Subversion Repositories SmartDukaan

'''

Created on 13-May-2010

@author: gaurav

'''

from scrapy.spider import BaseSpider

from scrapy.selector import HtmlXPathSelector

from scrapy.http import Request

from demo.items import DemoItem

from scrapy.contrib.spidermiddleware import referer

from scrapy.http.headers import Headers

from scrapy.http.request.form import FormRequest

from scrapy.log import msg

from scrapy.http.response import Response

from datastore.DataAccessor import *

from datastore.DataCodeAccessor import *

from html2text.unescaping import *

from elixir import *

class infi_spider(BaseSpider):

    """

    Documentation for class infi_spider

    This spider collects the information for the individual phones

    and store them in table datastore_datadefinition_infibeam_data

    """

    def __init__(self): 

       """

        Documentation for constructor

        initialize_table is called to make all the tables known in

        the scope of this class.

        Also start url needs to be feeded to the spider through start_urls.append

        Domainname is name by which this spider is known outside

        So this will be used as an argument for calling this spider.

        As the number of pages to be crawled are not fixed so ct and no are used to make it dynamic.  

        """

       initialize_table()

       da = DataHelper()

       #INFIBEAM_DOMAINNAME = "infibeam"   

       INFIBEAM_DOMAINNAME = get_code_word("INFIBEAM_DOMAINNAME")

       print INFIBEAM_DOMAINNAME

       self.domain_name = INFIBEAM_DOMAINNAME

       #INFIBEAM_CT = 15

       #INFIBEAM_CT = int(get_code_word("INFIBEAM_CT")) 

       #ct = 15     

       ct = int(da.get_extra_vars('infibeam_count'))

       #INFIBEAM_NO = 1

       no = 1

       if ct>14:

           no = ct

       #INFIBEAM_URL = "http://www.infibeam.com/Mobiles/search?page="

       INFIBEAM_URL = get_code_word("INFIBEAM_URL")

       while(no<=ct):

            url1 = INFIBEAM_URL + str(no)

            self.start_urls.append(url1)

            no=no+1

    def start_requests(self):

        """

        Documentation for method start_requests

        To set various properties of the request to be made

        like referer, headers and all.

        Also suppliers entry need to be done in the table

        datastore_datadefinition_suppliers.

        @return a list of well formed requests which will be 

        crawled by spider and spider will return the response

        """

        #adding entry for the supplier i.e its name and site

        #INFIBEAM_HOMEPAGE = "www.infibeam.com"

        INFIBEAM_HOMEPAGE = get_code_word("INFIBEAM_HOMEPAGE")

        da = DataHelper()

        da.add_supplier(self.domain_name, INFIBEAM_HOMEPAGE)

        listreq = []

        #for each request a referer has to be set

        #INFIBEAM_REFERER = "www.google.com/search"

        INFIBEAM_REFERER = get_code_word("INFIBEAM_REFERER") 

        for url1 in self.start_urls:

            request = Request(url = str(url1), callback=self.parse, dont_filter=True)

            request.headers.setdefault("Referer", INFIBEAM_REFERER)

            listreq.append(request)    

        return listreq

    def parse(self, response):

        """

        Documentation for method parse

        @param response of individual requests

        Using Xpaths needed information is extracted out of the response

        and added to the database

        Xpath1 = Give us section for individual phone

        Xpath2 = Give us name for individual phone

        Xpath3 = Give us quoted price for individual phone

        vatplustax = to get final price from quoted price

        Removelist = To filer the prices so as to make them integer for eg remove ',' or 'Rs'

        """

        da = DataHelper()

        msg(response.url)

        #INFIBEAM_VATPLUSTAX = 0

        INFIBEAM_VATPLUSTAX = int(get_code_word("INFIBEAM_VATPLUSTAX"))

        #list elements are separated by ';'

        #INFIBEAM_REMOVELIST = ["Rs.",",","-","/","Rs"]

        INFIBEAM_REMOVELIST = get_code_word("INFIBEAM_REMOVELIST")

        INFIBEAM_REMOVELIST = INFIBEAM_REMOVELIST.split(';') 

        hxs = HtmlXPathSelector(response)

        #INFIBEAM_XPATH1 = '//ul[@class="srch_result portrait"]/li'

        INFIBEAM_XPATH1 = get_code_word("INFIBEAM_XPATH1")

        phone_info = hxs.select(INFIBEAM_XPATH1)

        #INFIBEAM_XPATH2 = './/p/span[@class="title"]/text()'

        INFIBEAM_XPATH2 = get_code_word("INFIBEAM_XPATH2")

        #INFIBEAM_XPATH3 = './/p/span[@class="price"]/text()'

        INFIBEAM_XPATH3 = get_code_word("INFIBEAM_XPATH3")

        items = []

        msg(response.url)

        if not phone_info:

            ct = int(da.get_extra_vars('infibeam_count'))

            if ct>14:

                fails = int(da.get_extra_vars('infibeam_fails'))

                fails = fails+1

                if fails > 0:

                    da.set_extra_vars('infibeam_flag','FALSE','')

                da.set_extra_vars('infibeam_fails',str(fails),'')

        else:

            for i in phone_info:

                item = {}

                item['name'] = i.select(INFIBEAM_XPATH2)[0].extract()

                item['price'] = i.select(INFIBEAM_XPATH3)[0].extract()

                items.append(item)

            for i in items:

                amnt = i['price']

                if amnt != '':        

                    for r in INFIBEAM_REMOVELIST: 

                        while amnt.find(r) != -1:

                            amnt = amnt.replace(r, "") 

                        amnt = amnt.strip()

                pr = int(amnt) + INFIBEAM_VATPLUSTAX

                da.add_infiphone(i['name'], amnt,pr)

SPIDER = infi_spider()