From de8ffd4598d6c1e45273b50642870a661b4bcad4 Mon Sep 17 00:00:00 2001
From: Chris Dyer <cdyer@allegro.clab.cs.cmu.edu>
Date: Mon, 10 Mar 2014 18:40:13 -0400
Subject: few tokenization bugs

---
 corpus/support/quote-norm.pl | 1 +
 corpus/support/token_list    | 6 ++++--
 2 files changed, 5 insertions(+), 2 deletions(-)

(limited to 'corpus/support')

diff --git a/corpus/support/quote-norm.pl b/corpus/support/quote-norm.pl
index 33604027..0366fad5 100755
--- a/corpus/support/quote-norm.pl
+++ b/corpus/support/quote-norm.pl
@@ -39,6 +39,7 @@ while(<STDIN>) {
   s/&\#([0-9]+);/pack("U", $1)/ge;
 
   # Regularlize spaces:
+  s/\x{ad}//g;        # soft hyphen
   s/\x{a0}/ /g;       # non-breaking space
   s/\x{2009}/ /g;     # thin space
   s/\x{2028}/ /g;     # "line separator"
diff --git a/corpus/support/token_list b/corpus/support/token_list
index 228663f6..d38638cf 100644
--- a/corpus/support/token_list
+++ b/corpus/support/token_list
@@ -156,8 +156,9 @@ Mass.
 Md.
 Mfg.
 Mgr.
-Mexican-U.S.
-Mich.
+Mio.
+Mrd.
+Bio.
 Minn.
 Mo.
 Mon.
@@ -187,6 +188,7 @@ Rd.
 Rev.
 R.J.
 C.L
+Rs.
 Rte.
 Sat.
 W.T
-- 
cgit v1.2.3