Using loginform with scrapy

Question

The scrapy framework (https://github.com/scrapy/scrapy) provides a library for use when logging into websites that require authentication, https://github.com/scrapy/loginfor

Answer 1

I managed to get it working without the loginform library, my solution is below.

import scrapy
import requests

class Spider(scrapy.Spider):
    name = 'spider'

    start_urls = [
        'http://start.com',
    ]

    def start_requests(self):
        return [scrapy.FormRequest("login.php",
                               formdata={'username': 'user', 'password': 'pass'},
                               callback=self.start_crawl)]

    def start_crawl(self, response):
        #start crawling

Answer 2

loginform is just a library, totally decoupled from Scrapy.

You have to write the code to plug it in the spider you want, probably in a callback method.

Here is an example of a structure to do this:

import scrapy
from loginform import fill_login_form

class MySpiderWithLogin(scrapy.Spider):
    name = 'my-spider'

    start_urls = [
        'http://somewebsite.com/some-login-protected-page',
        'http://somewebsite.com/another-protected-page',
    ]

    login_url = 'http://somewebsite.com/login-page'

    login_user = 'your-username'
    login_password = 'secret-password-here'

    def start_requests(self):
        # let's start by sending a first request to login page
        yield scrapy.Request(self.login_url, self.parse_login)

    def parse_login(self, response):
        # got the login page, let's fill the login form...
        data, url, method = fill_login_form(response.url, response.body,
                                            self.login_user, self.login_password)

        # ... and send a request with our login data
        return scrapy.FormRequest(url, formdata=dict(data),
                           method=method, callback=self.start_crawl)

    def start_crawl(self, response):
        # OK, we're in, let's start crawling the protected pages
        for url in self.start_urls:
            yield scrapy.Request(url)

    def parse(self, response):
        # do stuff with the logged in response