Subversion Repositories SmartDukaan

Rev

Rev 15265 | Rev 17103 | Go to most recent revision | Details | Compare with Previous | Last modification | View Log | RSS feed

Rev Author Line No. Line
15950 kshitij.so 1
from dtr.utils.utils import fetchResponseUsingProxy, transformUrl
14122 kshitij.so 2
from sys import exit
3
import json
14746 kshitij.so 4
import re
15265 kshitij.so 5
import traceback
15950 kshitij.so 6
import datetime
7
from pyquery import PyQuery
14122 kshitij.so 8
 
14746 kshitij.so 9
 
15950 kshitij.so 10
 
14122 kshitij.so 11
headers = { 
12
            'User-agent':'Mozilla/5.0 (X11; Linux i686) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/37.0.2062.120 Safari/537.36',
13
            'Accept' : 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',      
14
            'Accept-Language' : 'en-US,en;q=0.8',                     
15
            'Accept-Charset' : 'ISO-8859-1,utf-8;q=0.7,*;q=0.3',
14746 kshitij.so 16
            'Connection':'keep-alive',
17
            'Accept-Encoding' : 'gzip,deflate,sdch'
14122 kshitij.so 18
        }
19
 
15950 kshitij.so 20
 
14122 kshitij.so 21
class FlipkartProductPageScraper:
22
    def __init__(self):
23
        self.count_trials = 0
24
        self.redirectCount = 0
25
 
26
    def read(self, url):
27
        response_data = ""
28
        redirect_url = ""
15950 kshitij.so 29
        url = transformUrl(url,2)
14122 kshitij.so 30
        try:
14538 kshitij.so 31
 
32
            """quick fix,need to add it conf""" 
33
 
15950 kshitij.so 34
            response_data = fetchResponseUsingProxy(url, headers, proxy=True)                
14122 kshitij.so 35
            print "Fetched response from flipkart for %s" %(url)
14538 kshitij.so 36
            #redirect_url = response.url
14122 kshitij.so 37
 
38
        except Exception as e:
14746 kshitij.so 39
            traceback.print_exc()
14122 kshitij.so 40
            print 'ERROR: ', e
41
            print 'Retrying'
42
            self.count_trials += 1
43
 
44
            if self.count_trials < 3:
45
                return self.read(url)
46
 
47
        self.response_data=response_data
15950 kshitij.so 48
        print datetime.datetime.now()
14122 kshitij.so 49
        return self.createData(url,redirect_url)
50
 
51
    def createData(self,url, redirect_url):
15950 kshitij.so 52
        pq = PyQuery(self.response_data)
53
        buyBoxPrice = float(pq('span.selling-price.omniture-field').attr['data-evar48'])
54
        inStock = 1
14122 kshitij.so 55
        try:
15950 kshitij.so 56
            sellerJson = pq('div.seller-table-wrap').attr['data-config']
57
            x = json.loads(sellerJson)
58
            lines = sorted(x['dataModel'], key=lambda k: k['priceInfo'].get('sellingPrice', 0), reverse=False)
14122 kshitij.so 59
            sellingPrice =  float(lines[0]['priceInfo']['sellingPrice'])
60
            try:
61
                offerText = lines[0]['offerInfo']['listingOffers'][0]['description']
62
            except:
63
                offerText = ""
15950 kshitij.so 64
            return {'lowestSp':sellingPrice,'inStock':inStock,'buyBoxPrice':buyBoxPrice}
14122 kshitij.so 65
        except:
15950 kshitij.so 66
            """Not able to parse seller wrap section, probably due to only single seller option"""
15265 kshitij.so 67
            sellingPrice = buyBoxPrice
15950 kshitij.so 68
            stockDiv = pq('div.out-of-stock')
69
            if len(stockDiv) > 0:
15265 kshitij.so 70
                inStock = 0
71
            return {'lowestSp':sellingPrice,'inStock':inStock,'buyBoxPrice':buyBoxPrice}
15950 kshitij.so 72
 
14122 kshitij.so 73
if __name__ == '__main__':
15950 kshitij.so 74
    print datetime.datetime.now()
14122 kshitij.so 75
    scraper = FlipkartProductPageScraper()
15950 kshitij.so 76
 
77
    print scraper.read('http://www.flipkart.com/redmi-2/p/itme8ygtcfax6w39')
78
    print datetime.datetime.now()