using System;
using System.Diagnostics;
using System.IO;
using System.Linq;
using System.Net;
using System.Reflection;
using System.Threading;
using System.Threading.Tasks;
using NewLife.Collections;
using NewLife.Log;
#if !NET4
using System.Net.Http;
using TaskEx = System.Threading.Tasks.Task;
#endif
namespace NewLife.Web
{
/// <summary>扩展的Web客户端</summary>
public class WebClientX : DisposeBase
{
#region 属性
/// <summary>超时，默认15000毫秒</summary>
public Int32 Timeout { get; set; } = 15000;
/// <summary>最后使用的连接名</summary>
public Link LastLink { get; set; }
#endregion
#region 构造
static WebClientX()
{
#if NET4
try
{
ServicePointManager.SecurityProtocol |= SecurityProtocolType.Ssl3 | SecurityProtocolType.Tls;
}
catch
{
ServicePointManager.SecurityProtocol |= SecurityProtocolType.Tls;
}
//#elif STD21
// ServicePointManager.SecurityProtocol |= SecurityProtocolType.Tls | SecurityProtocolType.Tls11 | SecurityProtocolType.Tls12;
#else
try
{
ServicePointManager.SecurityProtocol |= SecurityProtocolType.Ssl3 | SecurityProtocolType.Tls | SecurityProtocolType.Tls11 | SecurityProtocolType.Tls12;
}
catch
{
ServicePointManager.SecurityProtocol |= SecurityProtocolType.Tls | SecurityProtocolType.Tls11 | SecurityProtocolType.Tls12;
}
#endif
}
/// <summary>实例化</summary>
public WebClientX() { }
/// <summary>销毁</summary>
/// <param name="disposing"></param>
protected override void Dispose(Boolean disposing)
{
base.Dispose(disposing);
#if !NET4
_client.TryDispose();
#endif
}
#endregion
#region 核心方法
#if NET4
/// <summary>创建客户端会话</summary>
/// <param name="uri"></param>
/// <returns></returns>
protected virtual HttpWebRequest Create(String uri)
{
var req = WebRequest.Create(uri) as HttpWebRequest;
return req;
}
/// <summary>下载数据</summary>
/// <param name="address"></param>
/// <param name="data"></param>
/// <returns></returns>
protected virtual async Task<Byte[]> SendAsync(String address, Byte[] data = null)
{
var time = Timeout;
if (time <= 0) time = 3000;
while (true)
{
var http = Create(address);
http.Timeout = time;
http.Method = data == null || data.Length == 0 ? "GET" : "POST";
Log.Info("WebClientX.SendAsync {0}", address);
// 发送请求
var rs = (await Task.Factory.FromAsync(http.BeginGetResponse, http.EndGetResponse, null)) as HttpWebResponse;
// 如果是重定向
switch (rs.StatusCode)
{
case HttpStatusCode.MovedPermanently:
case HttpStatusCode.Redirect:
case HttpStatusCode.RedirectMethod:
var url = rs.Headers[HttpResponseHeader.Location + ""] + "";
if (!url.IsNullOrEmpty())
{
address = url;
data = null;
continue;
}
break;
default:
break;
}
var ms = NewLife.Collections.Pool.MemoryStream.Get();
var ns = rs.GetResponseStream();
ns.CopyTo(ms);
while (rs.ContentLength > 0 && ms.Length < rs.ContentLength)
{
Thread.Sleep(10);
ns.CopyTo(ms);
}
return ms.Put(true);
}
}
/// <summary>下载数据</summary>
/// <param name="address"></param>
/// <returns></returns>
public virtual async Task<Byte[]> DownloadDataAsync(String address) => await SendAsync(address);
/// <summary>下载字符串</summary>
/// <param name="address"></param>
/// <returns></returns>
public virtual async Task<String> DownloadStringAsync(String address) => (await SendAsync(address)).ToStr();
/// <summary>下载文件</summary>
/// <param name="address"></param>
/// <param name="fileName"></param>
public virtual async Task DownloadFileAsync(String address, String fileName)
{
