import pymongo
import json
from pprint import pprint
import numpy
import scipy
from tokenizer import Tokenizer
import datetime

client = pymongo.MongoClient("localhost", 27017)
db = client.yelp
users_collection = db.users
reviews_collection = db.reviews
businesses_collection = db.businesses
restaurant_review_collection = db.restaurant_reviews
food_review_collection = db.food_reviews
restaurants_collection = db.restaurants
culinary_businesses_collection = db.culinary_businesses
print db.name

# GET USER, TEXT PAIRS
user_ids = list()
user_texts = list()
business_ids = list()
business_texts = list()

print "#####################"
print "# CALCULATING USERS #"
print "#####################"

t1 = datetime.datetime.now()
for user in users_collection.find().limit(10):
	u_id = user["user_id"]
        #print u_id
	reviews_current = restaurant_review_collection.find({"user_id":u_id})
	text = u_id # since some users have no reviews with text
	for review in reviews_current:
		text = text + " " + review["text"]
	user_ids.append(u_id)
	user_texts.append(text)
        #print "Review text length: " + str(len(text))
t2 = datetime.datetime.now()

print "Execution time: %s" % (t2-t1)

print "##########################"
print "# CALCULATING BUSINESSES #"
print "##########################"

t1 = datetime.datetime.now()
for restaurant in restaurants_collection.find().limit(100):
	b_id = restaurant["business_id"]
	print b_id
	reviews_current = restaurant_review_collection.find({"business_id":b_id})
	text = b_id # since some users have no reviews with text
	for review in reviews_current:
		text = text + " " + review["text"]
        text = text[0: min(len(text), 100000)]
	business_ids.append(b_id)
	business_texts.append(text)
        print "Review text length: " + str(len(text))

myTokenizer = Tokenizer(business_ids, business_texts)
t2 = datetime.datetime.now()

print "Execution time: %s" % (t2-t1)

#myTokenizer = Tokenizer(user_ids, user_texts)
#myTokenizer = Tokenizer(business_ids, business_texts)

#results = myTokenizer.getSimilarDocuments("KXJbnHT4PDS1JZNCFKdmMg", 5)
results = myTokenizer.getSimilarDocuments("qw5gR8vW7mSOK4VROSwdMA", 5)

#print results
print "USER 1-----------------------------------------"
print "qw5gR8vW7mSOK4VROSwdMA"
#print user_texts[user_ids.index("KXJbnHT4PDS1JZNCFKdmMg")]
#print business_texts[business_ids.index("qw5gR8vW7mSOK4VROSwdMA")]
print "USER 2-----------------------------------------"
#print user_texts[user_ids.index(results[0])]
#print business_texts[business_ids.index(results[0])]
print results[0]
print "USER 3-----------------------------------------"
print results[1]