simplify Directory and add exceptions;

[scpubgit/stemmatology.git] / lib / Text / Tradition / Collation.pm

diff --git a/lib/Text/Tradition/Collation.pm b/lib/Text/Tradition/Collation.pm
index ee8e9da..77a367e 100644 (file)
--- a/lib/Text/Tradition/Collation.pm
+++ b/lib/Text/Tradition/Collation.pm
@@ -558,9 +558,13 @@ sub as_dot {
                my $label = $self->path_display_label( $self->path_witnesses( $edge ) );
                        my $variables = { %edge_attrs, 'label' => $label };
                        # Account for the rank gap if necessary
-                       my $rankgap = $self->reading( $edge->[1] )->rank 
+                       if( $self->reading( $edge->[1] )->has_rank 
+                               && $self->reading( $edge->[0] )->has_rank
+                               && $self->reading( $edge->[1] )->rank 
+                               - $self->reading( $edge->[0] )->rank > 1 ) {
+                               $variables->{'minlen'} = $self->reading( $edge->[1] )->rank 
                                - $self->reading( $edge->[0] )->rank;
-                       $variables->{'minlen'} = $rankgap if $rankgap > 1;
+                       }
                        my $varopts = _dot_attr_string( $variables );
                        $dot .= sprintf( "\t\"%s\" -> \"%s\" %s;\n", 
                                $edge->[0], $edge->[1], $varopts );
@@ -879,7 +883,7 @@ sub make_alignment_table {
                { 'witness' => $wit->sigil, 'tokens' => \@row } );
         if( $wit->is_layered ) {
                my @wit_ac_path = $self->reading_sequence( $self->start, $self->end, 
-                       $wit->sigil.$self->ac_label, $wit->sigil );
+                       $wit->sigil.$self->ac_label );
             my @ac_row = _make_witness_row( \@wit_ac_path, \@all_pos, $noderefs );
                        push( @{$table->{'alignment'}},
                                { 'witness' => $wit->sigil.$self->ac_label, 'tokens' => \@ac_row } );
@@ -954,9 +958,10 @@ used wherever no path exists for $sigil or $backup.
 =cut
 
 # TODO Think about returning some lazy-eval iterator.
+# TODO Get rid of backup; we should know from what witness is whether we need it.
 
 sub reading_sequence {
-    my( $self, $start, $end, $witness, $backup ) = @_;
+    my( $self, $start, $end, $witness ) = @_;
 
     $witness = $self->baselabel unless $witness;
     my @readings = ( $start );
@@ -969,7 +974,7 @@ sub reading_sequence {
         }
         $seen{$n->id} = 1;
         
-        my $next = $self->next_reading( $n, $witness, $backup );
+        my $next = $self->next_reading( $n, $witness );
         unless( $next ) {
             warn "Did not find any path for $witness from reading " . $n->id;
             last;
@@ -1016,7 +1021,15 @@ sub prior_reading {
 }
 
 sub _find_linked_reading {
-    my( $self, $direction, $node, $path, $alt_path ) = @_;
+    my( $self, $direction, $node, $path ) = @_;
+    
+    # Get a backup if we are dealing with a layered witness
+    my $alt_path;
+    my $aclabel = $self->ac_label;
+    if( $path && $path =~ /^(.*)\Q$aclabel\E$/ ) {
+       $alt_path = $1;
+    }
+    
     my @linked_paths = $direction eq 'next' 
         ? $self->sequence->edges_from( $node ) 
         : $self->sequence->edges_to( $node );
@@ -1072,8 +1085,25 @@ sub _witnesses_of_label {
     my $regex = $self->wit_list_separator;
     my @answer = split( /\Q$regex\E/, $label );
     return @answer;
-}    
+}
+
+=head2 path_text( $sigil, $mainsigil [, $start, $end ] )
+
+Returns the text of a witness (plus its backup, if we are using a layer)
+as stored in the collation.  The text is returned as a string, where the
+individual readings are joined with spaces and the meta-readings (e.g.
+lacunae) are omitted.  Optional specification of $start and $end allows
+the generation of a subset of the witness text.
+
+=cut
 
+sub path_text {
+       my( $self, $wit, $start, $end ) = @_;
+       $start = $self->start unless $start;
+       $end = $self->end unless $end;
+       my @path = grep { !$_->is_meta } $self->reading_sequence( $start, $end, $wit );
+       return join( ' ', map { $_->text } @path );
+}
 
 =head1 INITIALIZATION METHODS
 
@@ -1246,6 +1276,28 @@ sub flatten_ranks {
     }
 }
 
+=head2 text_from_paths
+
+Calculate the text array for all witnesses from the path, for later consistency
+checking.  Only to be used if there is no non-graph-based way to know the
+original texts.
+
+=cut
+
+sub text_from_paths {
+       my $self = shift;
+    foreach my $wit ( $self->tradition->witnesses ) {
+       my @text = split( /\s+/, 
+               $self->reading_sequence( $self->start, $self->end, $wit->sigil ) );
+       $wit->text( \@text );
+       if( $wit->is_layered ) {
+                       my @uctext = split( /\s+/, 
+                               $self->reading_sequence( $self->start, $self->end, 
+                                       $wit->sigil.$self->ac_label ) );
+                       $wit->text( \@uctext );
+       }
+    }    
+}
 
 =head1 UTILITY FUNCTIONS