| 13829 |
kshitij.so |
1 |
import urllib2
|
|
|
2 |
import simplejson as json
|
|
|
3 |
import pymongo
|
|
|
4 |
from dtr.utils.utils import to_java_date
|
|
|
5 |
from datetime import datetime
|
|
|
6 |
from operator import itemgetter
|
|
|
7 |
from shop2020.model.v1.catalog.script import FlipkartScraper
|
|
|
8 |
from dtr.utils.AmazonPriceOnlyScraper import AmazonScraper
|
|
|
9 |
|
|
|
10 |
con = None
|
|
|
11 |
now = datetime.now()
|
|
|
12 |
SOURCE_MAP = {'AMAZON':1,'FLIPKART':2,'SNAPDEAL':3}
|
|
|
13 |
scraperFk = FlipkartScraper.FlipkartScraper()
|
|
|
14 |
scraperAmazon = AmazonScraper()
|
|
|
15 |
|
|
|
16 |
headers = {
|
|
|
17 |
'User-agent':'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/535.11 (KHTML, like Gecko) Chrome/17.0.963.56 Safari/535.11',
|
|
|
18 |
'Accept' : 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
|
|
|
19 |
'Accept-Language' : 'en-US,en;q=0.8',
|
|
|
20 |
'Accept-Charset' : 'ISO-8859-1,utf-8;q=0.7,*;q=0.3'
|
|
|
21 |
}
|
|
|
22 |
|
|
|
23 |
def get_mongo_connection(host='localhost', port=27017):
|
|
|
24 |
global con
|
|
|
25 |
if con is None:
|
|
|
26 |
print "Establishing connection %s host and port %d" %(host,port)
|
|
|
27 |
try:
|
|
|
28 |
con = pymongo.MongoClient(host, port)
|
|
|
29 |
except Exception, e:
|
|
|
30 |
print e
|
|
|
31 |
return None
|
|
|
32 |
return con
|
|
|
33 |
|
|
|
34 |
def returnLatestPrice(data, source_id):
|
|
|
35 |
if source_id == 1:
|
|
|
36 |
try:
|
|
|
37 |
if data['identifier'] is None or len(data['identifier'])==0:
|
|
|
38 |
return {}
|
|
|
39 |
url = "http://www.amazon.in/gp/offer-listing/%s/ref=olp_sort_ps"%(data['identifier'])
|
|
|
40 |
lowestPrice = 0.0
|
|
|
41 |
lowestPrice = scraperAmazon.read(url)
|
|
|
42 |
inStock = 0
|
|
|
43 |
if lowestPrice > 0:
|
|
|
44 |
inStock = 1
|
|
|
45 |
if lowestPrice > 0:
|
|
|
46 |
get_mongo_connection().Catalog.MasterData.update({'identifier':data['identifier'],'source_id':source_id}, {'$set' : {'available_price':lowestPrice,'updatedOn':to_java_date(now),'in_stock':inStock}}, multi=True)
|
|
|
47 |
else:
|
|
|
48 |
pass
|
|
|
49 |
get_mongo_connection().Catalog.MasterData.update({'identifier':data['identifier'],'source_id':source_id}, {'$set' : {'updatedOn':to_java_date(now),'in_stock':0}}, multi=True)
|
|
|
50 |
return {'_id':data['_id'],'available_price':lowestPrice,'in_stock':inStock,'source_id':1,'source_product_name':data['source_product_name'],'marketPlaceUrl':data['marketPlaceUrl'],'thumbnail':data['thumbnail']}
|
|
|
51 |
except:
|
|
|
52 |
return {'_id':data['_id'],'available_price':data['available_price'],'in_stock':data['inStock'],'source_id':1,'source_product_name':data['source_product_name'],'marketPlaceUrl':data['marketPlaceUrl'],'thumbnail':data['thumbnail']}
|
|
|
53 |
|
|
|
54 |
elif source_id ==3:
|
|
|
55 |
try:
|
|
|
56 |
if data['identifier'] is None or len(data['identifier'])==0:
|
|
|
57 |
return {}
|
|
|
58 |
url="http://www.snapdeal.com/acors/json/gvbps?supc=%s&catId=175&sort=sellingPrice"%(data['identifier'])
|
|
|
59 |
req = urllib2.Request(url,headers=headers)
|
|
|
60 |
response = urllib2.urlopen(req)
|
|
|
61 |
json_input = response.read()
|
|
|
62 |
vendorInfo = json.loads(json_input)
|
|
|
63 |
lowestOfferPrice = 0
|
|
|
64 |
inStock = 0
|
|
|
65 |
for vendor in vendorInfo:
|
|
|
66 |
lowestOfferPrice = float(vendor['sellingPrice'])
|
|
|
67 |
stock = vendor['buyableInventory']
