diff options
| author | Chris Dyer <cdyer@allegro.clab.cs.cmu.edu> | 2013-01-21 17:28:19 -0500 |
|---|---|---|
| committer | Chris Dyer <cdyer@allegro.clab.cs.cmu.edu> | 2013-01-21 17:28:19 -0500 |
| commit | ee681f3d0337dc3d93c25de373a50fbda69252fe (patch) | |
| tree | 66a89cdc31969aa36440bb7061fbe39f2ffdd4f6 | |
| parent | e1f71a6ce868d116f04082b697a8d61afcd625f1 (diff) | |
tokenizer support for utf8 patterns
| -rw-r--r-- | corpus/support/token_patterns | 1 | ||||
| -rwxr-xr-x | corpus/support/tokenizer.pl | 2 |
2 files changed, 3 insertions, 0 deletions
diff --git a/corpus/support/token_patterns b/corpus/support/token_patterns index 8e69432b..b25ac6de 100644 --- a/corpus/support/token_patterns +++ b/corpus/support/token_patterns @@ -1,3 +1,4 @@ /^(al|el|ul|e)\-[a-z]+$/ +/^((а|А)(ль|ш)|уль)-\p{Cyrillic}+$/ /^(\d|\d\d|\d\d\d)\.$/ diff --git a/corpus/support/tokenizer.pl b/corpus/support/tokenizer.pl index b5190858..0350a894 100755 --- a/corpus/support/tokenizer.pl +++ b/corpus/support/tokenizer.pl @@ -73,6 +73,7 @@ my $dict_file = "$workdir/token_list"; my $word_patt_file = "$workdir/token_patterns"; open(my $dict_fp, "$dict_file") or die; +binmode($dict_fp, ":utf8"); # read in the list of words that should not be segmented, ## e.g.,"I.B.M.", co-operation. @@ -89,6 +90,7 @@ while(<$dict_fp>){ } open(my $patt_fp, "$word_patt_file") or die; +binmode($patt_fp, ":utf8"); my @word_patts = (); my $word_patt_num = 0; while(<$patt_fp>){ |