From: Tara L Andrews <tla@mit.edu>
Date: Sun, 24 Jun 2012 12:37:09 +0000 (+0200)
Subject: make normal_form default to ->text
X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=367e901bd7644e4f197cb157380cfa7dff12ae41;p=scpubgit%2Fstemmatology.git

make normal_form default to ->text
---

diff --git a/lib/Text/Tradition/Collation/Reading.pm b/lib/Text/Tradition/Collation/Reading.pm
index 108548b..2d4322e 100644
--- a/lib/Text/Tradition/Collation/Reading.pm
+++ b/lib/Text/Tradition/Collation/Reading.pm
@@ -159,7 +159,8 @@ has 'is_nonsense' => (
 has 'normal_form' => (
 	is => 'rw',
 	isa => 'Str',
-	predicate => 'has_normal_form',
+	predicate => '_has_normal_form',
+	clearer => '_clear_normal_form',
 	);
 
 # Holds the lexemes for the reading.
@@ -236,6 +237,21 @@ sub BUILD {
 	}
 }
 
+# Make normal_form default to text, transparently.
+around 'normal_form' => sub {
+	my $orig = shift;
+	my $self = shift;
+	my( $arg ) = @_;
+	if( $arg && $arg eq $self->text ) {
+		$self->_clear_normal_form;
+		return $arg;
+	} elsif( !$arg && !$self->_has_normal_form ) {
+		return $self->text;
+	} else {
+		$self->$orig( @_ );
+	}
+};
+
 =head2 is_meta
 
 A meta attribute (ha ha), which should be true if any of our 'special'
diff --git a/lib/Text/Tradition/Language/Base.pm b/lib/Text/Tradition/Language/Base.pm
index d0d1caf..f0fe304 100644
--- a/lib/Text/Tradition/Language/Base.pm
+++ b/lib/Text/Tradition/Language/Base.pm
@@ -113,8 +113,7 @@ sub _lemmatize_treetagger_sequence {
 				warn "Ran out of readings in sequence at $lexeme";
 				last;
 			}
-			my $curr_rdg_text = $curr_rdg->has_normal_form 
-				? $curr_rdg->normal_form : $curr_rdg->text;
+			my $curr_rdg_text = $curr_rdg->normal_form;
 			if( $unused_rdg_part &&
 				$unused_rdg_part =~ /^\Q$lexeme\E(\s*)(.*)$/ ) {
 				# Nth part of curr_rdg
@@ -134,7 +133,7 @@ sub _lemmatize_treetagger_sequence {
 				my @lookahead;
 				my $matched;
 				while( my $nr = shift @path ) {
-					my $nrtext = $nr->has_normal_form ? $nr->normal_form : $nr->text;
+					my $nrtext = $nr->normal_form;
 					if( $nrtext =~ /^\Q$lexeme\E/ ) {
 						$curr_rdg = $lookahead[-1] if @lookahead;
 						$matched = 1;
@@ -209,8 +208,7 @@ sub _text_from_path {
 		unless ( $r->join_prior || !$last || $last->join_next ) {
 			$pathtext .= ' ';
 		} 
-		$pathtext .= ( $normalize && $r->has_normal_form ) 
-			? $r->normal_form : $r->text;
+		$pathtext .= $normalize ? $r->normal_form : $r->text;
 		$last = $r;
 	}
 	return $pathtext;