var rs = await SendAsync(address);
fileName.EnsureDirectory(true);
using (var fs = new FileStream(fileName, FileMode.OpenOrCreate, FileAccess.ReadWrite))
{
await fs.WriteAsync(rs, 0, rs.Length);
}
}
#else
private HttpClient _client;
/// <summary>创建客户端会话</summary>
/// <returns></returns>
public virtual HttpClient EnsureCreate()
{
var http = _client;
if (http == null)
{
var handler = new HttpClientHandler();
http = new HttpClient(handler);
_client = http;
http.Timeout = new TimeSpan(0, 0, 0, 0, Timeout);
}
return http;
}
/// <summary>发送请求，获取响应</summary>
/// <param name="address"></param>
/// <param name="content"></param>
/// <returns></returns>
public virtual async Task<HttpContent> SendAsync(String address, HttpContent content = null)
{
var time = Timeout;
if (time <= 0) time = 3000;
var http = EnsureCreate();
Log.Info("{2}.{1} {0}", address, content != null ? "Post" : "Get", GetType().Name);
// 发送请求
var source = new CancellationTokenSource(time);
var task = content != null ? http.PostAsync(address, content, source.Token) : http.GetAsync(address, source.Token);
var rs = await task;
return rs.Content;
}
/// <summary>下载字符串</summary>
/// <param name="address"></param>
/// <returns></returns>
public virtual async Task<String> DownloadStringAsync(String address)
{
var rs = await SendAsync(address);
return await rs.ReadAsStringAsync();
}
/// <summary>下载文件</summary>
/// <param name="address"></param>
/// <param name="fileName"></param>
public virtual async Task DownloadFileAsync(String address, String fileName)
{
var rs = await SendAsync(address);
fileName.EnsureDirectory(true);
using (var fs = new FileStream(fileName, FileMode.OpenOrCreate, FileAccess.ReadWrite))
{
await rs.CopyToAsync(fs);
}
}
#endif
#endregion
#region 方法
/// <summary>获取指定地址的Html，自动处理文本编码</summary>
/// <param name="url"></param>
/// <returns></returns>
public String GetHtml(String url) => TaskEx.Run(() => DownloadStringAsync(url)).Result;
/// <summary>获取指定地址的Html，分析所有超链接</summary>
/// <param name="url"></param>
/// <returns></returns>
public Link[] GetLinks(String url)
{
var html = GetHtml(url);
if (html.IsNullOrWhiteSpace()) return new Link[0];
return Link.Parse(html, url);
}
/// <summary>分析指定页面指定名称的链接，并下载到目标目录，返回目标文件</summary>
/// <remarks>
/// 根据版本或时间降序排序选择
/// </remarks>
/// <param name="urls">指定页面</param>
/// <param name="name">页面上指定名称的链接</param>
/// <param name="destdir">要下载到的目标目录</param>
/// <returns>返回已下载的文件，无效时返回空</returns>
public String DownloadLink(String urls, String name, String destdir)
{
Log.Info("下载链接 {0}，目标 {1}", urls, name);
var names = name.Split(",", ";");
var file = "";
Link link = null;
Exception lastError = null;
foreach (var url in urls.Split(",", ";"))
{
try
{
var ls = GetLinks(url);
if (ls.Length == 0) return file;
// 过滤名称后降序排序，多名称时，先确保前面的存在，即使后面名称也存在并且也时间更新都不能用
foreach (var item in names)
{
link = ls.Where(e => !e.Url.IsNullOrWhiteSpace())
.Where(e => e.Name.EqualIgnoreCase(item) || e.FullName.Equals(item))
.OrderByDescending(e => e.Version)
.ThenByDescending(e => e.Time)
.FirstOrDefault();
if (link != null) break;
}
}
catch (WebException ex)
{
Log.Error(ex.Message);
}
catch (Exception ex)
{
lastError = ex;
}
if (link != null) break;
}
if (link == null)
{
if (lastError != null) throw lastError;
return file;
}
LastLink = link;
var linkName = link.FullName;
var file2 = destdir.CombinePath(linkName).EnsureDirectory();
