class Consumer(multiprocessing.Process):
def __init__(self, task_queue, host, port, db, collection):
self.coll = MongoClient(host, port)[db][collection]
self.mongo_bulk = self.coll.initialize_unordered_bulk_op()
multiprocessing.Process.__init__(self)
self.task_queue = task_queue
def run(self):
count = 0
def execute_mongo_bulk():
try:
self.mongo_bulk.execute()
self.mongo_bulk = self.coll.initialize_unordered_bulk_op()
logger.info("executing mongo bulk")
except Exception, e:
logger.info(e)
while True:
next_task = self.task_queue.get()
if next_task is None:
self.task_queue.task_done()
break
next_task(self.mongo_bulk)
if count % 1000 == 0:
execute_mongo_bulk()
self.task_queue.task_done()
execute_mongo_bulk()
return
def worker(identifier, skip, count):
stopwords = load_stopwords()
reviews_collection = MongoClient(Settings.MONGO_CONNECTION_STRING)[Settings.REVIEWS_DATABASE][
Settings.REVIEWS_COLLECTION]
tags_collection = MongoClient(Settings.MONGO_CONNECTION_STRING)[Settings.TAGS_DATABASE][
Settings.REVIEWS_COLLECTION]
# Tags creation and storage is insert heavey operation , so Adding data in Bulk
print tags_collection
bulk = tags_collection.initialize_unordered_bulk_op();
bulkCounter = 0
counter = 0
batch_size = 1000
for batch in range(0, count, batch_size):
reviews_cursor = reviews_collection.find().skip(skip + batch).limit(batch_size)
for review in reviews_cursor:
words = []
sentences = nltk.sent_tokenize(review["text"].lower())
for sentence in sentences:
tokens = nltk.word_tokenize(sentence)
text = [word for word in tokens if word not in stopwords]
tagged_text = nltk.pos_tag(text)
for word, tag in tagged_text:
words.append({"word": word, "pos": tag})
bulk.insert({
"rating" : review ["rating"],
"userId" : review ["userId"],
"reviewId": review["reviewId"],
"business": review["business"],
"text": review["text"],
"words": words
})
bulkCounter=bulkCounter + 1
counter = counter + 1
if bulkCounter % 10000 == 0 :
bulk.execute()
print str(counter) + 'Entries are inserted in Thread' + str(multiprocessing.current_process())
bulkCounter=0
bulk= tags_collection.initialize_unordered_bulk_op()
if bulkCounter > 0 :
bulk.execute()
from pymongo import MongoClient
from nltk.stem.wordnet import WordNetLemmatizer
from settings import Settings
tags_collection = MongoClient(Settings.MONGO_CONNECTION_STRING)[Settings.TAGS_DATABASE][Settings.REVIEWS_COLLECTION]
corpus_collection = MongoClient(Settings.MONGO_CONNECTION_STRING)[Settings.TAGS_DATABASE][Settings.CORPUS_COLLECTION]
reviews_cursor = tags_collection.find()
reviewsCount = reviews_cursor.count()
reviews_cursor.batch_size(5000)
lem = WordNetLemmatizer()
## Do Bulk insertion
bulk = corpus_collection.initialize_unordered_bulk_op()
bulkCounter=0
counter=0
for review in reviews_cursor:
nouns = []
words = [word for word in review["words"] if word["pos"] in ["NN", "NNS"]]
for word in words:
nouns.append(lem.lemmatize(word["word"]))
bulk.insert({
"userId" : review["userId"],
"rating" : review["rating"],
"reviewId": review["reviewId"],
"business": review["business"],
"text": review["text"],
import json
import sys
sys.path.insert(0, '/home/ish/DataScience/yelp_topic')
from pymongo import MongoClient
from settings import Settings
dataset_file = Settings.BUSINESS_DATASET_FILE
reviews_collection = MongoClient(Settings.MONGO_CONNECTION_STRING)[Settings.REVIEWS_DATABASE][
Settings.BUSINESS_COLLECTION]
# Bulk insertion is used to make insertion faster
bulk= reviews_collection.initialize_unordered_bulk_op()
bulkCounter=0
counter=0
#Read each line and insert data in MongoDB in Json format
with open(dataset_file) as dataset:
next(dataset)
for line in dataset:
try:
data = json.loads(line)
except ValueError:
print 'Oops!'
if data["type"] == "business":
bulk.insert({
"business_id":data["business_id"],
"name": data["name"],
"neighborhoods": data["neighborhoods"],
"full_address": data["full_address"],
reviews_collection = MongoClient(Settings.MONGO_CONNECTION_STRING)[Settings.REVIEWS_DATABASE][
Settings.REVIEWS_COLLECTION]
tags_collection = MongoClient(Settings.MONGO_CONNECTION_STRING)[Settings.TAGS_DATABASE][Settings.REVIEWS_COLLECTION]
reviews_cursor = reviews_collection.find()
reviewsCount = reviews_cursor.count()
reviews_cursor.batch_size(1000)
stopwords = {}
with open('stopwords.txt', 'rU') as f:
for line in f:
stopwords[line.strip()] = 1
bulk = tags_collection.initialize_unordered_bulk_op()
bulkCounter=0
for review in reviews_cursor:
words = []
sentences = nltk.sent_tokenize(review["text"].lower())
for sentence in sentences:
tokens = nltk.word_tokenize(sentence)
text = [word for word in tokens if word not in stopwords]
tagged_text = nltk.pos_tag(text)
for word, tag in tagged_text:
words.append({"word": word, "pos": tag})
bulk.insert({
