将硒响应网址传递给scrapy

Question

我正在学习Python,并试图在下拉菜单中抓取此页面以获取特定值.之后,我需要单击结果表上的每个项目以检索特定信息.我能够选择项目并检索webdriver上的信息.但我不知道如何将响应URL传递给crawlspider.

driver = webdriver.Firefox()
driver.get('http://www.cppcc.gov.cn/CMS/icms/project1/cppcc/wylibary/wjWeiYuanList.jsp')
more_btn = WebDriverWait(driver, 20).until(
     EC.visibility_of_element_located((By.ID, '_button_select'))
            )  
more_btn.click()

## select specific value from the dropdown
driver.find_element_by_css_selector("select#tabJcwyxt_jiebie >     option[value='teyaoxgrs']").click()
driver.find_element_by_css_selector("select#tabJcwyxt_jieci > option[value='d11jie']").click()
search2 = driver.find_element_by_class_name('input_a2')
search2.click()
time.sleep(5)

## convert html to "nice format"
text_html=driver.page_source.encode('utf-8')
html_str=str(text_html)

## this is a hack that initiates a "TextResponse" object (taken from the Scrapy module)
resp_for_scrapy=TextResponse('none',200,{},html_str,[],None)

## convert html to "nice format"
text_html=driver.page_source.encode('utf-8')
html_str=str(text_html)

resp_for_scrapy=TextResponse('none',200,{},html_str,[],None)

所以这就是我被困住的地方.我能够使用上面的代码进行查询.但是如何将resp_for_scrapy传递给crawlspider？我把resp_for_scrapy替换为项目但是没有用.

## spider 
class ProfileSpider(CrawlSpider):
name = 'pccprofile2'
allowed_domains = ['cppcc.gov.cn']
start_urls = ['http://www.cppcc.gov.cn/CMS/icms/project1/cppcc/wylibary/wjWeiYuanList.jsp']    

def parse(self, resp_for_scrapy):

    hxs = HtmlXPathSelector(resp_for_scrapy)
    for post in resp_for_scrapy.xpath('//div[@class="table"]//ul//li'):
        items = []
        item = Ppcprofile2Item()
        item ["name"] = hxs.select("//h1/text()").extract()
        item ["title"] = hxs.select("//div[@id='contentbody']//tr//td//text()").extract()
        items.append(item)

    ##click next page      
    while True:
        next = self.driver.findElement(By.linkText("???"))
        try:
            next.click()
        except:
            break

    return(items)

任何建议将不胜感激!!!!

编辑我包括一个中间件类,可以从蜘蛛类之前的下拉列表中进行选择.但现在没有错误也没有结果.

class JSMiddleware(object):
    def process_request(self, request, spider):
        driver = webdriver.PhantomJS()
         driver.get('http://www.cppcc.gov.cn/CMS/icms/project1/cppcc/wylibary/wjWeiYuanList.jsp')


    # select from the dropdown
        more_btn = WebDriverWait(driver, 20).until(
        EC.visibility_of_element_located((By.ID, '_button_select'))
                )
        more_btn.click()


        driver.find_element_by_css_selector("select#tabJcwyxt_jiebie > option[value='teyaoxgrs']").click()
        driver.find_element_by_css_selector("select#tabJcwyxt_jieci > option[value='d11jie']").click()
        search2 = driver.find_element_by_class_name('input_a2')
        search2.click()
        time.sleep(5)

        #get the response 
        body = driver.page_source
        return HtmlResponse(driver.current_url, body=body, encoding='utf-8', request=request)



class ProfileSpider(CrawlSpider):
    name = 'pccprofile2'
    rules = [Rule(SgmlLinkExtractor(allow=(),restrict_xpaths=("//div[@class='table']")), callback='parse_item')]  

    def parse_item(self, response):
    hxs = HtmlXPathSelector(response)
    items = []
    item = Ppcprofile2Item()
    item ["name"] = hxs.select("//h1/text()").extract()
    item ["title"] = hxs.select("//div[@id='contentbody']//tr//td//text()").extract()
    items.append(item)

    #click next page      
    while True:
        next = response.findElement(By.linkText("???"))
        try:
            next.click()
        except:
            break

    return(items)

Answer 1

在使用Scrapy定期处理页面之前,使用下载器中间件来捕获需要硒的页面:

下载器中间件是Scrapy的请求/响应处理的钩子框架.它是一个轻量级的低级系统,用于全局改变Scrapy的请求和响应.

