如何在scrapy中集成selenium爬取網頁
1.背景
- 我們在爬取網頁時一般會使用到三個爬蟲庫:requests,scrapy,selenium。requests一般用於小型爬蟲,scrapy用於構建大的爬蟲項目,而selenium主要用來應付負責的頁面(複雜js渲染的頁面,請求非常難構造,或者構造方式經常變化)。
- 在我們面對大型爬蟲項目時,肯定會優選scrapy框架來開發,但是在解析複雜JS渲染的頁面時,又很麻煩。 儘管使用selenium瀏覽器渲染來抓取這樣的頁面很方便,這種方式下,我們不需要關心頁面後臺發生了怎樣的請求,也不需要分析整個頁面的渲染過程,我們只需要關心頁面最終結果即可,可見即可爬,但是selenium的效率又太低。
- 所以,如果可以在scrapy中,集成selenium,讓selenium負責複雜頁面的爬取,那麼這樣的爬蟲就無敵了,可以爬取任何網站了。
2. 環境
- python 3.6.1
- 系統:win7
- IDE:pycharm
- 安裝過chrome瀏覽器
- 配置好chromedriver(設置好環境變量)
- selenium 3.7.0
- scrapy 1.4.0
3.原理分析
3.1. 分析request請求的流程
首先看一下scrapy最新的架構圖(https://blog.csdn.net/zwq912318834/article/details/79720742):
部分流程:
- 第一:爬蟲引擎生成requests請求,送往scheduler調度模塊,進入等待隊列,等待調度。
- 第二:scheduler模塊開始調度這些requests,出隊,發往爬蟲引擎。
- 第三:爬蟲引擎將這些requests送到下載中間件(多個,例如加header,代理,自定義等等)進行處理。
- 第四:處理完之後,送往Downloader模塊進行下載。
- 從這個處理過程來看,突破口就在下載中間件部分,用selenium直接處理掉request請求。
3.2. requests和response中間處理件源碼分析
相關代碼位置:
源碼解析:
# 文件:E:\Miniconda\Lib\site-packages\scrapy\core\downloader\middleware.py
"""
Downloader Middleware manager
See documentation in docs/topics/downloader-middleware.rst
"""
import six
from twisted.internet import defer
from scrapy.http import Request, Response
from scrapy.middleware import MiddlewareManager
from scrapy.utils.defer import mustbe_deferred
from scrapy.utils.conf import build_component_list
class DownloaderMiddlewareManager(MiddlewareManager):
component_name = 'downloader middleware'
@classmethod
def _get_mwlist_from_settings(cls, settings):
# 從settings.py或這custom_setting中拿到自定義的Middleware中間件
'''
'DOWNLOADER_MIDDLEWARES': {
'mySpider.middlewares.ProxiesMiddleware': 400,
# SeleniumMiddleware
'mySpider.middlewares.SeleniumMiddleware': 543,
'scrapy.downloadermiddlewares.useragent.UserAgentMiddleware': None,
},
'''
return build_component_list(
settings.getwithbase('DOWNLOADER_MIDDLEWARES'))
# 將所有自定義Middleware中間件的處理函數添加到對應的methods列表中
def _add_middleware(self, mw):
if hasattr(mw, 'process_request'):
self.methods['process_request'].append(mw.process_request)
if hasattr(mw, 'process_response'):
self.methods['process_response'].insert(0, mw.process_response)
if hasattr(mw, 'process_exception'):
self.methods['process_exception'].insert(0, mw.process_exception)
# 整個下載流程
def download(self, download_func, request, spider):
@defer.inlineCallbacks
def process_request(request):
# 處理request請求,依次經過各個自定義Middleware中間件的process_request方法,前面有加入到list中
for method in self.methods['process_request']:
response = yield method(request=request, spider=spider)
assert response is None or isinstance(response, (Response, Request)), \
'Middleware %s.process_request must return None, Response or Request, got %s' % \
(six.get_method_self(method).__class__.__name__, response.__class__.__name__)
# 這是關鍵地方
# 如果在某個Middleware中間件的process_request中處理完之後,生成了一個response對象
# 那麼會直接將這個response return 出去,跳出循環,不再處理其他的process_request
# 之前我們的header,proxy中間件,都只是加個user-agent,加個proxy,並不做任何return值
# 還需要注意一點:就是這個return的必須是Response對象
# 後面我們構造的HtmlResponse正是Response的子類對象
if response:
defer.returnValue(response)
# 如果在上面的所有process_request中,都沒有返回任何Response對象的話
# 最後,會將這個加工過的Request送往download_func,進行下載,返回的就是一個Response對象
# 然後依次經過各個Middleware中間件的process_response方法進行加工,如下
defer.returnValue((yield download_func(request=request,spider=spider)))
@defer.inlineCallbacks
def process_response(response):
assert response is not None, 'Received None in process_response'
if isinstance(response, Request):
defer.returnValue(response)
for method in self.methods['process_response']:
response = yield method(request=request, response=response,
spider=spider)
assert isinstance(response, (Response, Request)), \
