From f45693e7f06a60eaf273698550e613e33eb8b212 Mon Sep 17 00:00:00 2001
From: TomV <tom@viner.tv>
Date: Thu, 7 Jul 2011 20:45:42 +0100
Subject: [PATCH] initial parsing with nltk

---
 parse.py | 15 +++++++++++++++
 1 file changed, 15 insertions(+)
 create mode 100644 parse.py

diff --git a/parse.py b/parse.py
new file mode 100644
index 0000000..d44eb98
--- /dev/null
+++ b/parse.py
@@ -0,0 +1,15 @@
+import nltk
+
+def get_parts(text):
+    text = nltk.word_tokenize(text)
+    parts = nltk.pos_tag(text)
+    dic = {}
+    for work, part in parts:
+        if part in dic:
+            dic[part].append(work)
+        else:
+            dic[part] = [work]
+    return dic
+
+get_parts("hi how are you")
+