New script for mapping corpus files to classes

git-svn-id: https://ws10smt.googlecode.com/svn/trunk@117 ec762483-ff6d-05da-a07a-a48fb63a330f

1 files changed, 27 insertions, 0 deletions

diff --git a/gi/pyp-topics/scripts/tokens2classes.py b/gi/pyp-topics/scripts/tokens2classes.py
new file mode 100755
index 00000000..33df255f
--- /dev/null
+++ b/gi/pyp-topics/scripts/tokens2classes.py
@@ -0,0 +1,27 @@
+#!/usr/bin/python
+
+import sys
+
+if len(sys.argv) != 3:
+  print "Usage: tokens2classes.py source_classes target_classes"
+  exit(1)
+
+source_to_topics = {}
+for line in open(sys.argv[1],'r'):
+  term,cls = line.split()
+  source_to_topics[term] = cls
+
+target_to_topics = {}
+for line in open(sys.argv[2],'r'):
+  term,cls = line.split()
+  target_to_topics[term] = cls
+
+for line in sys.stdin:
+  source, target, tail = line.split(" ||| ")
+
+  for token in source.split():
+    print source_to_topics[token],
+  print "|||",
+  for token in target.split():
+    print target_to_topics[token],
+  print "|||", tail,