'Middleware %s.process_response must return Response or Request, got %s' % \
(six.get_method_self(method).__class__.__name__, type(response))
if isinstance(response, Request):
defer.returnValue(response)
defer.returnValue(response)
@defer.inlineCallbacks
def process_exception(_failure):
exception = _failure.value
for method in self.methods['process_exception']:
response = yield method(request=request, exception=exception,
spider=spider)
assert response is None or isinstance(response, (Response, Request)), \
'Middleware %s.process_exception must return None, Response or Request, got %s' % \
(six.get_method_self(method).__class__.__name__, type(response))
if response:
defer.returnValue(response)
defer.returnValue(_failure)
deferred = mustbe_deferred(process_request, request)
deferred.addErrback(process_exception)
deferred.addCallback(process_response)
return deferred
4. 代碼
- 在settings.py中,配置好selenium參數:
# 文件settings.py中
# ----------- selenium參數配置 -------------
SELENIUM_TIMEOUT = 25 # selenium瀏覽器的超時時間,單位秒
LOAD_IMAGE = True # 是否下載圖片
WINDOW_HEIGHT = 900 # 瀏覽器窗口大小
WINDOW_WIDTH = 900
- 在spider中,生成request時,標記哪些請求需要走selenium下載:
# 文件mySpider.py中
class mySpider(CrawlSpider):
name = "mySpiderAmazon"
allowed_domains = ['amazon.com']
custom_settings = {
'LOG_LEVEL':'INFO',
'DOWNLOAD_DELAY': 0,
'COOKIES_ENABLED': False, # enabled by default
'DOWNLOADER_MIDDLEWARES': {
# 代理中間件
'mySpider.middlewares.ProxiesMiddleware': 400,
# SeleniumMiddleware 中間件
'mySpider.middlewares.SeleniumMiddleware': 543,
# 將scrapy默認的user-agent中間件關閉
'scrapy.downloadermiddlewares.useragent.UserAgentMiddleware': None,
},
#.....................華麗的分割線.......................
# 生成request時,將是否使用selenium下載的標記,放入到meta中
yield Request(
url = "https://www.amazon.com/",
meta = {'usedSelenium': True, 'dont_redirect': True},
callback = self.parseIndexPage,
errback = self.error
)
- 在下載中間件middlewares.py中,使用selenium抓取頁面(核心部分)
# -*- coding: utf-8 -*-
from selenium import webdriver
from selenium.common.exceptions import TimeoutException
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.common.keys import Keys
from scrapy.http import HtmlResponse
from logging import getLogger
import time
class SeleniumMiddleware():
# 經常需要在pipeline或者中間件中獲取settings的屬性,可以通過scrapy.crawler.Crawler.settings屬性
@classmethod
def from_crawler(cls, crawler):
# 從settings.py中,提取selenium設置參數,初始化類
return cls(timeout=crawler.settings.get('SELENIUM_TIMEOUT'),
isLoadImage=crawler.settings.get('LOAD_IMAGE'),
windowHeight=crawler.settings.get('WINDOW_HEIGHT'),
windowWidth=crawler.settings.get('WINDOW_WIDTH')
)
def __init__(self, timeout=30, isLoadImage=True, windowHeight=None, windowWidth=None):
self.logger = getLogger(__name__)
self.timeout = timeout
self.isLoadImage = isLoadImage
# 定義一個屬於這個類的browser,防止每次請求頁面時,都會打開一個新的chrome瀏覽器
# 這樣,這個類處理的Request都可以只用這一個browser
self.browser = webdriver.Chrome()
if windowHeight and windowWidth:
self.browser.set_window_size(900, 900)
self.browser.set_page_load_timeout(self.timeout) # 頁面加載超時時間
self.wait = WebDriverWait(self.browser, 25) # 指定元素加載超時時間
def process_request(self, request, spider):
'''
用chrome抓取頁面
:param request: Request請求對象
:param spider: Spider對象
:return: HtmlResponse響應
'''
# self.logger.debug('chrome is getting page')
print(f"chrome is getting page")
# 依靠meta中的標記,來決定是否需要使用selenium來爬取
usedSelenium = request.meta.get('usedSelenium', False)
if usedSelenium:
try:
self.browser.get(request.url)
# 搜索框是否出現
input = self.wait.until(
EC.presence_of_element_located((By.XPATH, "//div[@class='nav-search-field ']/input"))
)
time.sleep(2)
input.clear()
input.send_keys("iphone 7s")
# 敲enter鍵, 進行搜索
input.send_keys(Keys.RETURN)
