From c81b0dc240f2233c3e5ecccd8982218115476f9a Mon Sep 17 00:00:00 2001
From: Chris Dyer <cdyer@allegro.clab.cs.cmu.edu>
Date: Tue, 4 Dec 2012 22:01:32 -0500
Subject: more flexible corpus cutting

---
 corpus/cut-corpus.pl | 29 ++++++++++++++++++++++++-----
 1 file changed, 24 insertions(+), 5 deletions(-)

(limited to 'corpus')

diff --git a/corpus/cut-corpus.pl b/corpus/cut-corpus.pl
index fc9cce3b..7daac0e2 100755
--- a/corpus/cut-corpus.pl
+++ b/corpus/cut-corpus.pl
@@ -3,14 +3,33 @@ use strict;
 die "Usage: $0 N\nSplits a corpus separated by ||| symbols and returns the Nth field\n" unless scalar @ARGV > 0;
 
 my $x = shift @ARGV;
-die "N must be numeric" unless $x =~ /^\d+$/;
-$x--;
+my @ind = split /,/, $x;
+my @o = ();
+for my $ff (@ind) {
+  if ($ff =~ /^\d+$/) {
+    push @o, $ff - 1;
+  } elsif ($ff =~ /^(\d+)-(\d+)$/) {
+    my $a = $1;
+    my $b = $2;
+    die "$a-$b is a bad range in input: $x\n" unless $b > $a;
+    for (my $i=$a; $i <= $b; $i++) {
+      push @o, $i - 1;
+    }
+  } else {
+    die "Bad input: $x\n";
+  }
+}
 
 while(<>) {
   chomp;
   my @fields = split / \|\|\| /;
-  my $y = $fields[$x];
-  if (!defined $y) { $y= ''; }
-  print "$y\n";
+  my @sf;
+  for my $i (@o) {
+    my $y = $fields[$i];
+    if (!defined $y) { $y= ''; }
+    push @sf, $y;
+  }
+  print join(' ||| ', @sf) . "\n";
 }
 
+
-- 
cgit v1.2.3