| 12256 |
kshitij.so |
1 |
from BeautifulSoup import BeautifulSoup
|
|
|
2 |
import mechanize
|
|
|
3 |
import re
|
|
|
4 |
|
|
|
5 |
class SellerCentralScraper:
|
|
|
6 |
|
|
|
7 |
def getBrowserObject(self):
|
|
|
8 |
import cookielib
|
|
|
9 |
br = mechanize.Browser(factory=mechanize.RobustFactory())
|
|
|
10 |
cj = cookielib.LWPCookieJar()
|
|
|
11 |
br.set_cookiejar(cj)
|
|
|
12 |
br.set_handle_equiv(True)
|
|
|
13 |
br.set_handle_redirect(True)
|
|
|
14 |
br.set_handle_referer(True)
|
|
|
15 |
br.set_handle_robots(False)
|
|
|
16 |
br.set_debug_http(False)
|
|
|
17 |
br.set_debug_redirects(False)
|
|
|
18 |
br.set_debug_responses(False)
|
|
|
19 |
|
|
|
20 |
br.set_handle_refresh(mechanize._http.HTTPRefreshProcessor(), max_time=1)
|
|
|
21 |
|
|
|
22 |
br.addheaders = [('User-agent','Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/535.11 (KHTML, like Gecko) Chrome/17.0.963.56 Safari/535.11'),
|
|
|
23 |
('Accept', 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8'),
|
|
|
24 |
('Accept-Encoding', 'gzip,deflate,sdch'),
|
|
|
25 |
('Accept-Language', 'en-US,en;q=0.8'),
|
|
|
26 |
('Accept-Charset', 'ISO-8859-1,utf-8;q=0.7,*;q=0.3')]
|
|
|
27 |
return br
|
|
|
28 |
|
|
|
29 |
def login(self,url):
|
|
|
30 |
br = self.getBrowserObject()
|
|
|
31 |
br.open(url)
|
|
|
32 |
response = br.open(url)
|
|
|
33 |
self.ungzipResponse(response, br)
|
|
|
34 |
#html = response.read()
|
|
|
35 |
#print html
|
|
|
36 |
br.select_form(name="signinWidget")
|
|
|
37 |
br.form['username'] = "kshitij.sood@saholic.com"
|
|
|
38 |
br.form['password'] = "pioneer"
|
|
|
39 |
response = br.submit()
|
|
|
40 |
print "********************"
|
|
|
41 |
print "Attempting to Login"
|
|
|
42 |
print "********************"
|
|
|
43 |
#ungzipResponse(response, br)
|
|
|
44 |
return br
|
|
|
45 |
|
|
|
46 |
def requestSku(self,br,skuUrl):
|
|
|
47 |
print "********************"
|
|
|
48 |
print "Requesting SKU Details"
|
|
|
49 |
print "********************"
|
|
|
50 |
response = br.open(skuUrl)
|
|
|
51 |
self.ungzipResponse(response, br)
|
|
|
52 |
page = response.read()
|
|
|
53 |
response = None
|
|
|
54 |
try:
|
|
|
55 |
return self.getSkuDetails(br, page)
|
|
|
56 |
except:
|
|
|
57 |
return '',0,0
|
|
|
58 |
|
|
|
59 |
def getSkuDetails(self,br,page):
|
|
|
60 |
soup = BeautifulSoup(page,convertEntities=BeautifulSoup.HTML_ENTITIES)
|
|
|
61 |
table_rows = soup.find("table" , {"class" : "data-display manageTable"})
|
|
|
62 |
skuDetails = table_rows.findAll("tr" , {"id" : re.compile('sku-.*')})
|
|
|
63 |
skuRow = skuDetails[0].findAll("td")
|
|
|
64 |
sku = str(skuRow[4].text)
|
|
|
65 |
if sku.startswith('FBA'):
|
|
|
66 |
isFba = True
|
|
|
67 |
else:
|
|
|
68 |
isFba = False
|
|
|
69 |
asin = skuRow[5].text
|
|
|
70 |
try:
|
|
|
71 |
inputTag = skuDetails[0].find(attrs={"name": "inv"})
|
|
|
72 |
inventory = inputTag['value']
|
|
|
73 |
except:
|
|
|
74 |
inventory = skuRow[8].text
|
|
|
75 |
try:
|
|
|
76 |
inputTag = skuDetails[0].find(attrs={"name": "price"})
|
|
|
77 |
ourPrice = inputTag['value']
|
|
|
78 |
except:
|
|
|
79 |
ourPrice = skuRow[10].text
|
|
|
80 |
|
|
|
81 |
return asin, inventory, ourPrice
|
|
|
82 |
|
|
|
83 |
def ungzipResponse(self,r,b):
|
|
|
84 |
headers = r.info()
|
|
|
85 |
if headers['Content-Encoding']=='gzip':
|
|
|
86 |
import gzip
|
|
|
87 |
print "********************"
|
|
|
88 |
print "Deflating gzip response"
|
|
|
89 |
print "********************"
|
|
|
90 |
gz = gzip.GzipFile(fileobj=r, mode='rb')
|
|
|
91 |
html = gz.read()
|
|
|
92 |
gz.close()
|
|
|
93 |
headers["Content-type"] = "text/html; charset=utf-8"
|
|
|
94 |
r.set_data( html )
|
|
|
95 |
b.set_response(r)
|
|
|
96 |
|
|
|
97 |
|
|
|
98 |
|
|
|
99 |
|
|
|
100 |
def main():
|
|
|
101 |
print "Opening Seller Central login page"
|
|
|
102 |
login_url = "https://sellercentral.amazon.in/gp/homepage.html"
|
|
|
103 |
sc = SellerCentralScraper()
|
|
|
104 |
br = sc.login(login_url)
|
| 12282 |
kshitij.so |
105 |
sku_url = "https://sellercentral.amazon.in/myi/search/ProductSummary?keyword=2287"
|
| 12256 |
kshitij.so |
106 |
print sc.requestSku(br,sku_url)
|
|
|
107 |
|
|
|
108 |
|
|
|
109 |
if __name__ == "__main__":
|
|
|
110 |
main()
|