Sync. Dictionaries/pruned.

DiceHoldingsInc · Jun 7, 2011 · 402f91f · 402f91f
1 parent 819183b
commit 402f91f
Show file tree

Hide file tree

Showing 6 changed files with 245,984 additions and 0 deletions.
diff --git a/.gitignore b/.gitignore
@@ -0,0 +1 @@
+/bin
diff --git a/src/comp-ccorpus.py b/src/comp-ccorpus.py
@@ -0,0 +1,13 @@
+import sys,re
+
+words = {}
+for line in open('ccorpus.txt','r'):
+    w = line.split()[0]
+    words[w] = True
+
+print("Read input.")
+
+for line in open('combined.sorted','r'):
+    w, r = line.split()
+    if words.has_key(w):
+        print w + "\t" + r