Replace Zenkaku-Space to Zenkaku-Underscore

Remove disallowed UTF-8 whitespace character
kaldi-asr · Oct 11, 2017 · e8102c2 · e8102c2
1 parent 5fbdc66
commit e8102c2
Showing 1 changed file with 5 additions and 0 deletions.
diff --git a/egs/csj/s5/local/csj_make_trans/csj2kaldi4m.pl b/egs/csj/s5/local/csj_make_trans/csj2kaldi4m.pl
@@ -245,6 +245,11 @@
     $morph =~ s/\/\//\//g;
     $morph =~ s/\/\//\//g;
     $morph =~ s/\/$//g;
+    # Replace Zenkaku-Space to Zenkaku-Underscore
+    # Input: っしゃっ+動詞/ラ行五段/連用形/促音便　省略 r a q sh a q
+    # ->
+    # Output: っしゃっ+動詞/ラ行五段/連用形/促音便＿省略 r a q sh a q
+    $morph =~ s/　/＿/g;
 
     unless( $word =~ /skip_word/){
 	if ($word && $pos){