Question

我正在抓住马德里大会的网站，用aspx构建，我不知道如何模拟链接上的点击，我需要从中获取相应的政治家。我试过这个：

import scrapy

class AsambleaMadrid(scrapy.Spider):

name        =   "Asamblea_Madrid"
start_urls  =   ['http://www.asambleamadrid.es/ES/QueEsLaAsamblea/ComposiciondelaAsamblea/LosDiputados/Paginas/RelacionAlfabeticaDiputados.aspx']

def parse(self, response):

    for id in response.css('div#moduloBusqueda div.sangria div.sangria ul li a::attr(id)'):
        target                  =   id.extract()
        url                     =   "http://www.asambleamadrid.es/ES/QueEsLaAsamblea/ComposiciondelaAsamblea/LosDiputados/Paginas/RelacionAlfabeticaDiputados.aspx"

        formdata=   {'__EVENTTARGET': target,
                     '__VIEWSTATE': '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',
                     '__EVENTVALIDATION': '/wEWCALIhqvYAwKh2YVvAuDF1KUDAqCK1bUOAqCKybkPAqCKnbQCAqCKsZEJAvejv84Dtkx5dCFr3QGqQD2wsFQh8nP3iq8',
                     '__VIEWSTATEGENERATOR': 'BAB98CB3',
                     '__REQUESTDIGEST': '0x476239970DCFDABDBBDF638A1F9B026BD43022A10D1D757B05F1071FF3104459B4666F96A47B4845D625BCB2BE0D88C6E150945E8F5D82C189B56A0DA4BC859D'}

        yield scrapy.FormRequest(url=url, formdata= formdata, callback=self.takeEachParty)


def takeEachParty(self, response):

     print response.css('ul.listadoVert02 ul li::text').extract()

进入网站的源代码，我可以看到链接的样子，以及它们如何发送JavaScript查询。这是我需要访问的链接之一：

<a id="ctl00_m_g_36ea0310_893d_4a19_9ed1_88a133d06423_ctl00_Repeater1_ctl00_lnk_Grupo" href="javascript:WebForm_DoPostBackWithOptions(new WebForm_PostBackOptions(&quot;ctl00$m$g_36ea0310_893d_4a19_9ed1_88a133d06423$ctl00$Repeater1$ctl00$lnk_Grupo&quot;, &quot;&quot;, true, &quot;&quot;, &quot;&quot;, false, true))">Grupo Parlamentario Popular de la Asamblea de Madrid</a>

我一直在读这么多篇文章，但问题可能是我的无知。

提前致谢。

编辑：

解决方案：我终于做到了！将Padraic Cunningham的帮助代码翻译成Scrapy方式。当我为Scrapy指定问题时，我想发布结果，以防有人遇到与我相同的问题。

所以这就是：

import scrapy
import js2xml

class AsambleaMadrid(scrapy.Spider):

     name        =   "AsambleaMadrid"
     start_urls  =   ['http://www.asambleamadrid.es/ES/QueEsLaAsamblea/ComposiciondelaAsamblea/LosDiputados/Paginas/RelacionAlfabeticaDiputados.aspx']

    def parse(self, response):

         source  =   response
         hrefs   =   response.xpath("//*[@id='moduloBusqueda']//div[@class='sangria']/ul/li/a/@href").extract()
         form_data = self.validate(source)
         for ref in hrefs:
             # js2xml allows us to parse the JS function and params, and so to grab the __EVENTTARGET
             js_xml            = js2xml.parse(ref)
             _id               = js_xml.xpath(
                            "//identifier[@name='WebForm_PostBackOptions']/following-sibling::arguments/string[starts-with(.,'ctl')]")[0]
             form_data["__EVENTTARGET"] = _id.text

             url_diputado    =   'http://www.asambleamadrid.es/ES/QueEsLaAsamblea/ComposiciondelaAsamblea/LosDiputados/Paginas/RelacionAlfabeticaDiputados.aspx'
             # The proper way to send a POST in scrapy is by using the FormRequest
             yield scrapy.FormRequest(url=url_diputado, formdata=form_data, callback=self.extract_parties, method='POST')

     def validate(self, source):
         # these fields are the minimum required as cannot be hardcoded
         data = {"__VIEWSTATEGENERATOR": source.xpath("//*[@id='__VIEWSTATEGENERATOR']/@value")[0].extract(),
             "__EVENTVALIDATION": source.xpath("//*[@id='__EVENTVALIDATION']/@value")[0].extract(),
             "__VIEWSTATE": source.xpath("//*[@id='__VIEWSTATE']/@value")[0].extract(),
             " __REQUESTDIGEST": source.xpath("//*[@id='__REQUESTDIGEST']/@value")[0].extract()}
         return data

     def extract_parties(self, response):
         source      =   response
         name        =   source.xpath("//ul[@class='listadoVert02']/ul/li/a/text()").extract()
         print name

