is_common } $self->readings;
1180 =head2 path_text( $sigil, $mainsigil [, $start, $end ] )
1182 Returns the text of a witness (plus its backup, if we are using a layer)
1183 as stored in the collation. The text is returned as a string, where the
1184 individual readings are joined with spaces and the meta-readings (e.g.
1185 lacunae) are omitted. Optional specification of $start and $end allows
1186 the generation of a subset of the witness text.
1191 my( $self, $wit, $start, $end ) = @_;
1192 $start = $self->start unless $start;
1193 $end = $self->end unless $end;
1194 my @path = grep { !$_->is_meta } $self->reading_sequence( $start, $end, $wit );
1195 return join( ' ', map { $_->text } @path );
1198 =head1 INITIALIZATION METHODS
1200 These are mostly for use by parsers.
1202 =head2 make_witness_path( $witness )
1204 Link the array of readings contained in $witness->path (and in
1205 $witness->uncorrected_path if it exists) into collation paths.
1206 Clear out the arrays when finished.
1208 =head2 make_witness_paths
1210 Call make_witness_path for all witnesses in the tradition.
1214 # For use when a collation is constructed from a base text and an apparatus.
1215 # We have the sequences of readings and just need to add path edges.
1216 # When we are done, clear out the witness path attributes, as they are no
1218 # TODO Find a way to replace the witness path attributes with encapsulated functions?
1220 sub make_witness_paths {
1222 foreach my $wit ( $self->tradition->witnesses ) {
1223 # print STDERR "Making path for " . $wit->sigil . "\n";
1224 $self->make_witness_path( $wit );
1228 sub make_witness_path {
1229 my( $self, $wit ) = @_;
1230 my @chain = @{$wit->path};
1231 my $sig = $wit->sigil;
1232 foreach my $idx ( 0 .. $#chain-1 ) {
1233 $self->add_path( $chain[$idx], $chain[$idx+1], $sig );
1235 if( $wit->is_layered ) {
1236 @chain = @{$wit->uncorrected_path};
1237 foreach my $idx( 0 .. $#chain-1 ) {
1238 my $source = $chain[$idx];
1239 my $target = $chain[$idx+1];
1240 $self->add_path( $source, $target, $sig.$self->ac_label )
1241 unless $self->has_path( $source, $target, $sig );
1245 $wit->clear_uncorrected_path;
1248 =head2 calculate_ranks
1250 Calculate the reading ranks (that is, their aligned positions relative
1251 to each other) for the graph. This can only be called on linear collations.
1255 use Text::Tradition;
1257 my $cxfile = 't/data/Collatex-16.xml';
1258 my $t = Text::Tradition->new(
1260 'input' => 'CollateX',
1263 my $c = $t->collation;
1266 my $table = $c->alignment_table;
1267 ok( $c->has_cached_table, "Alignment table was cached" );
1268 is( $c->alignment_table, $table, "Cached table returned upon second call" );
1269 $c->calculate_ranks;
1270 is( $c->alignment_table, $table, "Cached table retained with no rank change" );
1271 $c->add_relationship( 'n9', 'n23', { 'type' => 'spelling' } );
1272 isnt( $c->alignment_table, $table, "Alignment table changed after relationship add" );
1278 sub calculate_ranks {
1280 # Save the existing ranks, in case we need to invalidate the cached SVG.
1282 # Walk a version of the graph where every node linked by a relationship
1283 # edge is fundamentally the same node, and do a topological ranking on
1284 # the nodes in this graph.
1285 my $topo_graph = Graph->new();
1289 foreach my $r ( $self->readings ) {
1290 next if exists $rel_containers{$r->id};
1291 my @rels = $r->related_readings( 'colocated' );
1293 # Make a relationship container.
1295 my $rn = 'rel_container_' . $rel_ctr++;
1296 $topo_graph->add_vertex( $rn );
1298 $rel_containers{$_->id} = $rn;
1301 # Add a new node to mirror the old node.
1302 $rel_containers{$r->id} = $r->id;
1303 $topo_graph->add_vertex( $r->id );
1308 foreach my $r ( $self->readings ) {
1309 $existing_ranks{$r} = $r->rank;
1310 foreach my $n ( $self->sequence->successors( $r->id ) ) {
1311 my( $tfrom, $tto ) = ( $rel_containers{$r->id},
1312 $rel_containers{$n} );
1313 # $DB::single = 1 unless $tfrom && $tto;
1314 $topo_graph->add_edge( $tfrom, $tto );
1318 # Now do the rankings, starting with the start node.
