summaryrefslogtreecommitdiff
path: root/corpus/support/quote-norm.pl
diff options
context:
space:
mode:
authorChris Dyer <cdyer@allegro.clab.cs.cmu.edu>2014-03-10 18:40:13 -0400
committerChris Dyer <cdyer@allegro.clab.cs.cmu.edu>2014-03-10 18:40:13 -0400
commit284383880f043edb2d67afbe2f64237c466245c1 (patch)
tree2751eaccdfc82586cf8aad1cfe257dbc3cf27436 /corpus/support/quote-norm.pl
parent8c2d8217f068d8f107f95171496a25013d4e35fe (diff)
few tokenization bugs
Diffstat (limited to 'corpus/support/quote-norm.pl')
-rwxr-xr-xcorpus/support/quote-norm.pl1
1 files changed, 1 insertions, 0 deletions
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"