一、背景:
每年终都有一个习惯,就是整理资料进行归档,结果发现手机照片全备份在华为云里,在官网上找了一圈,没找到官方的pc工具用来同步照片。
于是找出上次写的程序,看看能不能爬到数据,然而……果然不好用。因为华为在登录上又增加了一些验证机制,譬如:账号保护
抓了一下报文,发现逻辑变复杂了很多,部分逻辑还封装在js里。
算了,懒得琢磨了,直接用selenium吧。
二、实现思路:
1、用Python + selenium +浏览器 ,人工登录,保存cookie及签名信息。
2、再调用requests加第一步保存的cookie和前面,直接向后台发post请求,获取数据。
思路确定,开干。
三、开发环境:
1、python3.6,在最近的一个项目中由于多次遇到中文问题,实在是烦不胜烦,所以就把开发工具升级到了py3,确实方便多了。
说到py2升到py3,虽然还是有些写法调整,有些包在py3下不支持,但总体来说,迁移很平稳,写法问题,百度一下基本就可以解决。
我用的Anaconda的python包。
3.6.3 |Anaconda custom (64-bit)| (default, Oct 15 2017, 03:27:45) [MSC v.1900 64 bit (AMD64)] Python Type "help", "copyright", "credits" or "license" for more information.
2、selenium 3.9.0,用conda现安装的。
conda install selenium
3、浏览器,试用了firefox,edge,chrome,phantomjs,分别版本如下:
firefox: 58.0.2 (64 位) edge: Microsoft Edge 41.16299.248.0 ,Microsoft Edge 41.16299.248.0 chrome: 版本 63.0.3239.132(正式版本) (32 位) phantomjs: 2.1.1 另外,操作系统:Microsoft Windows [版本 10.0.16299.248]
4、浏览器驱动:
firefox驱动,https://github.com/mozilla/geckodriver/releases/,支持 Firefox 55及以上版本。
edge驱动,https://developer.microsoft.com/en-us/microsoft-edge/tools/webdriver/#downloads,最新版本 Release 16299,Version: 5.16299,支持 Edge version supported: 16.16299 。注意edge驱动只有在edge浏览器未启动的情况下才能正常运行,否则会报错。
chrome驱动,https://sites.google.com/a/chromium.org/chromedriver/downloads,这里需要注意的是:最新版本是2.35(不是2.9),2.35才支持chrome 61-63版本。
phantomjs,http://phantomjs.org/download.html,phantomjs可以理解成没有界面的浏览器,所以驱动跟浏览器是一体的。
驱动版本一定要选对,否则会有奇奇怪怪的问题。
四、实现代码
huaweiphoto_sele.py,如下:
#-*-coding=utf-8-*- #Create by : zhongtang #Create Date : 2018.2.28 from selenium import webdriver from selenium.webdriver.common.desired_capabilities import DesiredCapabilities from selenium.webdriver.common.proxy import ProxyType from selenium.webdriver.support.ui import WebDriverWait from selenium.webdriver.support import expected_conditions as EC from selenium.webdriver.common.by import By from PIL import Image import json,re,os,time,requests,socket #下载函数 from huaweiphoto_py3 import HuaWei class hwSele: SeleBrowser=None TimeOUT=30 Headers=None Username=\'*****\' Passwd=\'****\' DriverType="Edge".lower() def __init__(self,ip=None,port=None,SeleDriver="Edge",SeleHeader=None): print (u\'proxy %s %s...\' %(ip,port)) if not SeleHeader : self.Headers = "Mozilla/5.0 (Windows NT 10.0; WOW64; rv:56.0) Gecko/20100101 Firefox/56.0" else: self.Headers = SeleHeader if SeleDriver: self.DriverType= SeleDriver.lower() #加代理的目的是为了更便于抓报文。 if self.DriverType==\'chrome\' : chromeOptions = webdriver.ChromeOptions() if ip: chromeOptions.add_argument(\'--proxy-server=http://%s:%s\' %(ip,port)) self.SeleBrowser = webdriver.Chrome(chrome_options=chromeOptions) else: self.SeleBrowser = webdriver.Chrome() #DriverType=\'Edge\' elif self.DriverType==\'phantomjs\': #设置userAgent dcap = dict(DesiredCapabilities.PHANTOMJS) dcap["phantomjs.page.settings.userAgent"] = (self.Headers) self.SeleBrowser = webdriver.PhantomJS(executable_path=r\'D:\python\toupiao\phantomjs\bin\phantomjs.exe\',desired_capabilities=dcap) if ip: proxy=webdriver.Proxy() proxy.proxy_type=ProxyType.MANUAL proxy.http_proxy=\'%s:%s\' %(ip,port) proxy.add_to_capabilities(webdriver.DesiredCapabilities.PHANTOMJS) else: self.SeleBrowser.start_session(webdriver.DesiredCapabilities.PHANTOMJS) elif self.DriverType==\'edge\': self.KillSeleProc() #edge,默认先kill掉已启动的浏览器。 self.SeleBrowser = webdriver.Edge() elif self.DriverType==\'firefox\': webdriver.DesiredCapabilities.FIREFOX[\'firefox.page.settings.userAgent\'] = self.Headers profile = webdriver.FirefoxProfile() if ip: profile.set_preference(\'network.proxy.type\', 1) # 默认值0,就是直接连接;1就是手工配置代理。 profile.set_preference(\'network.proxy.http\', ip) profile.set_preference(\'network.proxy.http_port\', port) profile.set_preference(\'network.proxy.ssl\', ip) profile.set_preference(\'network.proxy.ssl_port\', port) profile.update_preferences() self.SeleBrowser = webdriver.Firefox(profile) else: self.SeleBrowser = webdriver.Firefox() socket.setdefaulttimeout(self.TimeOUT) # 设置10秒页面超时返回,类似于requests.get()的timeout选项,driver.get()没有timeout选项 # 以前遇到过driver.get(url)一直不返回,但也不报错的问题,这时程序会卡住,设置超时选项能解决这个问题。 self.SeleBrowser.set_page_load_timeout(self.TimeOUT) # 设置10秒脚本超时时间 self.SeleBrowser.set_script_timeout(self.TimeOUT) # 隐式等待30秒,可以自己调节 self.SeleBrowser.implicitly_wait(self.TimeOUT) def KillSeleProc(self): if self.DriverType==\'edge\': command = \'taskkill /F /IM MicrosoftWebDriver.exe & taskkill /F /IM MicrosoftEdge.exe\' #比如这里关闭edge进程 elif self.DriverType==\'chrome\': command = \'taskkill /F /IM chromedriver.exe & taskkill /F /IM chrome.exe\' elif self.DriverType==\'firefox\': command = \'taskkill /F /IM geckodriver.exe & taskkill /F /IM firefox.exe\' elif self.DriverType=="phantomjs": command = \'taskkill /F /IM phantomjs.exe \' if command: os.system(command) def QuitSele(self,e,mess=None,iRet= -1): print (mess,e) if self.SeleBrowser: self.SeleBrowser.save_screenshot(\'error.png\') self.SeleBrowser.close() self.KillSeleProc() return iRet def LoginHW(self): \'\'\' try: element = WebDriverWait(driver, 10).until(EC.presence_of_element_located((By.ID, "loadedButton"))) finally: print(driver.find_element_by_id("content").text) driver.close() #等待页面加载完毕1,显示等待 try: auth_img = WebDriverWait(self.SeleBrowser, 5).until(EC.presence_of_element_located((By.ID, "randomCodeImg"))) except Exception as e: print (u\'加载验证码超时...