Python 使用Selenium和Firefox版本40，如何下载文件？

通过Selenium下载文件的旧方法似乎不再有效

我的代码是：

    fp = webdriver.FirefoxProfile()
    fp.set_preference("browser.download.dir", os.getcwd())
    fp.set_preference("browser.download.folderList", 2)
    fp.set_preference("browser.download.manager.showWhenStarting", False)
    fp.set_preference("browser.helperApps.neverAsk.saveToDisk",
                      "application/pdf")

    self.driver = webdriver.Firefox(firefox_profile=fp)
    self.longMessage = True

但是，“文件”对话框仍会出现。我已经做了相当多的字段开关操作，但经过一番挖掘，我发现Selenium生成的默认Firefox配置文件的

prefs.js

文件与我手动选中“从现在开始自动为这种类型的文件执行此操作”的文件的

prefs.js

文件之间没有区别在下载对话框中

mimeTypes.rdf

文件确实发生了更改，但添加了以下行：

<RDF:Description RDF:about="urn:mimetype:handler:application/pdf"
               NC:alwaysAsk="false"
               NC:saveToDisk="true"
               NC:handleInternal="false">
<NC:externalApplication RDF:resource="urn:mimetype:externalApplication:application/pdf"/>

但是，我不知道在创建新的Firefox配置文件时如何设置自定义mimeTypes.rdf文件。有人知道吗

---

为了防止任何人建议我只卷曲下载URL，文件是为用户生成的，我需要特别验证.pdf文件是否已下载到驱动器。

您可以通过链接创建从internet下载文件的其他方法

我的c#代码示例：

如您所见，在这段代码中，我从image元素中获取src属性，并从浏览器外部下载它，以获得绝对正确的位图图像（之后，我可以将其保存到HDD）。

---

通过同样的方式，您可以从links=）下载任何文件。

我是R用户，所以只需使用R发布我的解决方案即可。如果您不能在python中转换相同的代码，请告诉我，我将提示您使用相同的代码

known_formats <- c("application/vnd.ms-excel","application/vnd.openxmlformats-officedocument.spreadsheetml.sheet")


firefox_profile.me <- makeFirefoxProfile(list(marionette = TRUE,
                                              # this is for certificate issues [may be ignored]
                                              webdriver_accept_untrusted_certs = TRUE,
                                              webdriver_assume_untrusted_issuer = TRUE,
                                              # download related settings
                                              browser.download.folderList = 2L,
                                              browser.download.manager.showWhenStarting = FALSE,
                                              # put your path here but remember to give path like C:\\DirOfYourChoice and not C:\\DirOfYourChoice\\ [last \\ is not going to work]
                                              browser.download.dir = normalizePath("TestDL"),
                                              browser.helperApps.alwaysAsk.force = FALSE,
                                              browser.helperApps.neverAsk.openFile = paste0(known_formats, collapse = ","),
                                              browser.helperApps.neverAsk.saveToDisk = paste0(known_formats, collapse = ","),
                                              browser.download.manager.showWhenStarting = FALSE,
                                              # this is for marionette and related security
                                              "browser.tabs.remote.force-enable" = TRUE,
                                              pdfjs.disabled = TRUE))

remDr <- remoteDriver(remoteServerAddr = "localhost",
                      port = 4444,
                      browserName = "firefox",
                      extraCapabilities = firefox_profile.me)

remDr$open()

remDr$navigate("https://www.google.com/search?q=sample+xlsx")

remDr$findElement(using = "css selector", value = ".g:nth-child(1) a")$clickElement()

remDr$navigate("https://www.google.com/search?q=test+xls")

remDr$findElement(using = "css selector", value = ".g:nth-child(1) a")$clickElement()

希望您能够将其移植到python。只需尝试在firefox

makeFirefoxProfile

进一步理解的参考资料：-

---

在我的原始帖子中，我试图避开这个答案。该文件是在用户单击按钮时生成的。我没有办法抢先获得下载链接：（@Staunch，你可以点击按钮获得链接并下载。有什么问题吗？你如何获得链接？这是一个

javascript

方法和一个

ajax

请求。我也在努力寻找解决方案。我发现我使用的是Firefox版本50.1.0。当出现提示时，RSelenium.未能下载。然而，在一些情况下它起作用了。我会写一封回复，对同样的内容进行重新评分。

known_formats <- c("application/vnd.ms-excel","application/vnd.openxmlformats-officedocument.spreadsheetml.sheet")


firefox_profile.me <- makeFirefoxProfile(list(marionette = TRUE,
                                              # this is for certificate issues [may be ignored]
                                              webdriver_accept_untrusted_certs = TRUE,
                                              webdriver_assume_untrusted_issuer = TRUE,
                                              # download related settings
                                              browser.download.folderList = 2L,
                                              browser.download.manager.showWhenStarting = FALSE,
                                              # put your path here but remember to give path like C:\\DirOfYourChoice and not C:\\DirOfYourChoice\\ [last \\ is not going to work]
                                              browser.download.dir = normalizePath("TestDL"),
                                              browser.helperApps.alwaysAsk.force = FALSE,
                                              browser.helperApps.neverAsk.openFile = paste0(known_formats, collapse = ","),
                                              browser.helperApps.neverAsk.saveToDisk = paste0(known_formats, collapse = ","),
                                              browser.download.manager.showWhenStarting = FALSE,
                                              # this is for marionette and related security
                                              "browser.tabs.remote.force-enable" = TRUE,
                                              pdfjs.disabled = TRUE))

remDr <- remoteDriver(remoteServerAddr = "localhost",
                      port = 4444,
                      browserName = "firefox",
                      extraCapabilities = firefox_profile.me)

remDr$open()

remDr$navigate("https://www.google.com/search?q=sample+xlsx")

remDr$findElement(using = "css selector", value = ".g:nth-child(1) a")$clickElement()

remDr$navigate("https://www.google.com/search?q=test+xls")

remDr$findElement(using = "css selector", value = ".g:nth-child(1) a")$clickElement()

Firefox 50.1.0 [while I'm writing this post]
Selenium [3.0.1]
R [3.3.2 (2016-10-31)]