// 已经提前检查过，这里几乎不可能有文件存在
if (File.Exists(file2))
{
// 如果连接名所表示的文件存在，并且带有时间，那么就智能是它啦
var p = linkName.LastIndexOf("_");
if (p > 0 && (p + 8 + 1 == linkName.Length || p + 14 + 1 == linkName.Length))
{
Log.Info("分析得到文件 {0}，目标文件已存在，无需下载 {1}", linkName, link.Url);
return file2;
}
}
Log.Info("分析得到文件 {0}，准备下载 {1}，保存到 {2}", linkName, link.Url, file2);
// 开始下载文件，注意要提前建立目录，否则会报错
file2 = file2.EnsureDirectory();
var sw = Stopwatch.StartNew();
TaskEx.Run(() => DownloadFileAsync(link.Url, file2)).Wait();
sw.Stop();
if (File.Exists(file2))
{
Log.Info("下载完成，共{0:n0}字节，耗时{1:n0}毫秒", file2.AsFile().Length, sw.ElapsedMilliseconds);
file = file2;
}
return file;
}
/// <summary>分析指定页面指定名称的链接，并下载到目标目录，解压Zip后返回目标文件</summary>
/// <param name="urls">提供下载地址的多个目标页面</param>
/// <param name="name">页面上指定名称的链接</param>
/// <param name="destdir">要下载到的目标目录</param>
/// <param name="overwrite">是否覆盖目标同名文件</param>
/// <returns></returns>
public String DownloadLinkAndExtract(String urls, String name, String destdir, Boolean overwrite = false)
{
var file = "";
// 下载
try
{
file = DownloadLink(urls, name, destdir);
}
catch (Exception ex)
{
Log.Error(ex?.GetTrue()?.ToString());
// 这个时候出现异常，删除zip
if (!file.IsNullOrEmpty() && File.Exists(file))
{
try
{
File.Delete(file);
}
catch { }
}
}
if (file.IsNullOrEmpty()) return null;
// 解压缩
try
{
Log.Info("解压缩到 {0}", destdir);
//ZipFile.ExtractToDirectory(file, destdir);
file.AsFile().Extract(destdir, overwrite);
// 删除zip
File.Delete(file);
return file;
}
catch (Exception ex)
{
Log.Error(ex?.GetTrue()?.ToString());
//// 这个时候出现异常，删除zip
//if (!file.IsNullOrEmpty() && File.Exists(file))
//{
// try
// {
// File.Delete(file);
// }
// catch { }
//}
}
return null;
}
#endregion
#region 辅助
private static Boolean? _useUnsafeHeaderParsing;
/// <summary>设置是否允许不安全头部</summary>
/// <remarks>
/// 微软WebClient默认要求严格的Http头部，否则报错
/// </remarks>
/// <param name="useUnsafe"></param>
/// <returns></returns>
public static Boolean SetAllowUnsafeHeaderParsing(Boolean useUnsafe)
{
if (_useUnsafeHeaderParsing != null && _useUnsafeHeaderParsing.Value == useUnsafe) return true;
#if __CORE__
_useUnsafeHeaderParsing = true;
#else
//Get the assembly that contains the internal class
var aNetAssembly = Assembly.GetAssembly(typeof(System.Net.Configuration.SettingsSection));
if (aNetAssembly == null) return false;
//Use the assembly in order to get the internal type for the internal class
var type = aNetAssembly.GetType("System.Net.Configuration.SettingsSectionInternal");
if (type == null) return false;
//Use the internal static property to get an instance of the internal settings class.
//If the static instance isn't created allready the property will create it for us.
var section = type.InvokeMember("Section", BindingFlags.Static | BindingFlags.GetProperty | BindingFlags.NonPublic, null, null, new Object[] { });
if (section != null)
{
//Locate the private bool field that tells the framework is unsafe header parsing should be allowed or not
var useUnsafeHeaderParsing = type.GetField("useUnsafeHeaderParsing", BindingFlags.NonPublic | BindingFlags.Instance);
if (useUnsafeHeaderParsing != null)
{
useUnsafeHeaderParsing.SetValue(section, useUnsafe);
_useUnsafeHeaderParsing = useUnsafe;
return true;
}
}
#endif
return false;
}
#endregion
#region 日志
/// <summary>日志</summary>
public ILog Log { get; set; } = Logger.Null;
#endregion
}
}
|