我希望很清楚。再次感谢大家！

Answer 1

如果您查看在chrome或firebug中发布到表单的数据，您可以看到在post请求中传递了许多字段，有一些是必不可少的，必须从原始页面解析，解析来自的div.sangria ul li a标签不够用，因为发布的实际数据略有不同，发布的内容是Javascript函数，WebForm_DoPostBackWithOptions位于 href 而不是 id 属性：

href='javascript:WebForm_DoPostBackWithOptions(new 
 WebForm_PostBackOptions("ctl00$m$g_36ea0310_893d_4a19_9ed1_88a133d06423$ctl00$Repeater1$ctl03$lnk_Grupo", "", true, "", "", false, true))'>

有时所有的下划线都被美元符号替换，所以很容易做一个 str.replace 来使它们按照正确的顺序排列，但在这种情况下并不是真的，我们可以使用正则表达式进行解析但我喜欢js2xml lib，它可以将javascript函数及其args解析为xml树。

以下使用请求的代码向您展示了如何从初始请求获取数据并获取所需的所有页面：

import requests
from  lxml import html
import js2xml

post = "http://www.asambleamadrid.es/ES/QueEsLaAsamblea/ComposiciondelaAsamblea/LosDiputados/Paginas/RelacionAlfabeticaDiputados.aspx"


def validate(xml):
    # these fields are the minimum required as cannot be hardcoded
    data = {"__VIEWSTATEGENERATOR": xml.xpath("//*[@id='__VIEWSTATEGENERATOR']/@value")[0],
            "__EVENTVALIDATION": xml.xpath("//*[@id='__EVENTVALIDATION']/@value")[0],
            "__VIEWSTATE": xml.xpath("//*[@id='__VIEWSTATE']/@value")[0],
            " __REQUESTDIGEST": xml.xpath("//*[@id='__REQUESTDIGEST']/@value")[0]}
    return data



with requests.Session() as s:
    # make initial requests to get the links/hrefs and the from fields
    r = s.get(
        "http://www.asambleamadrid.es/ES/QueEsLaAsamblea/ComposiciondelaAsamblea/LosDiputados/Paginas/RelacionAlfabeticaDiputados.aspx")
    xml = html.fromstring(r.content)
    hrefs = xml.xpath("//*[@id='moduloBusqueda']//div[@class='sangria']/ul/li/a/@href")
    form_data = validate(xml)
    for h in hrefs:
        js_xml = js2xml.parse(h)
        _id = js_xml.xpath(
            "//identifier[@name='WebForm_PostBackOptions']/following-sibling::arguments/string[starts-with(.,'ctl')]")[
            0]
        form_data["__EVENTTARGET"] = _id.text
        r = s.post(post, data=form_data)
        xml = html.fromstring(r.content)
        print(xml.xpath("//ul[@class='listadoVert02']/ul/li/a/text()"))