# 查看搜索結果是否出現
searchRes = self.wait.until(
EC.presence_of_element_located((By.XPATH, "//div[@id='resultsCol']"))
)
except Exception as e:
# self.logger.debug(f'chrome getting page error, Exception = {e}')
print(f"chrome getting page error, Exception = {e}")
return HtmlResponse(url=request.url, status=500, request=request)
else:
time.sleep(3)
return HtmlResponse(url=request.url,
body=self.browser.page_source,
request=request,
# 最好根據網頁的具體編碼而定
encoding='utf-8',
status=200)
5. 執行結果
6. 存在的問題
6.1. Spider關閉了,chrome沒有退出。
2018-04-04 09:26:18 [scrapy.statscollectors] INFO: Dumping Scrapy stats:
{'downloader/response_bytes': 2092766,
'downloader/response_count': 2,
'downloader/response_status_count/200': 2,
'finish_reason': 'finished',
'finish_time': datetime.datetime(2018, 4, 4, 1, 26, 16, 763602),
'log_count/INFO': 7,
'request_depth_max': 1,
'response_received_count': 2,
'scheduler/dequeued': 2,
'scheduler/dequeued/memory': 2,
'scheduler/enqueued': 2,
'scheduler/enqueued/memory': 2,
'start_time': datetime.datetime(2018, 4, 4, 1, 25, 48, 301602)}
2018-04-04 09:26:18 [scrapy.core.engine] INFO: Spider closed (finished)
- 上面,我們是把browser對象放到了Middleware中間件中,只能做process_request和process_response, 沒有說在中間件中介紹如何調用scrapy的close方法。
- 解決方案:利用信號量的方式,當收到spider_closed信號時,調用browser.quit()
6.2. 當一個項目同時啓動多個spider,會共用到Middleware中的selenium,不利於併發。
- 因爲用scrapy + selenium的方式,只有部分,甚至是一小部分頁面會用到chrome,既然把chrome放到Middleware中有這麼多限制,那爲什麼不能把chrome放到spider裏面呢。這樣的好處在於:每個spider都有自己的chrome,這樣當啓動多個spider時,就會有多個chrome,不是所有的spider共用一個chrome,這對我們的併發是有好處的。
- 解決方案:將chrome的初始化放到spider中,每個spider獨佔自己的chrome
7. 改進版代碼
- 在settings.py中,配置好selenium參數:
# 文件settings.py中
# ----------- selenium參數配置 -------------
SELENIUM_TIMEOUT = 25 # selenium瀏覽器的超時時間,單位秒
LOAD_IMAGE = True # 是否下載圖片
WINDOW_HEIGHT = 900 # 瀏覽器窗口大小
WINDOW_WIDTH = 900
- 在spider中,生成request時,標記哪些請求需要走selenium下載:
# 文件mySpider.py中
# selenium相關庫
from selenium import webdriver
from selenium.webdriver.support.ui import WebDriverWait
# scrapy 信號相關庫
from scrapy.utils.project import get_project_settings
# 下面這種方式,即將廢棄,所以不用
# from scrapy.xlib.pydispatch import dispatcher
from scrapy import signals
# scrapy最新採用的方案
from pydispatch import dispatcher
class mySpider(CrawlSpider):
name = "mySpiderAmazon"
allowed_domains = ['amazon.com']
custom_settings = {
'LOG_LEVEL':'INFO',
'DOWNLOAD_DELAY': 0,
'COOKIES_ENABLED': False, # enabled by default
'DOWNLOADER_MIDDLEWARES': {
# 代理中間件
'mySpider.middlewares.ProxiesMiddleware': 400,
# SeleniumMiddleware 中間件
'mySpider.middlewares.SeleniumMiddleware': 543,
# 將scrapy默認的user-agent中間件關閉
'scrapy.downloadermiddlewares.useragent.UserAgentMiddleware': None,
},
# 將chrome初始化放到spider中,成爲spider中的元素
def __init__(self, timeout=30, isLoadImage=True, windowHeight=None, windowWidth=None):
# 從settings.py中獲取設置參數
self.mySetting = get_project_settings()
self.timeout = self.mySetting['SELENIUM_TIMEOUT']
self.isLoadImage = self.mySetting['LOAD_IMAGE']
self.windowHeight = self.mySetting['WINDOW_HEIGHT']
self.windowWidth = self.mySetting['windowWidth']
# 初始化chrome對象
self.browser = webdriver.Chrome()
if self.windowHeight and self.windowWidth:
self.browser.set_window_size(900, 900)
self.browser.set_page_load_timeout(self.timeout) # 頁面加載超時時間
self.wait = WebDriverWait(self.browser, 25) # 指定元素加載超時時間
super(mySpider, self).__init__()
# 設置信號量,當收到spider_closed信號時,調用mySpiderCloseHandle方法,關閉chrome
dispatcher.connect(receiver = self.mySpiderCloseHandle,
signal = signals.spider_closed
)
# 信號量處理函數:關閉chrome瀏覽器
def mySpiderCloseHandle(self, spider):
print(f"mySpiderCloseHandle: enter ")
self.browser.quit()
#.....................華麗的分割線.......................