\',e) SeleBrowser.save_screenshot(r\'d:\python\toupiao\error.jpg\') self.SeleBrowser.close() return -1 #等待页面加载完毕2,隐式等待 dr=WebDriverWait(self.SeleBrowser,20,0.5) dr.until(lambda the_driver:the_driver.find_element_by_xpath("//img[@id=\'randomCodeImg\']").is_displayed()) \'\'\' try: self.SeleBrowser.get(\'http://cloud.huawei.com\') except Exception as e: return self.QuitSele(e,"打开主页出错!") try: #等待页面加载完毕 dr=WebDriverWait(self.SeleBrowser,self.TimeOUT,0.5) dr.until(lambda the_driver:the_driver.find_element_by_id("randomCodeImg").is_displayed()) except Exception as e: return self.QuitSele(e,"加载验证码超时!") elem_user = self.SeleBrowser.find_element_by_id("login_userName") elem_user.clear() elem_user.send_keys(self.Username) elem_pwd = self.SeleBrowser.find_element_by_id("login_password") elem_pwd.clear() elem_pwd.send_keys(self.Passwd) auth_img = self.SeleBrowser.find_element_by_id("randomCodeImg") if not auth_img.is_displayed() : if not auth_img.is_displayed(): return self.QuitSele(e,"验证码未正常显示!") if self.DriverType==\'firefox\': #firefox驱动支持直接 元素另存图片 auth_img.screenshot("captcha.png") im = Image.open(\'captcha.png\') else: #chrome ,edge 都不支持,phantomjs存的还是整个窗口 self.SeleBrowser.save_screenshot(\'captcha.png\') im = Image.open(\'captcha.png\') x= eval(auth_img.get_attribute("x")) y= eval(auth_img.get_attribute("y")) width= eval(auth_img.get_attribute("width")) height= eval(auth_img.get_attribute("height")) im = im.crop((x, y, x+width, y+height)) #这里采用最原始、最准确的方法:显示图片,人工识别^_^,智能输入验证码。 #当然也可以调用三方的图像识别api进行识别,譬如pytesseract或者鹅厂的图像识别api,不复杂,但懒得写了。 im.show() authCode= input(u\'请输入验证码:\') # 先获取焦点,再赋值,再点击登录 \'\'\' js= \'$("#randomCode").attr("value","%s");$("#randomCode").trigger("onchange");\' %authCode self.SeleBrowser.execute_script(js) js= \'$("#btnLogin").trigger("click");\' self.SeleBrowser.execute_script(js) \'\'\' randomCode = self.SeleBrowser.find_element_by_id("randomCode") randomCode.clear() randomCode.send_keys(authCode) #休息五秒,等待完成后台预验证交互 time.sleep(5) btnLogin = self.SeleBrowser.find_element_by_id("btnLogin") btnLogin.click() #账号保护有时候会提示 \'\'\' <div class="global_dialog_confirm_main" style="display: block; margin-top: -163.5px;"> <div class="global_dialog_confirm_title"> <h3 class="ellipsis" title="帐号保护">帐号保护</h3> </div> <div class="global_dialog_confirm_content" style="padding-bottom: 0px;"><div> <div id="authenDialog"><p class="inptips2">您已开启帐号保护,请输入验证码以完成登录。</p> <div class="margin10-EMUI5"><div id="accountDiv" class="fixAccountDrt ddrop-EMU5"> \'\'\' try : #loginConfirm = self.SeleBrowser.find_element_by_class_name("global_dialog_confirm_main") loginConfirm =WebDriverWait(self.SeleBrowser, 5, 0.5).until(EC.presence_of_element_located((By.CLASS_NAME, \'global_dialog_confirm_main\') )) #需要验证,这块懒得实现了,休眠60秒,手动操作吧。 if loginConfirm.is_displayed(): time.sleep(self.TimeOUT*2) except: #不需要验证,直接下一步 pass #等待页面加载完毕 \'\'\' <span class="index-span" data-bind="lang.common.album">图库</span> \'\'\' try : #loginConfirm = self.SeleBrowser.find_element_by_class_name("global_dialog_confirm_main") success =WebDriverWait(self.SeleBrowser, 20, 0.5).until(EC.presence_of_element_located((By.XPATH, \'//span[@data-bind="lang.common.album"]\') )) except Exception as e: #登录失败 return