1319 my $topo_start = $rel_containers{$self->start->id};
1320 my $node_ranks = { $topo_start => 0 };
1321 my @curr_origin = ( $topo_start );
1322 # A little iterative function.
1323 while( @curr_origin ) {
1324 @curr_origin = _assign_rank( $topo_graph, $node_ranks, @curr_origin );
1326 # Transfer our rankings from the topological graph to the real one.
1327 foreach my $r ( $self->readings ) {
1328 if( defined $node_ranks->{$rel_containers{$r->id}} ) {
1329 $r->rank( $node_ranks->{$rel_containers{$r->id}} );
1331 # Die. Find the last rank we calculated.
1332 my @all_defined = sort { $node_ranks->{$rel_containers{$a->id}}
1333 <=> $node_ranks->{$rel_containers{$b->id}} }
1335 my $last = pop @all_defined;
1336 throw( "Ranks not calculated after $last - do you have a cycle in the graph?" );
1339 # Do we need to invalidate the cached data?
1340 if( $self->has_cached_svg || $self->has_cached_table ) {
1341 foreach my $r ( $self->readings ) {
1342 next if $existing_ranks{$r} == $r->rank;
1343 # Something has changed, so clear the cache
1344 $self->_clear_cache;
1345 # ...and recalculate the common readings.
1346 $self->calculate_common_readings();
1350 # The graph calculation information is now up to date.
1351 $self->_graphcalc_done(1);
1355 my( $graph, $node_ranks, @current_nodes ) = @_;
1356 # Look at each of the children of @current_nodes. If all the child's
1357 # parents have a rank, assign it the highest rank + 1 and add it to
1358 # @next_nodes. Otherwise skip it; we will return when the highest-ranked
1359 # parent gets a rank.
1361 foreach my $c ( @current_nodes ) {
1362 warn "Current reading $c has no rank!"
1363 unless exists $node_ranks->{$c};
1364 # print STDERR "Looking at child of node $c, rank "
1365 # . $node_ranks->{$c} . "\n";
1366 foreach my $child ( $graph->successors( $c ) ) {
1367 next if exists $node_ranks->{$child};
1368 my $highest_rank = -1;
1370 foreach my $parent ( $graph->predecessors( $child ) ) {
1371 if( exists $node_ranks->{$parent} ) {
1372 $highest_rank = $node_ranks->{$parent}
1373 if $highest_rank <= $node_ranks->{$parent};
1380 my $c_rank = $highest_rank + 1;
1381 # print STDERR "Assigning rank $c_rank to node $child \n";
1382 $node_ranks->{$child} = $c_rank;
1383 push( @next_nodes, $child );
1391 $self->wipe_svg if $self->has_cached_svg;
1392 $self->wipe_table if $self->has_cached_table;
1396 =head2 flatten_ranks
1398 A convenience method for parsing collation data. Searches the graph for readings
1399 with the same text at the same rank, and merges any that are found.
1405 my %unique_rank_rdg;
1406 foreach my $rdg ( $self->readings ) {
1407 next unless $rdg->has_rank;
1408 my $key = $rdg->rank . "||" . $rdg->text;
1409 if( exists $unique_rank_rdg{$key} ) {
1411 # print STDERR "Combining readings at same rank: $key\n";
1412 $self->merge_readings( $unique_rank_rdg{$key}, $rdg );
1413 # TODO see if this now makes a common point.
