summaryrefslogtreecommitdiff
path: root/sa-extract/escape-testset.pl
diff options
context:
space:
mode:
authorKenneth Heafield <github@kheafield.com>2012-08-03 07:46:54 -0400
committerKenneth Heafield <github@kheafield.com>2012-08-03 07:46:54 -0400
commitbe1ab0a8937f9c5668ea5e6c31b798e87672e55e (patch)
treea13aad60ab6cced213401bce6a38ac885ba171ba /sa-extract/escape-testset.pl
parente5d6f4ae41009c26978ecd62668501af9762b0bc (diff)
parent9fe0219562e5db25171cce8776381600ff9a5649 (diff)
Merge branch 'master' of github.com:redpony/cdec
Diffstat (limited to 'sa-extract/escape-testset.pl')
-rwxr-xr-xsa-extract/escape-testset.pl35
1 files changed, 0 insertions, 35 deletions
diff --git a/sa-extract/escape-testset.pl b/sa-extract/escape-testset.pl
deleted file mode 100755
index 02fd7445..00000000
--- a/sa-extract/escape-testset.pl
+++ /dev/null
@@ -1,35 +0,0 @@
-#!/usr/bin/perl -w
-
-use utf8;
-use strict;
-
-binmode(STDIN,":utf8");
-binmode(STDOUT,":utf8");
-
-my @fh = ();
-if (scalar @ARGV == 0) {
- push @fh, \*STDIN;
-} else {
- for my $file (@ARGV) {
- my $f;
- open $f, "<$file" or die "Can't read $file: $!\n";
- binmode $f, ":utf8";
- push @fh, $f;
- }
-}
-
-my $id = -1;
-for my $f (@fh) {
- while(<$f>) {
- chomp;
- die "Empty line in test set" if /^\s*$/;
- die "Please remove <seg> tags from input:\n$_" if /^\s*<seg/i;
- $id++;
- s/&/\&amp;/g;
- s/</\&lt;/g;
- s/>/\&gt;/g;
- print "<seg id=\"$id\"> $_ </seg>\n";
- }
-}
-
-