如果我们运行上面的代码，我们会看到所有teh锚标记的不同文本输出：

In [2]: with requests.Session() as s:
   ...:         r = s.get(
   ...:             "http://www.asambleamadrid.es/ES/QueEsLaAsamblea/ComposiciondelaAsamblea/LosDiputados/Paginas/RelacionAlfabeticaDiputados.aspx")
   ...:         xml = html.fromstring(r.content)
   ...:         hrefs = xml.xpath("//*[@id='moduloBusqueda']//div[@class='sangria']/ul/li/a/@href")
   ...:         form_data = validate(xml)
   ...:         for h in hrefs:
   ...:                 js_xml = js2xml.parse(h)
   ...:                 _id = js_xml.xpath(
   ...:                     "//identifier[@name='WebForm_PostBackOptions']/following-sibling::arguments/string[starts-with(.,'ctl')]")[
   ...:                     0]
   ...:                 form_data["__EVENTTARGET"] = _id.text
   ...:                 r = s.post(post, data=form_data)
   ...:                 xml = html.fromstring(r.content)
   ...:                 print(xml.xpath("//ul[@class='listadoVert02']/ul/li/a/text()"))
   ...:         
[u'Abo\xedn Abo\xedn, Sonsoles Trinidad', u'Adrados Gautier, M\xaa Paloma', u'Aguado Del Olmo, M\xaa Josefa', u'\xc1lvarez Padilla, M\xaa Nadia', u'Arribas Del Barrio, Jos\xe9 M\xaa', u'Ballar\xedn Valc\xe1rcel, \xc1lvaro C\xe9sar', u'Berrio Fern\xe1ndez-Caballero, M\xaa In\xe9s', u'Berzal Andrade, Jos\xe9 Manuel', u'Cam\xedns Mart\xednez, Ana', u'Carballedo Berlanga, M\xaa Eugenia', 'Cifuentes Cuencas, Cristina', u'D\xedaz Ayuso, Isabel Natividad', u'Escudero D\xedaz-Tejeiro, Marta', u'Fermosel D\xedaz, Jes\xfas', u'Fern\xe1ndez-Quejo Del Pozo, Jos\xe9 Luis', u'Garc\xeda De Vinuesa Gardoqui, Ignacio', u'Garc\xeda Mart\xedn, Mar\xeda Bego\xf1a', u'Garrido Garc\xeda, \xc1ngel', u'G\xf3mez Ruiz, Jes\xfas', u'G\xf3mez-Angulo Rodr\xedguez, Juan Antonio', u'Gonz\xe1lez Gonz\xe1lez, Isabel Gema', u'Gonz\xe1lez Jim\xe9nez, Bartolom\xe9', u'Gonz\xe1lez Taboada, Jaime', u'Gonz\xe1lez-Mo\xf1ux V\xe1zquez, Elena', u'Gonzalo L\xf3pez, Rosal\xeda', 'Izquierdo Torres, Carlos', u'Li\xe9bana Montijano, Pilar', u'Mari\xf1o Ortega, Ana Isabel', u'Moraga Valiente, \xc1lvaro', u'Mu\xf1oz Abrines, Pedro', u'N\xfa\xf1ez Guijarro, Jos\xe9 Enrique', u'Olmo Fl\xf3rez, Luis Del', u'Ongil Cores, M\xaa Gador', 'Ortiz Espejo, Daniel', u'Ossorio Crespo, Enrique Mat\xedas', 'Peral Guerra, Luis', u'P\xe9rez Baos, Ana Isabel', u'P\xe9rez Garc\xeda, David', u'Pla\xf1iol De Lacalle, Regina M\xaa', u'Redondo Alcaide, M\xaa Isabel', u'Roll\xe1n Ojeda, Pedro', u'S\xe1nchez Fern\xe1ndez, Alejandro', 'Sanjuanbenito Bonal, Diego', u'Serrano Guio, Jos\xe9 Tom\xe1s', u'Serrano S\xe1nchez-Capuchino, Alfonso Carlos', 'Soler-Espiauba Gallo, Juan', 'Toledo Moreno, Lucila', 'Van-Halen Acedo, Juan']
[u'Andaluz Andaluz, M\xaa Isabel', u'Ardid Jim\xe9nez, M\xaa Isabel', u'Carazo G\xf3mez, M\xf3nica', u'Casares D\xedaz, M\xaa Luc\xeda Inmaculada', u'Cepeda Garc\xeda De Le\xf3n, Jos\xe9 Carmelo', 'Cruz Torrijos, Diego', u'Delgado G\xf3mez, Carla', u'Franco Pardo, Jos\xe9 Manuel', u'Freire Campo, Jos\xe9 Manuel', u'Gabilondo Pujol, \xc1ngel', 'Gallizo Llamas, Mercedes', u"Garc\xeda D'Atri, Ana", u'Garc\xeda-Rojo Garrido, Pedro Pablo', u'G\xf3mez Montoya, Rafael', u'G\xf3mez-Chamorro Torres, Jos\xe9 \xc1ngel', u'Gonz\xe1lez Gonz\xe1lez, M\xf3nica Silvana', u'Leal Fern\xe1ndez, M\xaa Isaura', u'Llop Cuenca, M\xaa Pilar', 'Lobato Gandarias, Juan', u'L\xf3pez Ruiz, M\xaa Carmen', u'Manguan Valderrama, Eva M\xaa', u'Maroto Illera, M\xaa Reyes', u'Mart\xednez Ten, Carmen', u'Mena Romero, M\xaa Carmen', u'Moreno Navarro, Juan Jos\xe9', u'Moya Nieto, Encarnaci\xf3n', 'Navarro Lanchas, Josefa', 'Nolla Estrada, Modesto', 'Pardo Ortiz, Josefa Dolores', u'Quintana Viar, Jos\xe9', u'Rico Garc\xeda-Hierro, Enrique', u'Rodr\xedguez Garc\xeda, Nicol\xe1s', u'S\xe1nchez Acera, Pilar', u'Sant\xedn Fern\xe1ndez, Pedro', 'Segovia Noriega, Juan', 'Vicente Viondi, Daniel', u'Vinagre Alc\xe1zar, Agust\xedn']
['Abasolo Pozas, Olga', 'Ardanuy Pizarro, Miguel', u'Beirak Ulanosky, Jazm\xedn', u'Camargo Fern\xe1ndez, Ra\xfal', 'Candela Pokorna, Marco', 'Delgado Orgaz, Emilio', u'D\xedaz Rom\xe1n, Laura', u'Espinar Merino, Ram\xf3n', u'Espinosa De La Llave, Mar\xeda', u'Fern\xe1ndez Rubi\xf1o, Eduardo', u'Garc\xeda G\xf3mez, M\xf3nica', 'Gimeno Reinoso, Beatriz', u'Guti\xe9rrez Benito, Eduardo', 'Huerta Bravo, Raquel', u'L\xf3pez Hern\xe1ndez, Isidro', u'L\xf3pez Rodrigo, Jos\xe9 Manuel', u'Mart\xednez Abarca, Hugo', u'Morano Gonz\xe1lez, Jacinto', u'Ongil L\xf3pez, Miguel', 'Padilla Estrada, Pablo', u'Ruiz-Huerta Garc\xeda De Viedma, Lorena', 'Salazar-Alonso Revuelta, Cecilia', u'San Jos\xe9 P\xe9rez, Carmen', u'S\xe1nchez P\xe9rez, Alejandro', u'Serra S\xe1nchez, Isabel', u'Serra S\xe1nchez, Clara', 'Sevillano De Las Heras, Elena']
[u'Aguado Crespo, Ignacio Jes\xfas', u'\xc1lvarez Cabo, Daniel', u'Gonz\xe1lez Pastor, Dolores', u'Iglesia Vicente, M\xaa Teresa De La', 'Lara Casanova, Francisco', u'Marb\xe1n De Frutos, Marta', u'Marcos Arias, Tom\xe1s', u'Meg\xedas Morales, Jes\xfas Ricardo', u'N\xfa\xf1ez S\xe1nchez, Roberto', 'Reyero Zubiri, Alberto', u'Rodr\xedguez Dur\xe1n, Ana', u'Rubio Ruiz, Juan Ram\xf3n', u'Ruiz Fern\xe1ndez, Esther', u'Sol\xeds P\xe9rez, Susana', 'Trinidad Martos, Juan', 'Veloso Lozano, Enrique', u'Zafra Hern\xe1ndez, C\xe9sar']

