怎样截获IE下载的文件？

[SnHnBn]

能不能在下载之前截获准备要下载的文件？好像用BHO只能截获一个连接，但不知道是不是可以下载的文件。怎样才能知道？或者能否预先分析要下载的内容？

[zhucde]

截获链接后,分析其名称,比如后缀满足"*.zip,*.rar,*.exe,*.jpg....."等则视为可下载的,
大部分下载软件也是这样的.

[SnHnBn]

可惜了，现在很多链接是没有明显的后缀的，不信你们仔细观察看看。很多都是用GET METHOD来做的，比如http://www.site.com/link?fileid=245667775，从连接上根本看不出来是个可以下载的文件。可是浏览器怎么能够判断是个文件，连文件名都知道呢？

[SnHnBn]

都没有办法吗？版主看看我这贴转移到什么地方比较合适?

[珍惜生命远离CPP]

http://www.site.com/link?fileid=245667775
这样的链接都是重定向的链接，向这个链接GET一下，就会得到下载文件的具体地址。

[SnHnBn]

akirya 你说得很对，可是这样又有新问题：怎样向这个链接GET一下？单独开个Socket？或是怎样能从IE上获取得到Get之后的重定向？

[SnHnBn]

yjgx007 你指的是使用独立的WinInet来操作这个连接吧？会不会有些网站防盗链机制会起作用而导致无法HttpQueryInfo啊？

不过倒是感觉接近目标了。

[wlwlxj]

OnBeginDownload不晓得可以否

[yjgx007]

回复人：SnHnBn(大可达) ( 二级(初级)) 信誉：100 2006-10-7 12:16:27 得分:0
?

可惜了，现在很多链接是没有明显的后缀的，不信你们仔细观察看看。很多都是用GET METHOD来做的，比如http://www.site.com/link?fileid=245667775，从连接上根本看不出来是个可以下载的文件。可是浏览器怎么能够判断是个文件，连文件名都知道呢？

=========================================================
没关系,只要截获链接,然后打开这个链接,接着用HttpQueryInfo查询一下内容类型, 记着传参数HTTP_QUERY_CONTENT_TYPE, 通常htm,html格式返回text/html

[蒋晟]

It depends
If you start another HTTP web request in a webbrowser application, you may need to login again.

[lion_wing]

参考：
http://blog.csdn.net/lion_wing/articles/534716.aspx
http://blog.csdn.net/lion_wing/archive/2006/06/27/839134.aspx

[SnHnBn]

To jiangsheng(蒋晟.Net[MVP]) :
login again说得倒是挺容易的，不知道如何才能用程序自动做到？

[SnHnBn]

To lion_wing(凤之焚):
似乎你说的这个正是我需要的，我先去试试，然后回来结贴。

[蒋晟]

private void webBrowser1_StartNavigate(object sender, BrowserExtendedNavigatingEventArgs e)
{
if (e.Url.LocalPath.Contains("exportDoc.asp"))
{
e.Cancel = true;
string[] QueryParameters = e.Url.Query.Split("&".ToCharArray());
string FileName = QueryParameters[0];
//remove "?filename="
FileName = FileName.Substring(
FileName.IndexOf('=')+1);
FileName = Microsoft.JScript.GlobalObject.unescape(FileName);
string LocalFile =
Path.Combine(Environment.GetFolderPath(Environment.SpecialFolder.MyDocuments), "InfraHealth");
Directory.CreateDirectory(LocalFile);
LocalFile = Path.Combine(LocalFile, FileName);
DownloadFile(e.Url, LocalFile);
}
Debug.WriteLine(
string.Format("Navigating to :{0}",e.Url));
}

void DownloadFile(Uri DownloadURL, string LocalFile)
{
try
{
string UserAgent = "CSharp HTTP Sample";
string GetMethod = "GET";
string PostMethod= "POST";
Uri LoginURL= new Uri("https://www.example.com/checklogon.asp");
string PostData = "username=Example&password=example";

CookieContainer CookieJar = new CookieContainer();
ServicePointManager.ServerCertificateValidationCallback +=
new RemoteCertificateValidationCallback(myCertificateValidation);