这是使用PhantomJS的一个非常基本的例子:

from scrapy.http import HtmlResponse
from selenium import webdriver

class JSMiddleware(object):
    def process_request(self, request, spider):
        driver = webdriver.PhantomJS()
        driver.get(request.url)

        body = driver.page_source
        return HtmlResponse(driver.current_url, body=body, encoding='utf-8', request=request)

一旦你返回HtmlResponse(或者TextResponse如果那是你真正想要的),Scrapy将停止处理下载程序并放入蜘蛛的parse方法:

如果它返回一个Response对象,Scrapy将不会打扰调用任何其他process_request()或process_exception()方法或相应的下载函数; 它会回复那个回应.始终在每个响应上调用已安装的中间件的process_response()方法.

在这种情况下,您可以parse像通常使用HTML一样继续使用spider的方法,除了页面上的JS已经执行.

提示:由于Downloader Middleware的process_request方法接受蜘蛛作为参数,你可以在spider中添加一个条件来检查你是否需要处理JS,这将让你处理完全相同的JS和非JS页面蜘蛛类.

Answer 2

这是Scrapy和Selenium的中间件

from scrapy.http import HtmlResponse
from scrapy.utils.python import to_bytes
from selenium import webdriver
from scrapy import signals


class SeleniumMiddleware(object):

    @classmethod
    def from_crawler(cls, crawler):
        middleware = cls()
        crawler.signals.connect(middleware.spider_opened, signals.spider_opened)
        crawler.signals.connect(middleware.spider_closed, signals.spider_closed)
        return middleware

    def process_request(self, request, spider):
        request.meta['driver'] = self.driver  # to access driver from response
        self.driver.get(request.url)
        body = to_bytes(self.driver.page_source)  # body must be of type bytes 
        return HtmlResponse(self.driver.current_url, body=body, encoding='utf-8', request=request)

    def spider_opened(self, spider):
        self.driver = webdriver.Firefox()

    def spider_closed(self, spider):
        self.driver.close()

还需要加入 settings.py

DOWNLOADER_MIDDLEWARES = {
    'youproject.middlewares.selenium.SeleniumMiddleware': 200
}

200根据文档确定天气或其他内容.

使用scrapy和selenium 更新firefox无头模式

如果你想在无头模式下运行firefox,那么安装xvfb

sudo apt-get install -y xvfb

和PyVirtualDisplay

sudo pip install pyvirtualdisplay

并使用此中间件

from shutil import which

from pyvirtualdisplay import Display
from scrapy import signals
from scrapy.http import HtmlResponse
from scrapy.utils.project import get_project_settings
from selenium import webdriver
from selenium.webdriver.firefox.firefox_binary import FirefoxBinary

settings = get_project_settings()

HEADLESS = True


class SeleniumMiddleware(object):

    @classmethod
    def from_crawler(cls, crawler):
        middleware = cls()
        crawler.signals.connect(middleware.spider_opened, signals.spider_opened)
        crawler.signals.connect(middleware.spider_closed, signals.spider_closed)
        return middleware

    def process_request(self, request, spider):
        self.driver.get(request.url)
        request.meta['driver'] = self.driver
        body = str.encode(self.driver.page_source)
        return HtmlResponse(self.driver.current_url, body=body, encoding='utf-8', request=request)

    def spider_opened(self, spider):
        if HEADLESS:
            self.display = Display(visible=0, size=(1280, 1024))
            self.display.start()
        binary = FirefoxBinary(settings.get('FIREFOX_EXE') or which('firefox'))
        self.driver = webdriver.Firefox(firefox_binary=binary)

    def spider_closed(self, spider):
        self.driver.close()
        if HEADLESS:
            self.display.stop()

哪里settings.py包含

FIREFOX_EXE = '/path/to/firefox.exe'

问题是某些版本的Firefox不能与selenium一起使用.要解决这个问题,你可以从这里下载firefox版本47.0.1(这个版本完美无缺),然后将其解压缩并放入你的selenium项目中.然后修改firefox路径为

FIREFOX_EXE = '/path/to/your/scrapyproject/firefox/firefox.exe'