1415 $unique_rank_rdg{$key} = $rdg;
1421 =head2 calculate_common_readings
1423 Goes through the graph identifying the readings that appear in every witness
1424 (apart from those with lacunae at that spot.) Marks them as common and returns
1429 use Text::Tradition;
1431 my $cxfile = 't/data/Collatex-16.xml';
1432 my $t = Text::Tradition->new(
1434 'input' => 'CollateX',
1437 my $c = $t->collation;
1439 my @common = $c->calculate_common_readings();
1440 is( scalar @common, 8, "Found correct number of common readings" );
1441 my @marked = sort $c->common_readings();
1442 is( scalar @common, 8, "All common readings got marked as such" );
1443 my @expected = qw/ n1 n12 n16 n19 n20 n5 n6 n7 /;
1444 is_deeply( \@marked, \@expected, "Found correct list of common readings" );
1450 sub calculate_common_readings {
1453 map { $_->is_common( 0 ) } $self->readings;
1454 # Implicitly calls calculate_ranks
1455 my $table = $self->alignment_table;
1456 foreach my $idx ( 0 .. $table->{'length'} - 1 ) {
1457 my @row = map { $_->{'tokens'}->[$idx]
1458 ? $_->{'tokens'}->[$idx]->{'t'} : '' }
1459 @{$table->{'alignment'}};
1461 foreach my $r ( @row ) {
1463 $hash{$r->id} = $r unless $r->is_meta;
1465 $hash{'UNDEF'} = $r;
1468 if( keys %hash == 1 && !exists $hash{'UNDEF'} ) {
1469 my( $r ) = values %hash;
1471 push( @common, $r );
1477 =head2 text_from_paths
1479 Calculate the text array for all witnesses from the path, for later consistency
1480 checking. Only to be used if there is no non-graph-based way to know the
1485 sub text_from_paths {
1487 foreach my $wit ( $self->tradition->witnesses ) {
1488 my @text = split( /\s+/,
1489 $self->reading_sequence( $self->start, $self->end, $wit->sigil ) );
1490 $wit->text( \@text );
1491 if( $wit->is_layered ) {
1492 my @uctext = split( /\s+/,
1493 $self->reading_sequence( $self->start, $self->end,
1494 $wit->sigil.$self->ac_label ) );
1495 $wit->text( \@uctext );
1500 =head1 UTILITY FUNCTIONS
1502 =head2 common_predecessor( $reading_a, $reading_b )
1504 Find the last reading that occurs in sequence before both the given readings.
1506 =head2 common_successor( $reading_a, $reading_b )
1508 Find the first reading that occurs in sequence after both the given readings.
1512 use Text::Tradition;
1514 my $cxfile = 't/data/Collatex-16.xml';
1515 my $t = Text::Tradition->new(
1517 'input' => 'CollateX',
1520 my $c = $t->collation;
1522 is( $c->common_predecessor( 'n9', 'n23' )->id,
1523 'n20', "Found correct common predecessor" );
1524 is( $c->common_successor( 'n9', 'n23' )->id,
1525 '#END#', "Found correct common successor" );
1527 is( $c->common_predecessor( 'n19', 'n17' )->id,
1528 'n16', "Found correct common predecessor for readings on same path" );
1529 is( $c->common_successor( 'n21', 'n26' )->id,
1530 '#END#', "Found correct common successor for readings on same path" );
1536 ## Return the closest reading that is a predecessor of both the given readings.
1537 sub common_predecessor {
1539 my( $r1, $r2 ) = $self->_objectify_args( @_ );
1540 return $self->_common_in_path( $r1, $r2, 'predecessors' );
1543 sub common_successor {
1545 my( $r1, $r2 ) = $self->_objectify_args( @_ );
1546 return $self->_common_in_path( $r1, $r2, 'successors' );
1549 sub _common_in_path {
1550 my( $self, $r1, $r2, $dir ) = @_;
1551 my $iter = $r1->rank > $r2->rank ? $r1->rank : $r2->rank;
1552 $iter = $self->end->rank - $iter if $dir eq 'successors';
1554 my @last_checked = ( $r1, $r2 );
1556 while( !@candidates ) {
1558 foreach my $lc ( @last_checked ) {
1559 foreach my $p ( $lc->$dir ) {
1560 if( $all_seen{$p->id} ) {
1561 push( @candidates, $p );
1563 $all_seen{$p->id} = 1;
1564 push( @new_lc, $p );
1568 @last_checked = @new_lc;
1570 my @answer = sort { $a->rank <=> $b->rank } @candidates;
1571 return $dir eq 'predecessors' ? pop( @answer ) : shift ( @answer );
1575 Text::Tradition::Error->throw(
1576 'ident' => 'Collation error',
1582 __PACKAGE__->meta->make_immutable;
1586 This package is free software and is provided "as is" without express
1587 or implied warranty. You can redistribute it and/or modify it under
1588 the same terms as Perl itself.
1592 Tara L Andrews E<lt>aurum@cpan.orgE<gt>