使用Kimurai gem进行Web抓取

Question

我正在用Kimurai Ruby gem做一些网页抓取。我有一个运行得很好的脚本：

require 'kimurai'

class SimpleSpider < Kimurai::Base
  @name = "simple_spider"
  @engine = :selenium_chrome
  @start_urls = ["https://apply.workable.com/taxjar/"]

  def parse(response, url:, data: {})
    # Update response to current response after interaction with a browser
    count = 0
    # browser.click_button "Show more"
    doc = browser.current_response
    returned_jobs = doc.css('.careers-jobs-list-styles__jobsList--3_v12')
    returned_jobs.css('li').each do |char_element|
        # puts char_element
        title = char_element.css('a')[0]['aria-label']
        link = "https://apply.workable.com" + char_element.css('a')[0]['href']

        #click on job link and get description
        browser.visit(link)
        job_page = browser.current_response
        description = job_page.xpath('/html/body/div[1]/div/div[1]/div[2]/div[2]/div[2]').text

        puts '*******'
        puts title
      puts link
        puts description
        puts count += 1
    end
    puts "There are #{count} jobs total"
  end
end

SimpleSpider.crawl!

但是，在本例中，我希望所有这些都返回一个objects...or作业数组。我想在解析方法中创建一个作业数组，并在returned_jobs循环中执行类似jobs << [title, link, description, company]的操作，并在调用SimpleSpider.crawl!时返回该作业数组，但这不起作用。

感谢您的帮助。

Answer 1

原来有一个允许返回值的parse method。以下是工作示例：

require 'open-uri'
require 'nokogiri'
require 'kimurai'

class TaxJar < Kimurai::Base
  @name = "tax_jar"
  @engine = :selenium_chrome
  @start_urls = ["https://apply.workable.com/taxjar/"]

  def parse(response, url:, data: {})
  jobs = Array.new

    doc = browser.current_response
    returned_jobs = doc.css('.careers-jobs-list-styles__jobsList--3_v12')
    returned_jobs.css('li').each do |char_element|
      title = char_element.css('a')[0]['aria-label']
      link = "https://apply.workable.com" + char_element.css('a')[0]['href']

      #click on job link and get description
      browser.visit(link)
      job_page = browser.current_response
      description = job_page.xpath('/html/body/div[1]/div/div[1]/div[2]/div[2]/div[2]').text
      company = 'TaxJar'
      puts "title is: #{title}, link is: #{link}, \n description is: #{description}"
      jobs << [title, link, description, company]
    end
    return jobs
  end
end

jobs = TaxJar.parse!(:parse, url: "https://apply.workable.com/taxjar/")
puts jobs.inspect

如果你正在抓取JS网站，与我尝试过的其他网站(waitr/selenium)相比，这个gem似乎相当健壮。

Answer 2

您可以稍微修改代码，如下所示：

class SimpleSpider < Kimurai::Base
  @name = "simple_spider"
  @engine = :selenium_chrome
  @start_urls = ["https://apply.workable.com/taxjar/"]

  def parse(response, url:, data: {})
    # Update response to current response after interaction with a browser
    count = 0
    # browser.click_button "Show more"
    doc = browser.current_response
    returned_jobs = doc.css('.careers-jobs-list-styles__jobsList--3_v12')

    jobs = []
    returned_jobs.css('li').each do |char_element|
        # puts char_element
        title = char_element.css('a')[0]['aria-label']
        link = "https://apply.workable.com" + char_element.css('a')[0]['href']

        #click on job link and get description
        browser.visit(link)
        job_page = browser.current_response
        description = job_page.xpath('/html/body/div[1]/div/div[1]/div[2]/div[2]/div[2]').text

        jobs << [title, link, description]
    end

    puts "There are #{jobs.count} jobs total"
    puts jobs
  end
end

我对公司不太确定，因为我在你的代码中看不到这个变量。但是，您可以看到调用上面的数组并在其上工作的想法。

以下是在终端中运行的输出的一部分：

​

​

我还有一篇关于如何使用Ruby on Rails应用程序中的Kimurai框架的博客文章here。