java - Selenium: descargue el archivo en Internet Explorer a la carpeta especificada sin enlace directo, sin Windows Forms, sin AutoIt o Robot

c# internet-explorer

Question

A menudo me he enfrentado a un problema, cómo descargar archivos en IE. A diferencia de Chrome de Firefox, no solo puede especificar la carpeta requerida, y todos los archivos se descargarán a esa carpeta. También necesita interactuar con formularios nativos de Windows, etc.

Hay múltiples opciones, como usar AutoIt, usar comandos de teclado, Robot, etc. Pero todas estas opciones no son estables, requieren una espera explícita, el uso de bibliotecas redundantes y no son apropiadas cuando se ejecutan pruebas en paralelo. El otro problema es qué hacer si el archivo no se descarga mediante un enlace directo, pero el enlace se genera desde el comando javascript o se recibe del servidor, y no se puede extraer de html.

Todos estos problemas pueden resolverse, aquí, en la respuesta, le mostraré cómo hacerlo. La solución está escrita en C #, creo que lo mismo se puede implementar en Java

Answer 1

Método DownloadFileIexplore descargará el archivo a la ruta de archivo especificada (carpeta + nombre de archivo), por ejemplo, DownloadFileExplore ("C: / my_folder / report.xslx")

    public void DownloadFileIexplore(string filePath)
    {
        //Click the link.
        //Simple click can be used instead, but in my case it didn''t work for all the links, so i''ve replaced it with click via action: 
        new Actions(Browser.Driver).MoveToElement(Element).Click(Element).Perform();

        //Different types of element can be used to download file.
        //If the element contains direct link, it can be extracter from ''href'' attribute.
        //If the element doesn''t contains href or it''s just an javascript command, link will be extracted from the browser http requests.
        //
        string downloadUrlOrLink = Element.GetAttribute("href") != null && !Element.GetAttribute("href").Contains("javascript")
            ? Element.GetAttribute("href")
            : GetRequestUrls().Last() ?? GetLinkInNewWindowIexplore();
        if (downloadUrlOrLink == null)
        {
            throw Log.Exception("Download url cannot be read from the element href attribute and from the recent http requests.");
        }

        /// the last step is to download file using CookieWebClient
        /// method DownloadFile is available in the System.Net.WebClient, 
        /// but we have to create new class CookieWebClient, that will be inherited from System.Net.WebClient with one overriden method
        new CookieWebClient(GetCookies()).DownloadFile(downloadUrlOrLink, filePath);
    }


    /// <summary>
    /// this method returns all the http requests sent from browser.
    /// the latest requests was send when link (or button) was clicked to download file
    /// so we will need just to get last element from list: GetRequestUrls().Last().
    /// or, if the request for file downloading isn''t the last, find the required request by part of url, in my case it was ''common/handler'', e.g.:
    /// GetRequestUrls().LastOrDefault(x => x.Contains("common/handler")) 
    /// <summary>
    public List<string> GetRequestUrls()
    {
        ReadOnlyCollection<object> requestsUrls = (ReadOnlyCollection<object>) 
            Driver.ExecuteScript("return window.performance.getEntries().map(function(x) { return x.name });");
        return requestsUrls.Select(x => (string) x).ToList();
    }

    /// <summary>
    /// In some cases after clicking the Download button new window is opened in IE.
    /// Driver.WindowHandles can return only one window instead of two.
    /// To solve this problem reset IE security settings and set Enable Protected Mode for each zone.
    /// </summary>
    private string GetLinkInNewWindowIexplore()
    {
        /// here it would be better to add waiting till new window is opened.
        /// in that case we have to calculate number of windows before click and send this number as argument to GetLinkInNewWindowIexplore()
        /// and wait till number will be increased by 1 
        var availableWindows = Driver.WindowHandles;
        if (availableWindows.Count > 1)
        {
            Driver.SwitchTo().Window(availableWindows[availableWindows.Count - 1]);
        }
        string url;
        try
        {
            url = Driver.Url;
        }
        catch (Exception)
        {
            url = Driver.ExecuteScript("return document.URL;").ToString();
        }
        Driver.SwitchTo().Window(Driver.WindowHandles[0]);
        return url;
    }

    public System.Net.CookieContainer GetCookies()
    {
        CookieContainer cookieContainer = new CookieContainer();
        foreach (OpenQA.Selenium.Cookie cookie in Driver.Manage().Cookies.AllCookies)
        {
            cookieContainer.Add(new System.Net.Cookie
            {
                Name = cookie.Name,
                Value = cookie.Value,
                Domain = "domain of your site, you can find, track http requests send from your site in browser dev tools, tab Network"
            });
        }
        return cookieContainer;
    }


public class CookieWebClient : WebClient
{
    private readonly CookieContainer _cookieContainer;

    public CookieWebClient(CookieContainer cookieContainer)
    {
        _cookieContainer = cookieContainer;
    }

    /// it''s necessary to override method to add cookies, because file cannot be download by non-authorized user
    /// ServerCertificateValidationCallback is set to true to avoid some possible certificate errors
    protected override WebRequest GetWebRequest(Uri address)
    {
        ServicePointManager.ServerCertificateValidationCallback = delegate { return true; };
        WebRequest request = base.GetWebRequest(address);
        HttpWebRequest webRequest = request as HttpWebRequest;
        if (webRequest != null)
        {
            webRequest.CookieContainer = _cookieContainer;
        }
        return request;
    }
}