summaryrefslogtreecommitdiff
path: root/word-aligner/ortho-norm
diff options
context:
space:
mode:
authorKenneth Heafield <github@kheafield.com>2011-11-03 19:53:08 +0000
committerKenneth Heafield <github@kheafield.com>2011-11-03 19:53:08 +0000
commitca5b590478a2cb1c23a706b76e6a881ce8701716 (patch)
tree3faea0090f9a91b52ea2d4c47cfa3b4eb04d15f8 /word-aligner/ortho-norm
parent437ee131bc4c2574768a187bf29d62dbb715abf4 (diff)
parenta769e4964db2443ce165043095e18dfc0d788910 (diff)
Merge branch 'master' of github.com:redpony/cdec
Diffstat (limited to 'word-aligner/ortho-norm')
-rwxr-xr-xword-aligner/ortho-norm/mg.pl13
-rwxr-xr-xword-aligner/ortho-norm/rw.pl13
2 files changed, 26 insertions, 0 deletions
diff --git a/word-aligner/ortho-norm/mg.pl b/word-aligner/ortho-norm/mg.pl
new file mode 100755
index 00000000..4cb0e8e7
--- /dev/null
+++ b/word-aligner/ortho-norm/mg.pl
@@ -0,0 +1,13 @@
+#!/usr/bin/perl -w
+use strict;
+use utf8;
+
+binmode(STDIN, ":utf8");
+binmode(STDOUT, ":utf8");
+
+while(<STDIN>) {
+ $_ = lc $_;
+ s/([a-z])'( |$)/$1$2/g;
+ print;
+}
+
diff --git a/word-aligner/ortho-norm/rw.pl b/word-aligner/ortho-norm/rw.pl
new file mode 100755
index 00000000..4cb0e8e7
--- /dev/null
+++ b/word-aligner/ortho-norm/rw.pl
@@ -0,0 +1,13 @@
+#!/usr/bin/perl -w
+use strict;
+use utf8;
+
+binmode(STDIN, ":utf8");
+binmode(STDOUT, ":utf8");
+
+while(<STDIN>) {
+ $_ = lc $_;
+ s/([a-z])'( |$)/$1$2/g;
+ print;
+}
+