|
|
|
68 |
if stock > 0 and lowestOfferPrice > 0:
|
|
|
69 |
inStock = 1
|
|
|
70 |
break
|
|
|
71 |
|
|
|
72 |
print lowestOfferPrice
|
|
|
73 |
print inStock
|
|
|
74 |
print stock
|
|
|
75 |
print "*************"
|
|
|
76 |
if inStock == 1:
|
|
|
77 |
pass
|
|
|
78 |
#get_mongo_connection().Catalog.MasterData.update({'identifier':data['identifier']}, {'$set' : {'available_price':lowestOfferPrice,'updatedOn':to_java_date(now),'in_stock':inStock}}, multi=True)
|
|
|
79 |
else:
|
|
|
80 |
pass
|
|
|
81 |
#get_mongo_connection().Catalog.MasterData.update({'identifier':data['identifier'],'source_id':3}, {'$set' : {'updatedOn':to_java_date(now),'in_stock':inStock}}, multi=True)
|
|
|
82 |
return {'_id':data['_id'],'available_price':lowestOfferPrice,'in_stock':inStock,'source_id':3,'source_product_name':data['source_product_name'],'marketPlaceUrl':data['marketPlaceUrl'],'thumbnail':data['thumbnail']}
|
|
|
83 |
except:
|
|
|
84 |
return {'_id':data['_id'],'available_price':data['available_price'],'in_stock':data['inStock'],'source_id':3,'source_product_name':data['source_product_name'],'marketPlaceUrl':data['marketPlaceUrl'],'thumbnail':data['thumbnail']}
|
|
|
85 |
|
|
|
86 |
elif source_id == 2:
|
|
|
87 |
try:
|
|
|
88 |
if data['identifier'] is None or len(data['identifier'])==0:
|
|
|
89 |
return {}
|
|
|
90 |
url = "http://www.flipkart.com/ps/%s"%(data['identifier'])
|
|
|
91 |
vendorsData = scraperFk.read(url)
|
|
|
92 |
sortedVendorsData = []
|
|
|
93 |
sortedVendorsData = sorted(vendorsData, key=itemgetter('sellingPrice'))
|
|
|
94 |
print "data",sortedVendorsData
|
|
|
95 |
lowestSp, iterator = (0,)*2
|
|
|
96 |
for vData in sortedVendorsData:
|
|
|
97 |
if iterator == 0:
|
|
|
98 |
lowestSp = vData['sellingPrice']
|
|
|
99 |
break
|
|
|
100 |
if lowestSp > 0:
|
|
|
101 |
inStock = 1
|
|
|
102 |
print lowestSp
|
|
|
103 |
print inStock
|
|
|
104 |
if lowestSp > 0:
|
|
|
105 |
get_mongo_connection().Catalog.MasterData.update({'identifier':data['identifier'],'source_id':2}, {'$set' : {'available_price':lowestSp,'updatedOn':to_java_date(now),'in_stock':inStock}}, multi=True)
|
|
|
106 |
else:
|
|
|
107 |
get_mongo_connection().Catalog.MasterData.update({'identifier':data['identifier'],'source_id':2}, {'$set' : {'updatedOn':to_java_date(now),'in_stock':inStock}}, multi=True)
|
|
|
108 |
return {'_id':data['_id'],'available_price':lowestSp,'in_stock':inStock,'source_id':2,'source_product_name':data['source_product_name'],'marketPlaceUrl':data['marketPlaceUrl'],'thumbnail':data['thumbnail']}
|
|
|
109 |
except:
|
|
|
110 |
return {'_id':data['_id'],'available_price':data['available_price'],'in_stock':data['inStock'],'source_id':2,'source_product_name':data['source_product_name'],'marketPlaceUrl':data['marketPlaceUrl'],'thumbnail':data['thumbnail']}
|
|
|
111 |
else:
|
|
|
112 |
return {}
|
|
|
113 |
|
|
|
114 |
|
|
|
115 |
|
|
|
116 |
def getLatestPrice(skuBundleId, source_id):
|
|
|
117 |
temp = []
|
|
|
118 |
itemIds = list(get_mongo_connection().Catalog.MasterData.find({'skuBundleId':skuBundleId,'source_id' : source_id}))
|
|
|
119 |
for item in itemIds:
|
|
|
120 |
temp.append(returnLatestPrice(item, source_id))
|
|
|
121 |
return temp
|
|
|
122 |
|
|
|
123 |
|
|
|
124 |
|
|
|
125 |
def main():
|
|
|
126 |
print getLatestPrice(1,2)
|
|
|
127 |
|
|
|
128 |
if __name__=='__main__':
|
|
|
129 |
main()
|