Google Doc API研究之一：模拟页面上传任意类型文件
一直以来想要做个程序，将google doc用作网盘，程序做 的差不多了才发现不是所有的人都可以上传任意类型的文件，只有商业用户才可以。商业用户是要交钱的的，这与我们倡导的免费精神相关太远。怎么办，我的心血 不能白费，google还算厚道没有把门关死，可以通过form的形式上传，我们可以模拟form的动作，就能上传了。好了费话少话，说实在的。
Google在上传时要进行身份验证。取得身份验证后，提出上传要求，这时返回一个上传地址，然后上传文件。下面一步步来。
第一步，通过网页访问DOC，其身份认证是通过cookie进行的，取得cookie的过程是
1、先访问http://docs.google.com，它通过自动跳转将你带到一个登录页面。按登录页面中的头部的set- cookie设置httpwebrequest的cookie，cookie中最重要的是HSID，然后设置url为 https://www.google.com/accounts/ServiceLoginAuth?service=writely 。然后设置提交数据ltmpl=homepage&continue=http://docs.google.com/& followup=http://docs.google.com/&service=writely&nui=1& rm=false&dsh={0}<mpl=homepage<mpl=homepage&GALX={1}& amp;Email={2}&Passwd={3}&rmShown=1&signIn=登录&asts=
其中Email填google帐户名，passwd填密码，GALX可以从cookie中找到，dsh可从页面的数据中得到。
2、再访问网址：https://www.google.com/accounts /CheckCookie?continue=http%3A%2F%2Fdocs.google.com%2F&followup=http %3A%2F%2Fdocs.google.com%2F&service=writely<mpl=homepage& chtml=LoginDoneHtml,取得下一步的访问地址，
3、再访问网 址:http://docs.google.com/?auth=.......,根据回应设置cookie，这时的cookie就是可以访问相当于 doc api中的token了。cookie中包含三条记录HSI，SID，writelySID。
下面是原程序：
private CookieContainer GetAuthenticationCookie(string User,string Passwd)
{
string GALX;
string dsh;
string resText = "";
CookieContainer Auth = new CookieContainer();
//第一步
HttpWebRequest req = CreatRequest("http://docs.google.com/");
req.Method = "POST";
req.ContentLength = 0;
HttpWebResponse res = (HttpWebResponse)req.GetResponse();
//第二步
resText = getResponseText(res);
GALX = resText.Substring(resText.IndexOf("GALX") + 26, 11);
if ((resText.Substring(resText.IndexOf("dsh") + 32, 1)) == "-")
dsh = resText.Substring(resText.IndexOf("dsh") + 32, 20);
else
dsh = resText.Substring(resText.IndexOf("dsh") + 32, 19);
string postData = string.Format(
"ltmpl=homepage&continue=http://docs.google.com/&followup=http: //docs.google.com/&service=writely&nui=1&rm=false&dsh= {0}<mpl=homepage<mpl=homepage&GALX={1}&Email={2}& amp; amp;Passwd={3}&rmShown=1&signIn=登录&asts="
, dsh, GALX, User, Passwd);
req = CreatRequest("https://www.google.com/accounts/ServiceLoginAuth?service=writely ");
req.AllowAutoRedirect = false;
req.Method = "POST";
//设置cookie及提交数据
Auth.Add(setCookie(res, "www.google.com"));
req.CookieContainer = Auth;
setPostData(req, postData);
res = (HttpWebResponse)req.GetResponse();
//第三步
req = CreatRequest(res.Headers["Location"]);
Auth.Add(setCookie(res, "www.google.com"));
req.CookieContainer = Auth;
res = (HttpWebResponse)req.GetResponse();
//第四步
resText = getResponseText(res);
string url = resText.Substring(resText.IndexOf("url=")).Split('\"')[0];
url = HttpUtility.HtmlDecode(url);
url = url.Substring(5, url.Length - 6);
req = CreatRequest(url);
req.Method = "GET";
req.AllowAutoRedirect = false;
Auth.Add(setCookie(res, "www.google.com"));
req.CookieContainer = Auth;
res = (HttpWebResponse)req.GetResponse();
Auth.Add(setCookie(res, "www.google.com"));
return Auth;
}
{
if (res.StatusCode != HttpStatusCode.OK)
return null;
Stream dataStream = res.GetResponseStream();
StreamReader reader = new StreamReader(dataStream);
string str = reader.ReadToEnd();
reader.Close();
return str;
}
/// <summary>
///设置传送的数据
/// </summary>
/// <param name="req">The req.</param>
/// <param name="postData">The post data.</param>
private void setPostData(HttpWebRequest req, string postData)
{
byte[] bytes = Encoding.UTF8.GetBytes(postData);
req.ContentLength = bytes.Length;
using (Stream dataStream = req.GetRequestStream())
{
dataStream.Write(bytes, 0, bytes.Length);
dataStream.Close();
}
}
/// <summary>
/// 根据response中头部的set-cookie对request中的cookie进行设置
/// </summary>
/// <param name="setCookie">The set cookie.</param>
/// <param name="defaultDomain">The default domain.</param>
/// <returns></returns>
private CookieCollection setCookie(HttpWebResponse res, string defaultDomain)
{
string[] setCookie = res.Headers.GetValues("Set-Cookie");
// there is bug in it,the datetime in "set-cookie" will be sepreated in two pieces.
