import os
def load_data(dir):
list = []
for file in os.listdir(dir):
with open(dir + ’/’ + file, ’rb’) as f:
body = f.read().decode(’utf-8’, errors=’ignore’).splitlines() list.append(’ ’.join(body))
return list
BASE_DATA_DIR=’enron1’
ham = [(text, ’ham’) for text in load_data(BASE_DATA_DIR + ’/ham’)] spam = [(text, ’spam’) for text in load_data(BASE_DATA_DIR +
’/spam’)]
    2
all = ham + spam