Issue
i use scrapy to get data from a website (www.imensa.de) and i need to persist those data/items into a sqlite3 database. The Spider works fine, but i think there's something wrong about the pipeline.
Within the pipeline, the new .db-file is created properly with the specified tables and columns. But its not getting populated.
Spider.py
# Import library
from __future__ import absolute_import
import scrapy
from scrapy.loader import Item
from scrapy.loader import ItemLoader
from ..items import Mensen_Table
from ..items import Mensa_Meal
import datetime
from datetime import date
from datetime import timedelta
global base_url
global meal_plan_day_list
meal_plan_day_list = []
# Create Spider class
class mensen_crawler(scrapy.Spider):
# Name of spider
name = 'mensen'
# Website to scrape
allowed_domains = ['imensa.de']
start_urls = ['https://www.imensa.de/']
# STATE_NAME / STATE_LINK
def parse(self, response):
base_url = 'https://www.imensa.de/'
# Loop through response to parse state_name & state_link
for i, (name, link) in enumerate(zip(response.css('.group a::text'),response.css('.group a::attr(href)'))):
# Condition to only get german states (16 states, index starting from 0 -> 15)
if i < 16:
state_name = name.get()
state_link = base_url + link.get()
yield scrapy.Request(state_link, callback=self.parse_layer2, cb_kwargs={'state_name': state_name, 'state_link':state_link})
# CITY_NAME / CITY_LINK
def parse_layer2(self, response, state_name, state_link):
global base_url
base_url = 'https://www.imensa.de/'
for (city, link) in zip(response.css('.group a::text'), response.css('.group a::attr(href)')):
city_name = city.get()
yield print('current_city: ',city_name,' (state: ',state_name,')')
city_link_part = link.get().split('/')[0]
yield print('city_link_part: ', city_link_part)
city_link_real = base_url + city_link_part + '/index.html'
yield scrapy.Request(city_link_real, callback=self.parse_layer3, cb_kwargs={'state_name': state_name, 'state_link':state_link, 'city_name': city_name,'city_link': city_link_real})
# MENSA_NAME/MENSA_LINK
def parse_layer3(self, response, state_name, state_link, city_name, city_link):
base_url = 'https://www.imensa.de/'
for group in response.css('.group'):
uni_name = group.css('h2::text').get()
yield print('UNI_NAME: ',uni_name)
for mensa in group.css('.element'):
mensa_name = mensa.css('a::text').get()
yield print('mensa_name: ',mensa_name,' (state: ',state_name,') (city: ',city_name,') (uni_name: ',uni_name,')')
mensa_link = base_url + city_link.replace('https://www.imensa.de/','').split('/')[0] + '/' + mensa.css('a::attr(href)').get()
yield print('mensa_link: ', mensa_link)
yield scrapy.Request(mensa_link, callback=self.parse_layer4, cb_kwargs={'state_name': state_name,
'state_link':state_link,
'city_name': city_name,
'city_link': city_link,
'uni_name': uni_name,
'mensa_name': mensa_name,
'mensa_link': mensa_link
})
# CREATE MENSA ITEM -----------------------------------------------------------------------------------------------------------------
def parse_layer4(self, response, state_name, state_link, city_name, city_link, uni_name, mensa_name, mensa_link):
l = ItemLoader(item=Mensen_Table(), response=response)
try:
rating_avg = response.css('.aw-ratings-average::text').get()
except:
rating_avg = 0
try:
rating_count = response.css('.aw-ratings-count::text').get()
except:
rating_count = 0
address_list = []
for address_element in response.css('a.panel-body::text'):
address_list.append(address_element.get())
mensa_location = ', '.join(address_list)
yield print('mensa_location: ', mensa_location)
yield print('parse_layer_3 -- rating_avg: ',rating_avg)
yield print('parse_layer_3 -- rating_count: ', rating_count)
l.add_value('state_name', state_name)
l.add_value('state_link', state_link)
l.add_value('city_name', city_name)
l.add_value('city_link', city_link)
l.add_value('uni_name', uni_name)
l.add_value('mensa_name', mensa_name)
l.add_value('mensen_link', mensa_link)
l.add_value('mensa_address', mensa_location)
l.add_value('mensen_rating_avg', rating_avg)
l.add_value('mensen_rating_count', rating_count)
yield l.load_item()
for i,x in enumerate(response.css('.col-md-4.no-padding-xs .list-group')):
if i == 0:
date_list = x.css('.pull-right::text').extract()
day_list = x.css('a::text').extract()
link_list = x.css('a::attr(href)').extract()
yield print('date_list: ',date_list)
yield print('day_list: ', day_list)
yield print('link_list: ',link_list)
yield print('mensa_link: ',mensa_link)
# PROCESS DATE LIST
#------------------------------------------
meal_plan_date_list = []
for ele in date_list:
if ele == 'heute':
today = datetime.date.today().strftime('%d.%m.%Y')
meal_plan_date_list.append(today)
elif ele == 'morgen':
today = datetime.date.today()
tomorrow = today + datetime.timedelta(days=1)
meal_plan_date_list.append(tomorrow.strftime('%d.%m.%Y'))
else:
meal_plan_date_list.append(ele)
yield print('meal_plan_date_list: ',meal_plan_date_list)
# PROCESS LINK LIST
#--------------------------------------------
meal_plan_link_list = []
for ele in day_list:
link = mensa_link.replace('index.html','') + ele.lower() + '.html'
meal_plan_link_list.append(link)