self.QuitSele(e,"登录失败!",iRet=-999) #判断登录结果 if not success.is_displayed(): return self.QuitSele(None,"登录失败!",iRet=-999) #再次判断,增加一次意外处理 source_code =self.SeleBrowser.page_source if \'联系人\' not in source_code or \'图库\' not in source_code : return self.QuitSele(None,"登录失败!",iRet = -9999 ) cookie = [item["name"] + "=" + item["value"] for item in self.SeleBrowser.get_cookies()] cookiestr = \';\'.join(item for item in cookie) #保存CSRFToken pattern = re.compile(\'CSRFToken = "(.*?)"\',re.S) content = re.search(pattern,source_code) if content : CSRFToken = content.group(1) else : print (\'获取CSRFToken出错!\') self.Headers={ \'User-Agent\': \'%s\' %self.Headers, \'CSRFToken\': \'%s\' %CSRFToken, \'Cookie\': \'%s\' %cookiestr } return 1 if __name__ == \'__main__\': photohw= HuaWei() count =0 while (count <100): count += 1 selehw= hwSele(SeleDriver=\'edge\') iRet = selehw.LoginHW() if iRet !=1: print( \'登录华为失败!!!\n\n\') continue photohw.loginHeaders = selehw.Headers page = photohw.getAlbumList() if page==\'\' : print( \'获取到相册列表失败!!!\n\n\') break #保存相册列表 iRet = photohw.getFileList(page,\'albumList\',\'albumId\') if iRet <=0 : print(\'保存相册出错,重新登录\') continue #保存公共相册列表 iRet = photohw.getFileList(page,\'ownShareList\',\'shareId\') if iRet ==0 : print(\'运行结束,可以用迅雷打开相册文件进行批量下载到本地!!!\n\n\') #运行结束 selehw.QuitSele(None) break else: continue
huaweiphoto_py3.py如下:
# -*- coding=utf-8 -*- # Create by : zhongtang # Create date : 2018.2.28 import json import requests from requests.adapters import HTTPAdapter import html class HuaWei: #华为云服务登录 def __init__(self): self.getalbumsUrl= \'https://www.hicloud.com/album/getCloudAlbums.action\' self.getalbumfileUrl = \'https://www.hicloud.com/album/getCloudFiles.action\' self.loginHeaders = { } self.SReq=requests.session() self.SReq.mount(\'http://\', HTTPAdapter(max_retries=3)) self.SReq.mount(\'https://\', HTTPAdapter(max_retries=3)) self.OnceMaxFile=100 #单次最大获取文件数量 self.FileNum=0 self.AlbumList={} #保存相册照片地址到文件 ,不同相册保存到不同的文件 def saveFileList2Txt(self,filename,hjsondata,flag): if len(hjsondata)<= 0 : return -1 hjson2 = {} try: hjson2 = json.loads(hjsondata) except: print(\'获取相册明细出错\n\') return -1 lfilename = filename+u".txt" if flag == 0 : #新建文件 print( u\'创建相册文件\'+lfilename+"\n") #新建文件,代表新的相册重新开始计数 self.FileNum = 0 f = open(lfilename, \'w\') else: #追加文件 f = open(lfilename, \'a\') i = 0 if hjson2.get("fileList"): for each in hjson2["fileList"]: fileurl= html.unescape(hjson2["fileList"][i]["fileUrl"]) f.write(fileurl+"\n") #每一千行分页 self.FileNum += 1 if self.FileNum%1000 ==0 :f.write(\'\n\n\n\n\n\n--------------------page %s ------------------\n\n\n\n\n\n\' %(int(self.FileNum/1000))) i += 1 f.close() return i #循环读取相册文件 def getFileList(self,hjsondata,parentkey,childkey): #step 3 getCoverFiles.action,循环取相册文件列表,单次最多取100条记录。 #每次count都是最大数量49,不管实际数量是否够,每次currentnum递增,直到返回空列表。 #albumIds[]=default-album-2&ownerId=220086000029851117&height=300&width=300&count=49¤tNum=0&thumbType=imgcropa&fileType=0 #albumIds[]=default-album-1&ownerId=220086000029851117&height=300&width=300&count=49¤tNum=49&thumbType=imgcropa&fileType=0 #albumIds[]=default-album-1&ownerId=220086000029851117&height=300&width=300&count=49¤tNum=98&thumbType=imgcropa&fileType=0 #albumIds[]=default-album-2&ownerId=220086000029851117&height=300&width=300&count=49¤tNum=101&thumbType=imgcropa&fileType=0 #最后一次返回 空列表 #{"albumSortFlag":true,"code":0,"info":"success!","fileList":[]} #第一次取文件时,例如文件总数量只有2个,count也是放最大值49。 #albumIds[]=default-album-102-220086000029851117&ownerId=220086000029851117&height=300&width=300&count=49¤tNum=0&thumbType=imgcropa&fileType=0 #[{u\'photoNum\': 2518, u\'albumName\': u\'default-album-1\', u\'iversion\': -1, u\'albumId\': u\'default-album-1\', u\'flversion\': -1, u\'createTime\': 1448065264550L, u\'size\': 0}, #{u\'photoNum\': 100, u\'albumName\': u\'default-album-2\', u\'iversion\': -1, u\'albumId\': u\'default-album-2\', u\'flversion\': -1, u\'createTime\': 1453090781646L, u\'size\': 0}] try: hjson = json.loads(hjsondata) except Exception: print (\'加载json出错!\') return -1 #字典获取出错 if not hjson.get(parentkey): print (\'加载json根节点[%s]出错!\' %parentkey) return -1 #初始化全局 albumlist if not self.AlbumList : self.AlbumList=hjson for idx,album in enumerate(self.AlbumList[parentkey]): if \'currentNum\' not in self.AlbumList[parentkey][idx].keys(): self.AlbumList[parentkey][idx][\'currentNum\']=0 #循环保存相册 for each in hjson[parentkey]: #该相册已经进入记录 paraAlbum={} paraAlbum[\'albumIds[]\'] = each[childkey] paraAlbum[\'ownerId\'] = hjson[\'ownerId\'] paraAlbum[\'height\'] = \'300\' paraAlbum[\'width\'] = \'300\' paraAlbum[\'count\'] = self.OnceMaxFile paraAlbum[\'thumbType\'] = \'imgcropa\' paraAlbum[\'fileType\'] = \'0\' itotal= each[\'photoNum\'] #取当前节点的当前记录 for idx,album in enumerate(self.AlbumList[parentkey]): if each[childkey]==album[childkey]: icurrentnum = self.AlbumList[parentkey][idx][\'currentNum\'] break #保存相册中所有文件 while icurrentnum<itotal: paraAlbum[\'currentNum\'] = icurrentnum response=self.SReq.post(self.getalbumfileUrl,headers=self.loginHeaders,data=paraAlbum,verify=False) page = response.text #保存下载地址到文本文件中,但不下载文件 iret = self.saveFileList2Txt(each[childkey],page,icurrentnum) if iret >0 : self.AlbumList[parentkey][idx][\'currentNum\'] += iret icurrentnum = self.AlbumList[parentkey][idx][\'currentNum\'] else: #出错!!! return -1 return 1 #step 1 getCloudAlbums,取相册列表 def getAlbumList(self): response=self.SReq.post(self.getalbumsUrl,headers=self.loginHeaders,verify=False) page=response.text \'\'\'#返回报文 {"ownerId":"220086000029851117","code":0, "albumList":[{"albumId":"default-album-1","albumName":"default-album-1","createTime":1448065264550,"photoNum":2521,"flversion":-1,"iversion":-1,"size":0}, {"albumId":"default-album-2","albumName":"default-album-2","createTime":1453090781646,"photoNum":101,"flversion":-1,"iversion":-1,"size":0}], "ownShareList":[{"ownerId":"220086000029851117","resource":"album","shareId":"default-album-102-220086000029851117","shareName":"微信","photoNum":2,"flversion":-1,"iversion":-1,"createTime":1448070407055,"source":"HUAWEI MT7-TL00","size":0,"ownerAcc":"****","receiverList":[]}], "recShareList":[]}\' \'\'\' if len(page)<=0 : print( u\'取相册列表出错,无返回报文!!!\n\n\') return page
五、运行结果:
程序会在当前目录生成华为云相册照片下载地址文件,内容如下:
https://d167.g03.dbankcloud.com/file/MDAwMTZBODissQaaaaaaaaaaaaaaaaaaaaQc2CR-znjyRnw../162807b277aaaaaaaaaaaaaaaaaa9ee1/IMG_20170606_141952.jpg?key=AWqIQFqVkEaaaaaaaaaaaaaaaaaaaaaaWNLIosPR_EKv8VQ..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
https://d167.g03.dbankcloud.com/file/MDAwMTZBODhhoFaaaaaaaaaaaaaaaaaaaa7r6jPU67bWTQA../4039a1be5caaaaaaaaaaaaaaaaaac726/IMG_20170605_203519.jpg?key=AWqIQFqVkEaaaaaaaaaaaaaaaaaaaaaajvduIL8cXufhNhQ..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
https://d167.g03.dbankcloud.com/file/MDAwMTZBODgpoWaaaaaaaaaaaaaaaaaaaaaciAQlIVHRbXg../9e336da286aaaaaaaaaaaaaaaaaaf89d/IMG_20170604_171032.jpg?key=AWqIQFqVkEaaaaaaaaaaaaaaaaaaaaaaThkDgJKHpBtiG5w..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
https://d167.g03.dbankcloud.com/file/MDAwMTZBODgUz2aaaaaaaaaaaaaaaaaaaatyFMDr71YpXGg../b3c17582ccaaaaaaaaaaaaaaaaac278b/IMG_20170603_134831.jpg?key=AWqIQFqVkEaaaaaaaaaaaaaaaaaaaaaaTI_xPSzF_VzUsJA..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
https://d167.g03.dbankcloud.com/file/MDAwMTZBODgGfwaaaaaaaaaaaaaaaaaaaad7DLSHH4rwKVA../2722df087baaaaaaaaaaaaaaaaa915e4/IMG_20170603_133833.jpg?key=AWqIQFqVkEaaaaaaaaaaaaaaaaaaaaaaluTQ8grDHok9BzQ..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
https://d167.g03.dbankcloud.com/file/MDAwMTZBODgsrHaaaaaaaaaaaaaaaaaaaatq2yOJ-OnkDtQ../77e0ef0560aaaaaaaaaaaaaaaaa44702/IMG_20170602_183736.jpg?key=AWqIQFqVkEaaaaaaaaaaaaaaaaaaaaaa20WJbfxn-qoqIeQ..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
https://d167.g03.dbankcloud.com/file/MDAwMTZBODiAcDaaaaaaaaaaaaaaaaaaaaEXW-ONoF0Shuw../df033e69ffaaaaaaaaaaaaaaaaa8c1b1/IMG_20170601_185446.jpg?key=AWqIQFqVkEaaaaaaaaaaaaaaaaaaaaaaxIC-spleDG_xxVg..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
https://d167.g03.dbankcloud.com/file/MDAwMTZBODjY8EaaaaaaaaaaaaaaaaaaaaVbr9kC-JU0M8g../d5230d2032aaaaaaaaaaaaaaaaa903b3/IMG_20170601_102059.jpg?key=AWqIQFqVkEaaaaaaaaaaaaaaaaaaaaaa49iM03bK-Cm-Z9g..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
https://d167.g03.dbankcloud.com/file/MDAwMTZBODi04Caaaaaaaaaaaaaaaaaaaaaaw41bSNB4pxBw../6ee510e28aaaaaaaaaaaaaaaaa57cb5a/IMG_20170601_102042.jpg?key=AWqIQFqVkaaaaaaaaaaaaaaaaaaaaaaxlapdsHLoRCSITVw..&a=220086000029851117-3da1ab76-92808-5840&nsp_ver=3.0
把上述下载链接复制到迅雷,添加批量任务就可以下载图片到本地。
以上,-- End --