import os
import csv
#os.chdir('/path/to/wherever/you/downloaded/data/from/textcleaning')
os.chdir('/Users/rweiss/Dropbox/presentations/MozFest2013/data/')

review_data = []
labels = set()
tsvfile = open('../sentiment_examples/reviews_sample.csv', 'r') # take a look at what this .csv looks like and check the file structure
csv_reader = csv.DictReader(tsvfile, delimiter='\t')
for line in csv_reader:
 temp = {line['label'] : line['content']}
 review_data.append(temp)
 labels.add(line['label'])
tsvfile.close()

print 'There are ' + str(len(review_data)) + ' total reviews.'

print 'The labels are '+ ', '.join(labels) + '.'

from random import shuffle
x = [review_data[i] for i in range(len(review_data))]
shuffle(x)

#shuffling just to show you how to

review_data = [''.join(el.values()) for el in x] 
target_labels = [''.join(el.keys()) for el in x]

#!/usr/bin/python
# code adapted from the SASA script classifyFromCmdLine.py
import sys

try:
 from sasa.classifier import Classifier
except ImportError:
 raise ImportError("Did you try to run '. setup.env'?\n" + 
 "(or add the sasa-tool directory to PYTHONPATH, ie export PYTHONPATH=<this director>)?")

classifier = Classifier() 
classified_reviews = []

for line in review_data:
 print "classifying %s" % line.strip()
 sentiment, valence, posterior = classifier.classifyFromText(line)
 if valence >= 0:
 score = "positive"
 elif valence <0:
 score = "negative"
 classified_reviews.append({score: line}) 

outfile = open('../sentiment_examples/reviews_sample_3.csv', 'w')

for line in classified_reviews:
 body = ''.join(line.values())
 label = ''.join(line.keys())
 body = body.decode('iso-8859-1') # this might not be the encoding of your data!
 body = body.encode('utf-8') # utf-8 is usually the best encoding to use
 label = label.encode('utf-8')
 try:
 outfile.write(label + '\t' + body + '\n')
 except UnicodeDecodeError:
 print "Unicode error" + line
outfile.close()