Add files via upload

Author: anjalysam

mapper.py

@@ -0,0 +1,33 @@
+# -*- coding: utf-8 -*-
+"""mapper.ipynb
+
+Automatically generated by Colaboratory.
+
+Original file is located at
+    https://colab.research.google.com/drive/1yCwGyMXJT2qt3_58aLOOiJXO0GIaPcJd
+"""
+
+
+
+import sys
+import io
+import re
+import nltk
+nltk.download('stopwords',quiet=True)
+from nltk.corpus import stopwords
+punctuations = '''!()-[]{};:'"\,<>./?@#$%^&*_~'''
+
+stop_words = set(stopwords.words('english'))
+input_stream = io.TextIOWrapper(sys.stdin.buffer, encoding='latin1')
+for line in input_stream:
+  line = line.strip()
+  line = re.sub(r'[^\w\s]', '',line)
+  line = line.lower()
+  for x in line:
+    if x in punctuations:
+      line=line.replace(x, " ") 
+
+  words=line.split()
+  for word in words: 
+    if word not in stop_words:
+      print('%s\t%s' % (word, 1))