List<string> a = new List<string>(setCookie);
for (int i = setCookie.Length - 1; i > 0; i--)
{
if (a[i].Substring(a[i].Length - 3) == "GMT")
{
a[i - 1] = a[i - 1] + ", " + a[i];
a.RemoveAt(i);
i--;
}
}
setCookie = a.ToArray<string>();
CookieCollection cookies = new CookieCollection();
foreach (string str in setCookie)
{
NameValueCollection hs = new NameValueCollection();
foreach (string i in str.Split(';'))
{
int index = i.IndexOf("=");
if (index > 0)
hs.Add(i.Substring(0, index), i.Substring(index + 1));
else
switch (i)
{
case "HttpOnly":
hs.Add("HttpOnly", "True");
break;
case "Secure":
hs.Add("Secure", "True");
break;
}
}
Cookie ck = new Cookie();
foreach (string Key in hs.AllKeys)
{
switch (Key)
{
case "Path":
ck.Path = hs[Key];
break;
case "Expires":
ck.Expires = DateTime.Parse(hs[Key]);
break;
case "Domain":
ck.Domain = hs[Key];
break;
case "HttpOnly":
ck.HttpOnly = true;
break;
case "Secure":
ck.Secure = true;
break;
default:
ck.Name = Key;
ck.Value = hs[Key];
break;
}
}
if (ck.Domain == "") ck.Domain = defaultDomain;
if (ck.Name != "") cookies.Add(ck);
}
return cookies;
}
/// <summary>
/// 对request进行基本的设置
/// </summary>
/// <param name="URL">The URL.</param>
/// <returns></returns>
private HttpWebRequest CreatRequest(string URL)
{
HttpWebRequest req = (HttpWebRequest)WebRequest.Create(URL);
req.UserAgent = "Mozilla/5.0 (Windows; U; Windows NT 5.1; zh-CN; rv:1.9.2.2) Gecko/20100316 Firefox/3.6.2";
req.ServicePoint.Expect100Continue = false;
req.ContentType = "application/x-www-form-urlencoded";
return req;
}
整个过程很简单，但调试花了我很长时间，主要是在https下调试走了很多弯路，一 开始使用httpanalyzer,结果有bug显示的传送数据总是重复。后来还是使用fiddler才解决。但使用fiddler时在https下会出 现证书与网站 不符的错误。
解决办法是：
ServicePointManager.ServerCertificateValidationCallback = new RemoteCertificateValidationCallback(ValidateServerCertificate);
public static bool ValidateServerCertificate(object sender, X509Certificate certificate, X509Chain chain, SslPolicyErrors sslPolicyErrors)
{
return true;
}
第 一个函数在设置httpwebquest时调用就不会出现错误了。这时就可以用fiddler捕获https包了。今天就写到这里。
这里有几点猜 想未证实，授权cookie中的writelySID及SID可能与google 的client的LSID和SID相同，但HSID不知从何而来。
cookie 可以反复使用，不用每次对话都要取得cookie。
代码写得很笨拙，大家不吝赐教。
下一步，研究上传任意文件至指定目录。
