X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=lib%2FText%2FTradition%2FCollation.pm;h=fcdd1ff78959c66662fc5ac00a9481fba6afdf90;hb=2a8127263ef278f3f14b480a12b84f9aa4f92fdc;hp=7b261359f263c341f8b561cc6d7be6417e203d2c;hpb=d3e7842a9402304b1b701c2a72db001b324f1f2f;p=scpubgit%2Fstemmatology.git diff --git a/lib/Text/Tradition/Collation.pm b/lib/Text/Tradition/Collation.pm index 7b26135..fcdd1ff 100644 --- a/lib/Text/Tradition/Collation.pm +++ b/lib/Text/Tradition/Collation.pm @@ -284,6 +284,9 @@ sub add_reading { my( $self, $reading ) = @_; unless( ref( $reading ) eq 'Text::Tradition::Collation::Reading' ) { my %args = %$reading; + if( $self->tradition->has_language && !exists $args{'language'} ) { + $args{'language'} = $self->tradition->language; + } $reading = Text::Tradition::Collation::Reading->new( 'collation' => $self, %args ); @@ -639,15 +642,15 @@ sub as_dot { # Output substitute start/end readings if necessary if( $startrank ) { - $dot .= "\t\"#SUBSTART#\" [ label=\"...\" ];\n"; + $dot .= "\t\"__SUBSTART__\" [ label=\"...\",id=\"__START__\" ];\n"; } if( $endrank ) { - $dot .= "\t\"#SUBEND#\" [ label=\"...\" ];\n"; + $dot .= "\t\"__SUBEND__\" [ label=\"...\",id=\"__END__\" ];\n"; } if( $STRAIGHTENHACK ) { ## HACK part 1 - my $startlabel = $startrank ? 'SUBSTART' : 'START'; - $dot .= "\tsubgraph { rank=same \"#$startlabel#\" \"#SILENT#\" }\n"; + my $startlabel = $startrank ? '__SUBSTART__' : '__START__'; + $dot .= "\tsubgraph { rank=same \"$startlabel\" \"#SILENT#\" }\n"; $dot .= "\t\"#SILENT#\" [ shape=diamond,color=white,penwidth=0,label=\"\" ];" } my %used; # Keep track of the readings that actually appear in the graph @@ -669,6 +672,7 @@ sub as_dot { $label = "-$label" if $reading->join_prior; $label =~ s/\"/\\\"/g; $rattrs->{'label'} = $label; + $rattrs->{'id'} = $reading->id; $rattrs->{'fillcolor'} = '#b3f36d' if $reading->is_common && $color_common; $dot .= sprintf( "\t\"%s\" %s;\n", $reading->id, _dot_attr_string( $rattrs ) ); } @@ -718,18 +722,18 @@ sub as_dot { my $witstr = $self->_path_display_label ( $self->reading_witnesses( $self->reading( $node ) ) ); my $variables = { %edge_attrs, 'label' => $witstr }; my $varopts = _dot_attr_string( $variables ); - $dot .= "\t\"#SUBSTART#\" -> \"$node\" $varopts;"; + $dot .= "\t\"__SUBSTART__\" -> \"$node\" $varopts;"; } foreach my $node ( keys %subend ) { my $witstr = $self->_path_display_label ( $self->reading_witnesses( $self->reading( $node ) ) ); my $variables = { %edge_attrs, 'label' => $witstr }; my $varopts = _dot_attr_string( $variables ); - $dot .= "\t\"$node\" -> \"#SUBEND#\" $varopts;"; + $dot .= "\t\"$node\" -> \"__SUBEND__\" $varopts;"; } # HACK part 2 if( $STRAIGHTENHACK ) { - my $endlabel = $endrank ? 'SUBEND' : 'END'; - $dot .= "\t\"#$endlabel#\" -> \"#SILENT#\" [ color=white,penwidth=0 ];\n"; + my $endlabel = $endrank ? '__SUBEND__' : '__END__'; + $dot .= "\t\"$endlabel\" -> \"#SILENT#\" [ color=white,penwidth=0 ];\n"; } $dot .= "}\n"; @@ -883,6 +887,13 @@ is( scalar $st->collation->readings, $READINGS, "Reparsed collation has all read is( scalar $st->collation->paths, $PATHS, "Reparsed collation has all paths" ); is( scalar $st->collation->relationships, 3, "Reparsed collation has new relationships" ); +# Now add a stemma, write to GraphML, and parse again. +my $stemma = $tradition->add_stemma( 'dotfile' => 't/data/florilegium.dot' ); +is( ref( $stemma ), 'Text::Tradition::Stemma', "Parsed dotfile into stemma" ); +is( $tradition->stemmata, 1, "Tradition now has the