[Slashcode-development] [PATCH] initialize Lingua::Stem only if stem_uncommon_words is on

[Oliver M. B. <ol...@fa...>]

Hi!

My site uses a language (Japanese) that is not supported by Lingua::Stem
and though I've switched off stemming, Slash::Utility:::Data::findWords()
chokes trying to initialize Lingua::Stem.

Below is a patch that makes the initialization of the stemmer and all
operations on it depend on the stem_uncommon_words var. Please review and
apply.

diff -u -r1.6 Data.pm
--- ./Slash/Utility/Data/Data.pm	31 Dec 2004 12:35:48 -0000	1.6
+++ ./Slash/Utility/Data/Data.pm	2 Jan 2005 10:51:07 -0000
@@ -3243,11 +3243,14 @@
 	my $use_stemming = $constants->{stem_uncommon_words};
 	my $language = $constants->{rdflanguage} || "EN-US";
 	$language = uc($language);
-	my $stemmer = Lingua::Stem->new(-locale => $language);
-	$stemmer->stem_caching({ -level => 2 });
+	my $stemmer;
 	my $text_return_hr = {};
 	my @word_stems;
 
+	if ($use_stemming){
+		$stemmer = Lingua::Stem->new(-locale => $language);
+		$stemmer->stem_caching({ -level => 2 });
+	}
 
 	# Return a hashref;  keys are the words, values are hashrefs
 	# with the number of times they appear and so on.
@@ -3348,7 +3351,7 @@
 			$wordcount->{$word}{count}++;
 		}
 	}
-	$stemmer->clear_stem_cache();
+	$stemmer->clear_stem_cache() if $use_stemming;
 
 	return $wordcount;
 }

-- 
	Oliver M. Bolzer

GPG (PGP) Fingerprint = 621B 52F6 2AC1 36DB 8761  018F 8786 87AD EF50 D1FF