您可以为蜘蛛添加完全相同的逻辑，我只是使用请求向您展示一个有效的示例。您还应该知道，并非每个asp.net网站的行为都相同，您可能需要重新验证每个帖子，就像在此相关answer中一样。

Answer 2

我认为scrapy的from_response可以帮助你很多（也许这不是最好的，但是你会得到这个想法），尝试这样的事情：

import scrapy
import urllib
from scrapy.http.request.form import FormRequest


class AsambleaMadrid(scrapy.Spider):
    name = "Asamblea_Madrid"
    start_urls = ['http://www.asambleamadrid.es/ES/QueEsLaAsamblea/ComposiciondelaAsamblea/LosDiputados/Paginas/RelacionAlfabeticaDiputados.aspx']

    def parse(self, response):
        ids_re = r'WebForm_PostBackOptions\(([^,]*)'
        for id in response.css('#moduloBusqueda li a').re(ids_re):
            target = urllib.unquote(id).strip('"')
            formdata = {'__EVENTTARGET': target}
            request = FormRequest.from_response(response=response,
                                                formdata=formdata,
                                                callback=self.takeEachParty,
                                                dont_click=True)
            yield request

    def takeEachParty(self, response):
        print response.css('.listadoVert02 li a::text').extract()

Answer 3

同意ELRuLL - Firebug是你最好的朋友。如果你想避免JS模拟，那么你需要仔细重现正在发送的所有参数/标题。

例如我所看到的对于你刚刚发送的__EVENTTARGET id="ctl00_m_g_36ea0310_893d_4a19_9ed1_88a133d06423_ctl00_Repeater2_ctl01_lnk_Diputado")

并通过Firebug我们看到：

__EVENTTARGET=ctl00$m$g_36ea0310_893d_4a19_9ed1_88a133d06423$ctl00$Repeater2$ctl01$lnk_Diputado

这可能是原因，也许不是，只需重复和测试。

Firebug link以防万一。

Scrapy爬行不能在ASPX网站上运行

3 个答案: