From: Don Armstrong <don@donarmstrong.com>
Date: Fri, 24 Jun 2011 17:59:33 +0000 (+0000)
Subject:  * tweak utf8 issues
X-Git-Url: https://git.donarmstrong.com/?a=commitdiff_plain;h=1ce8a201f7e91a14f5dcaa1616d2f2e94afa6efd;p=bin.git

 * tweak utf8 issues
---

diff --git a/anamang b/anamang
index e14f213..81841ec 100755
--- a/anamang
+++ b/anamang
@@ -273,8 +273,9 @@ sub update_and_load_database {
 	while (<$wordlist_fh>) {
 	    chomp;
 	    next unless length $_;
-	    my $word = word_sanitize($_);
-	    $word =~ s/[^a-z]//;
+	    utf8::upgrade($_);
+	    my ($word) = word_sanitize($_);
+	    $word =~ s/[^a-z]//g;
 	    next unless length $_;
 	    next if exists $seen_words{$word};
 	    $seen_words{$word} = 1;