共通関数をまとめた perl モジュール ODIC.pm を追加した。

diff --git a/script/ODIC.pm b/script/ODIC.pm
new file mode 100644 (file)
index 0000000..d9cef0d
--- /dev/null
+++ b/script/ODIC.pm
@@ -0,0 +1,93 @@
+#
+# 沖縄辞書フォーマット共通ライブラリー
+#
+
+package ODIC;
+
+use strict;
+use Encode;
+
+our $MAX_PHONATE = 40;
+our $MAX_WORD    = 64;
+
+sub check_phonate {
+    my $phonate = shift;
+
+    if (length(Encode::decode('utf-8', $phonate)) > $MAX_PHONATE) {
+       print STDERR "Warning: $.: too long phonate `$phonate'\n";
+    }
+    if ($phonate =~ /[^あいうえおかきくけこさしすせそたちつてとなにぬねのはひふへほまみむめもらりるれろがぎぐげござじずぜぞだぢづでどばびぶべぼぁぃぅぇぉっょゃゅゎぱぴぷぺぽやゆよわをんヴー]/) {
+       print STDERR "Warning: $.: ilegal character in `$phonate'\n";
+    }
+}
+
+sub check_word {
+    my $word = shift;
+
+    if (length(Encode::decode('utf-8', $word)) > $MAX_WORD) {
+       print STDERR "Warning: $.: too long word `$word'\n";
+    }
+    if ($word =~ /[ \t",#]/) {
+       print STDERR "Warning: $.: ilegal character in `$word'\n";
+    }
+}
+
+sub to_eucjp {
+    my $utf8_string = Encode::decode('utf-8', shift);
+    return Encode::encode('euc-jp', $utf8_string);
+}
+
+sub to_shiftjis {
+    my $utf8_string = Encode::decode('utf-8', shift);
+    # 'cp932' は、「～」の変換でおかしくなるので使用しない。
+    return Encode::encode('shift_jis', $utf8_string);
+}
+
+1;
+__END__
+
+=head1 NAME
+
+ODIC - 沖縄辞書 <http://www.zukeran.org/o-dic/> フォーマット向けの共通関数
+
+=head1 EXAMPLE
+
+  require 'ODIC.pm';
+
+  while (<>) {
+    next if (/^\s*$|^\s*\#.*$/);       # 空行・コメントのみの行を読み飛ばす
+
+    if (/^(\S+)\s+(\S+)\s+(\S+)\s+#\s*([[:^cntrl:]]*).*$/) {
+      my $phonate = $1;        # 読み
+      my $word    = $2;        # 単語
+      my $class   = $3;        # 品詞
+      my $comment = $4;        # コメント
+      ODIC::check_phonate($phonate);
+      ODIC::check_word($word);
+      &convert_class;
+
+    } elsif (/^(\S+)\s+(\S+)\s+(\S+)/) {
+      my $phonate = $1;        # 読み
+      my $word    = $2;        # 単語
+      my $class   = $3;        # 品詞
+      my $comment = '';        # コメント
+      ODIC::check_phonate($phonate);
+      ODIC::check_word($word);
+      &convert_class;
+
+    } else {
+      print STDERR "Error: $.: too few field number `$_'\n";
+      print  "$_";
+    }
+  }
+  exit 0;
+
+  sub convert_class {
+    print "$phonate\t$word\t$class\t$comment\n";
+  }
+
+=head1 LICENSE
+
+Public domain.
+
+=cut