Changeset 7880 for lang/perl/Acme-Shukugawa-Atom

Timestamp:

03/13/08 10:42:55 (9 months ago)

Author:

daisuke

Message:

lang/perl/Acme-Shukugawa-Atom; more tests, more atom

Location:

lang/perl/Acme-Shukugawa-Atom/trunk

Files:

• lib/Acme/Shukugawa/Atom.pm (modified) (5 diffs)
• t/01_basic.t (modified) (2 diffs)

lang/perl/Acme-Shukugawa-Atom/trunk/lib/Acme/Shukugawa/Atom.pm

@@ -22 +22 @@
 }
 
-sub preprocess
+# Special case handling -- this could be optimized further
+# put it in a sharefile later
+my (@SPECIAL, $EXCEPTION);
+BEGIN
 {
-    my ($self, $strref) = @_;
-
-    # Special case handling -- this could be optimized further
-    # put it in a sharefile later
-    my @special = (
+    @SPECIAL = (
+        '銀座' => 'ザギン',
         '別に' => 'ジリサワゴネタ',
         '予約した' => 'バミった',
@@ -37 +37 @@
         '(?:おおきい|大きい)(?:のか?|か)?' => 'カイデー',
     );
+    $EXCEPTION = decode_utf8(join("|",
+        map { $SPECIAL[$_ * 2 + 1] } (0..$#SPECIAL/2) ));
+}
 
-    for(0..$#special/2) {
-        my $pattern = $special[$_ * 2];
-        my $replace = $special[$_ * 2 + 1];
+sub preprocess
+{
+    my ($self, $strref) = @_;
+
+    for(0..$#SPECIAL/2) {
+        my $pattern = $SPECIAL[$_ * 2];
+        my $replace = $SPECIAL[$_ * 2 + 1];
         $$strref =~ s/$pattern/$replace/g;
     }
@@ -54 +61 @@
     # in hiragana
     my $ret = '';
-    foreach (my $node = $mecab->parse($$strref);
-        $node; $node = $node->next)
-    {
-        my $surface = decode_utf8($node->surface);
-        next unless $surface;
-        if ($surface =~ /^\p{InHiragana}+$/ || $surface =~ /^\p{InKatakana}+$/) {
-            $ret .= $surface;
-        } else {
-            my $feature = decode_utf8($node->feature);
 
-            if (my $yomi = (split(/,/, $feature))[8]) {
-                $ret .= $self->atomize($yomi) || $surface;
+    foreach my $text (split(/($EXCEPTION)/, $$strref)) {
+        if ($text =~ /$EXCEPTION/) {
+            $ret .= $text;
+            next;
+        }
+
+        foreach (my $node = $mecab->parse($text); $node; $node = $node->next) {
+            my $surface = decode_utf8($node->surface);
+            next unless $surface;
+            if ($surface =~ /^\p{InHiragana}+$/) {
+                $ret .= $surface;
             } else {
-                $ret .= $surface;
+                my $feature = decode_utf8($node->feature);
+
+                if (my $yomi = (split(/,/, $feature))[8]) {
+                    $ret .= $self->atomize($yomi) || $surface;
+                } else {
+                    $ret .= $surface;
+                }
             }
         }
@@ -81 +94 @@
 my $small    = decode_utf8("[ャュョッー]");
 my $syllable = decode_utf8("(?:.$small?)");
-my $nbar     = decode_utf8("[^ンー]");
+my $nbar     = decode_utf8("^ンー");
 sub apply_shisu_rule
 {
@@ -95 +108 @@
     my ($self, $yomi) = @_;
 
-    if ($yomi =~ s/^($syllable[ー]?)(..)$/$2$1/) {
+    if ($yomi =~ s/^(${syllable}[$nbar]?)([^$nbar].)$/$2$1/) {
         $yomi =~ s/([^ー])$/$1ー/;
         return $yomi;

lang/perl/Acme-Shukugawa-Atom/trunk/t/01_basic.t

@@ -1 +1 @@
 use strict;
 use utf8;
-use Test::More (tests => 4);
+use Test::More (tests => 7);
 
 BEGIN
@@ -11 +11 @@
     "六本木の胸の大きいお姉さんがいる店を予約した"
         => "ギロッポンのパイオツカイデーチャンネーがいるセーミーをバミった" ,
+    "ハワイ" => "ワイハー",
+    "寿司"   => "シースー",
+    "銀座" => "ザギン",
     "狼" => "カミオー",
+#     "鋏" => "サミハー", <- mecabの辞書にない？
     "おばあさんの口はどうして大きいの？" =>
         "チャンバーのチークーはどうしてカイデー？"