stemma" ); +$graphml = $c->as_graphml; +like( $graphml, qr/digraph/, "Digraph declaration exists in GraphML" ); + =end testing =cut @@ -926,6 +937,7 @@ sub as_graphml { 'Str' => 'string', 'Int' => 'int', 'Bool' => 'boolean', + 'ReadingID' => 'string', 'RelationshipType' => 'string', 'RelationshipScope' => 'string', ); @@ -952,6 +964,8 @@ sub as_graphml { next unless $save_types{$attr->type_constraint->name}; $graph_attributes{$attr->name} = $save_types{$attr->type_constraint->name}; } + # Extra custom key for the tradition stemma(ta) + $graph_attributes{'stemmata'} = 'string'; foreach my $datum ( sort keys %graph_attributes ) { $graph_data_keys{$datum} = 'dg'.$gdi++; @@ -971,6 +985,9 @@ sub as_graphml { next unless $save_types{$attr->type_constraint->name}; $reading_attributes{$attr->name} = $save_types{$attr->type_constraint->name}; } + # Extra custom key for the reading morphology + $reading_attributes{'lexemes'} = 'string'; + my %node_data_keys; my $ndi = 0; foreach my $datum ( sort keys %reading_attributes ) { @@ -1022,11 +1039,16 @@ sub as_graphml { $sgraph->setAttribute( 'parse.nodes', 0 ); # fill in later $sgraph->setAttribute( 'parse.order', 'nodesfirst' ); - # Collation attribute data + # Tradition/collation attribute data foreach my $datum ( keys %graph_attributes ) { my $value; if( $datum eq 'version' ) { - $value = '3.1'; + $value = '3.2'; + } elsif( $datum eq 'stemmata' ) { + my @stemstrs; + map { push( @stemstrs, $_->editable( {linesep => ''} ) ) } + $self->tradition->stemmata; + $value = join( "\n", @stemstrs ); } elsif( $gattr_from{$datum} eq 'Tradition' ) { $value = $self->tradition->$datum; } else { @@ -1049,6 +1071,13 @@ sub as_graphml { $node_el->setAttribute( 'id', $node_xmlid ); foreach my $d ( keys %reading_attributes ) { my $nval = $n->$d; + # Custom serialization + if( $d eq 'lexemes' ) { + # If nval is a true value, we have lexemes so we need to + # serialize them. Otherwise set nval to undef so that the + # key is excluded from this reading. + $nval = $nval ? $n->_serialize_lexemes : undef; + } if( $rankoffset && $d eq 'rank' && $n ne $self->start ) { # Adjust the ranks within the subgraph. $nval = $n eq $self->end ? $end->rank - $rankoffset + 1 @@ -1386,21 +1415,13 @@ sub path_text { $start = $self->start unless $start; $end = $self->end unless $end; my @path = grep { !$_->is_meta } $self->reading_sequence( $start, $end, $wit ); - return _text_from_path( @path ); -} - -# Utility function so that we can cheat and use it when we need both the path -# and its text. -sub _text_from_path { - my( $self, @path ) = @_; my $pathtext = ''; my $last; foreach my $r ( @path ) { - if( $r->join_prior || !$last || $last->join_next ) { - $pathtext .= $r->text; - } else { - $pathtext .= ' ' . $r->text; - } + unless ( $r->join_prior || !$last || $last->join_next ) { + $pathtext .= ' '; + } + $pathtext .= $r->text; $last = $r; } return $pathtext; @@ -1722,12 +1743,12 @@ my $c = $t->collation; is( $c->common_predecessor( 'n24', 'n23' )->id, 'n20', "Found correct common predecessor" ); is( $c->common_successor( 'n24', 'n23' )->id, - '#END#', "Found correct common successor" ); + '__END__', "Found correct common successor" ); is( $c->common_predecessor( 'n19', 'n17' )->id, 'n16', "Found correct common predecessor for readings on same path" ); is( $c->common_successor( 'n21', 'n10' )->id, - '#END#', "Found correct common successor for readings on same path" ); + '__END__', "Found correct common successor for readings on same path" ); =end testing