yield print('meal_plan_link_list: ',meal_plan_link_list)
#meal_plan_list = []
#meal_plan_prep_list = zip(meal_plan_date_list, day_list, meal_plan_link_list)
#for item in meal_plan_prep_list:
# yield print('meal_plan_list_item: ', item)
for date, day, link in zip(meal_plan_date_list, day_list, meal_plan_link_list):
yield scrapy.Request(link, callback=self.parse_layer5, cb_kwargs={'mensa_name': mensa_name, 'mensa_link': link, 'day': day, 'date': date})
# PARSE MEAL PLAN --------------------------------------------------------------------------
def parse_layer5(self, response, mensa_name, mensa_link, day, date):
for element in response.css('.aw-meal-category'):
for sub in element.css('.aw-meal.row.no-margin-xs'):
l = ItemLoader(item=Mensa_Meal(),response=response,selector=sub)
meal_name = sub.css('p.aw-meal-description::text').get()
try:
meal_price = sub.css('.col-sm-2.no-padding-xs.aw-meal-price::text').get().replace('€','').strip()
except:
meal_price = 0
try:
meal_attributes = sub.css('.small.aw-meal-attributes span::text').extract_first().replace(u'\xa0', u'')
except:
meal_attributes = ''
#if not meal_price == None:
l.add_value('mensa_name', mensa_name)
l.add_value('date_of_meal_plan', date)
l.add_value('meal_name', meal_name)
l.add_value('meal_attributes', meal_attributes)
l.add_value('meal_price', meal_price)
yield l.load_item()
items.py
# Define here the models for your scraped items
#
# See documentation in:
# https://docs.scrapy.org/en/latest/topics/items.html
import scrapy
from itemadapter import ItemAdapter
from scrapy.loader import ItemLoader
from itemloaders.processors import TakeFirst, MapCompose
from w3lib.html import remove_tags
from scrapy.item import Item
class Mensen_Table(scrapy.Item):
state_name = scrapy.Field()
state_link = scrapy.Field()
city_name = scrapy.Field()
city_link = scrapy.Field()
uni_name = scrapy.Field()
mensa_name = scrapy.Field()
mensen_link = scrapy.Field()
mensa_address = scrapy.Field()
mensen_rating_avg = scrapy.Field()
mensen_rating_count = scrapy.Field()
five_star_ratings = scrapy.Field()
four_star_ratings = scrapy.Field()
three_star_ratings = scrapy.Field()
two_star_ratings = scrapy.Field()
one_star_ratings = scrapy.Field()
class Mensa_Meal(scrapy.Item):
mensa_name = scrapy.Field()
date_of_meal_plan = scrapy.Field()
meal_name = scrapy.Field()
meal_attributes = scrapy.Field()
meal_price = scrapy.Field()
pipelines.py
# Define your item pipelines here
#
# Don't forget to add your pipeline to the ITEM_PIPELINES setting
# See: https://docs.scrapy.org/en/latest/topics/item-pipeline.html
from scrapy import item
from scrapy.exceptions import DropItem
from itemadapter import ItemAdapter
import sqlite3
# useful for handling different item types with a single interface
from items import Mensen_Table, Mensa_Meal
class IwCrawlerPipeline:
def __init__(self):
self.con = sqlite3.connect('imensa.db')
self.cur = self.con.cursor()
self.create_table_mensen()
self.create_table_meal()
def create_table_mensen(self):
print('TABLE MENSEN CREATED')
self.cur.execute("""CREATE TABLE IF NOT EXISTS mensen_table (
state_name TEXT,
city_name TEXT,
uni_name TEXT,
mensa_name TEXT,
mensa_address LONG,
mensen_rating_avg FLOAT,
mensen_rating_count TEXT)
""")
def create_table_meal(self):
return print('TABLE MEAL CREATED')
self.cur.execute("""CREATE TABLE IF NOT EXISTS meal_table (
mensa_name TEXT,
date_of_meal_plan DATE,
meal_name LONG,
meal_attributes LONG,
meal_price FLOAT)
""")
def process_item(self, item, spider):
if isinstance(item, Mensen_Table):
print('MENSEN TABLE ITEM PROCESSED')
self.cur.execute("""INSERT INTO mensen_table (state_name, city_name, uni_name, mensa_name, mensa_address, mensen_rating_avg, mensen_rating_count) VALUES (?, ?, ?, ?, ?, ?, ?)""",
(item['state_name'], item['city_name'], item['uni_name'], item['mensa_name'], item['mensa_address'], item['mensen_rating_avg'], item['mensen_rating_count']))
self.con.commit()
return item
if isinstance(item, Mensa_Meal):
print('MEAL TABLE ITEM PROCESSED')
self.cur.execute("""INSERT INTO meal_table (mensa_name, date_of_meal_plan, meal_name, meal_attributes, meal_price) VALUES (?, ?, ?, ?, ?)""",
(item['mensa_name'], item['date_of_meal_plan'], item['meal_name'], item['meal_attributes'], item['meal_price']))
self.con.commit()
return item
What am i doing wrong? The items get displayed properly, but the dont reach the database-file.
Any help would be highly appreciated!
Solution
The only problem I see is that you never close the connection to the database, although I am not sure that this will solve your issue. It might though and it's good practice to do anyway.
In your pipeline add close_spider method that closes the connection to the database. This method is triggered automatically by the scrapy engine just before it deletes the spider and closes the twisted reactor.
class IwCrawlerPipeline:
def __init__(self):
self.con = sqlite3.connect('imensa.db')
self.cur = self.con.cursor()
self.create_table_mensen()
self.create_table_meal()
...
def close_spider(self, spider):
self.con.close()
Answered By - Alexander
0 comments:
Post a Comment
Note: Only a member of this blog may post a comment.