NLTK: add all files in a directory to a corpus

# NLTK code for building a corpus of Twitter messages (or any number of text files in a dir)

import glob, os
path = 'E:\Corpora\Twitter\plaintext\mostfrequent'
for infile in glob.glob (os.path.join(path, '*.txt') ):
f = open(infile)
raw = raw + ' ' + f.read()

tokens = nltk.word_tokenize(raw)
text = nltk.Text(tokens)
len(text)