# 生成request時,將是否使用selenium下載的標記,放入到meta中
yield Request(
url = "https://www.amazon.com/",
meta = {'usedSelenium': True, 'dont_redirect': True},
callback = self.parseIndexPage,
errback = self.error
)
- 在下載中間件middlewares.py中,使用selenium抓取頁面
# -*- coding: utf-8 -*-
from selenium import webdriver
from selenium.common.exceptions import TimeoutException
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.common.keys import Keys
from scrapy.http import HtmlResponse
from logging import getLogger
import time
class SeleniumMiddleware():
# Middleware中會傳遞進來一個spider,這就是我們的spider對象,從中可以獲取__init__時的chrome相關元素
def process_request(self, request, spider):
'''
用chrome抓取頁面
:param request: Request請求對象
:param spider: Spider對象
:return: HtmlResponse響應
'''
print(f"chrome is getting page")
# 依靠meta中的標記,來決定是否需要使用selenium來爬取
usedSelenium = request.meta.get('usedSelenium', False)
if usedSelenium:
try:
spider.browser.get(request.url)
# 搜索框是否出現
input = spider.wait.until(
EC.presence_of_element_located((By.XPATH, "//div[@class='nav-search-field ']/input"))
)
time.sleep(2)
input.clear()
input.send_keys("iphone 7s")
# 敲enter鍵, 進行搜索
input.send_keys(Keys.RETURN)
# 查看搜索結果是否出現
searchRes = spider.wait.until(
EC.presence_of_element_located((By.XPATH, "//div[@id='resultsCol']"))
)
except Exception as e:
print(f"chrome getting page error, Exception = {e}")
return HtmlResponse(url=request.url, status=500, request=request)
else:
time.sleep(3)
# 頁面爬取成功,構造一個成功的Response對象(HtmlResponse是它的子類)
return HtmlResponse(url=request.url,
body=spider.browser.page_source,
request=request,
# 最好根據網頁的具體編碼而定
encoding='utf-8',
status=200)
- 運行結果(spider結束,執行mySpiderCloseHandle關閉chrome瀏覽器):
['categorySelectorAmazon1.pipelines.MongoPipeline']
2018-04-04 11:56:21 [scrapy.core.engine] INFO: Spider opened
2018-04-04 11:56:21 [scrapy.extensions.logstats] INFO: Crawled 0 pages (at 0 pages/min), scraped 0 items (at 0 items/min)
chrome is getting page
parseProductDetail url = https://www.amazon.com/, status = 200, meta = {'usedSelenium': True, 'dont_redirect': True, 'download_timeout': 25.0, 'proxy': 'http://H37XPSB6V57VU96D:[email protected]:9020', 'depth': 0}
chrome is getting page
2018-04-04 11:56:54 [scrapy.core.engine] INFO: Closing spider (finished)
mySpiderCloseHandle: enter
2018-04-04 11:56:59 [scrapy.statscollectors] INFO: Dumping Scrapy stats:
{'downloader/response_bytes': 1938619,
'downloader/response_count': 2,
'downloader/response_status_count/200': 2,
'finish_reason': 'finished',
'finish_time': datetime.datetime(2018, 4, 4, 3, 56, 54, 301602),
'log_count/INFO': 7,
'request_depth_max': 1,
'response_received_count': 2,
'scheduler/dequeued': 2,
'scheduler/dequeued/memory': 2,
'scheduler/enqueued': 2,
'scheduler/enqueued/memory': 2,
'start_time': datetime.datetime(2018, 4, 4, 3, 56, 21, 642602)}
2018-04-04 11:56:59 [scrapy.core.engine] INFO: Spider closed (finished)