HttpWebRequest LoginRequest=null;
CreateHttpWebRequest(ref LoginRequest, LoginURL,ref CookieJar, UserAgent, PostMethod, PostData);
HttpWebResponse LoginResponse = (HttpWebResponse)LoginRequest.GetResponse();
DebugWriteContent(ref LoginResponse);
CookieCollection cookies = LoginResponse.Cookies;
LoginResponse.Close();
HttpWebRequest DownloadRequest = null;
CreateHttpWebRequest(ref DownloadRequest, DownloadURL, ref CookieJar, UserAgent, GetMethod, string.Empty);
HttpWebResponse DownloadResponse = (HttpWebResponse)DownloadRequest.GetResponse();
Stream DownloadStream = DownloadResponse.GetResponseStream();

string TempFile=Path.GetTempFileName();
int buffSize = 4 * 1024; //4KB
byte[] buffer = new byte[buffSize];
using (FileStream fileStream = File.OpenWrite(TempFile))
{
while (true)
{
int count = DownloadStream.Read(buffer, 0, buffSize);
if (count == 0) break;
fileStream.Write(buffer, 0, count);
}
}
DownloadStream.Close();
DownloadResponse.Close();
File.Copy(TempFile, LocalFile, true);
}
void CreateHttpWebRequest(ref HttpWebRequest webRequest, Uri URL, ref CookieContainer cookie,
string UserAgent, string Method, string PostDataText)
{
webRequest
= (HttpWebRequest)WebRequest.Create(URL);
// set the HttpWebRequest objects cookie container
// if you have any cookies that you want to go with the request you can add them
// to the cookiecontainer. If you had made a previous request that returned any cookies
// that needed to be sent on subsequent request this will make sure that they are sent.
webRequest.CookieContainer = cookie;
// the default credentials are usually the Windows credentials (user name, password, and domain) of the user running the application
webRequest.Credentials = CredentialCache.DefaultCredentials;
webRequest.UserAgent = UserAgent;
webRequest.KeepAlive = true;
webRequest.Headers.Set("Pragma", "no-cache");
//set the request timeout to 5 min.
webRequest.Timeout = 300000;
webRequest.Method = Method;
// See what the Method is a POST
if ("POST" == Method)
{
// add the content type so we can handle form data
webRequest.ContentType = "application/x-www-form-urlencoded";
if (PostDataText.Length < 1) // we don't have any data to post
{
return;
}
// we need to store the data into a byte array
byte[] PostData = System.Text.Encoding.ASCII.GetBytes(PostDataText);
webRequest.ContentLength = PostData.Length;
Stream tempStream = webRequest.GetRequestStream();
// write the data to be posted to the Request Stream
tempStream.Write(PostData, 0, PostData.Length);
tempStream.Close();
}
return;
}

[SnHnBn]

To jiangsheng(蒋晟.Net[MVP]) :
Uri LoginURL= new Uri("https://www.example.com/checklogon.asp");
string PostData = "username=Example&password=example";
似乎必须针对特定的网页内容来作处理，不具备通用性。


To lion_wing(凤之焚):
有没有全能的MIME Type？不知怎么拦截不到下载文件，连MIME Filter的DLL都没有加载。

[SnHnBn]

To lion_wing(凤之焚):
好像对下载的文件不起作用，只能处理要显示的内容？

[SnHnBn]

To lion_wing(凤之焚):
谢谢，问题搞定了。不过每种要下载的文件要对应一个mime type，似乎通用性还不够，不知道能不能直接截获未注册mime type的文件？

[蒋晟]

http://windowssdk.msdn.microsoft.com/en-us/library/ms673378(VS.80).aspx

[lion_wing]

得到要的下载文件要用RegisterNamespace
得到显示的内容要用RegisterMimeFilter

更多的MimeType可以查看MSDN！

[SnHnBn]

To jiangsheng(蒋晟.Net[MVP]):
可惜目标平台不允许我用IDownloadManager

To lion_wing(凤之焚):
我想知道的是可不可以不指定MimeType？就是说不针对特定MimeType的通用Filter？

[SnHnBn]

如果可以有通用Filter的话，另外给分，谢谢。