Translater-语言翻译类

  1. 云栖社区>
  2. 博客列表>
  3. 正文

Translater-语言翻译类

清山 2011-12-15 13:14:00 浏览401 评论0

摘要: using System; using System.Collections.Generic; using System.Text; using System.Net; using System.


using System;
using System.Collections.Generic;

using System.Text;
using System.Net;
using System.IO;
using Newtonsoft.Json;
using System.Web;
using System.Text.RegularExpressions;

namespace Framework
{
    /// <summary>
    /// 语言类型
    /// </summary>
    public class LanguageType
    {
        /// <summary>
        /// 中文
        /// </summary>
        public static string Chinese = "zh-cn";
        /// <summary>
        /// 英文
        /// </summary>
        public static string English = "en";
    }
    /// <summary>
    /// 翻译方式类型
    /// </summary>
    public class TranslationType
    {
        /// <summary>
        /// Google
        /// </summary>
        public static string Google = "GoogleTanslater";
        /// <summary>
        /// Bing
        /// </summary>
        public static string Bing = "MircsoftTanslater";
    }
    /// <summary>
    /// 语言翻译类
    /// </summary>
    public class Translater
    {
        /// <summary>
        /// 翻译方法 中文:"zh-cn", 英文:"en" type:MircsoftTanslater,GoogleTanslater
        /// </summary>
        /// <param name="sourceText">翻译原文</param>
        /// <param name="fromLanguage">原始语言</param>
        /// <param name="toLanguage">目标语言</param>
        /// <param name="type">翻译API</param>
        /// <returns>译文</returns>
        public static string Translate(string sourceText, string fromLanguage, string toLanguage, string type = "MircsoftTanslater")
        {
            string translateStr = string.Empty;
            switch (type)
            {
                case "MircsoftTanslater":
                    translateStr = MircsoftTanslater(sourceText, fromLanguage, toLanguage);//"zh-cn", "en";
                    break;
                case "GoogleTanslater":
                    translateStr = GoogleTranslater_PostMethod(sourceText, fromLanguage, toLanguage);//"zh-cn", "en";
                    break;
            }
            return translateStr;
        }
        #region Google 翻译: Get方式获取翻译
        /// <summary>
        /// Google 翻译: Get方式获取翻译
        /// </summary>
        /// <param name="sourceText"></param>
        /// <param name="fromType"></param>
        /// <param name="toType"></param>
        /// <returns></returns>
        private static string GoogleTranslater_GetMethod(string sourceText, string fromType, string toType)
        {
            string result;
            string langPair = fromType.ToLower() == "zh-cn" ? "zh|en" : "en|zh";
            string url = "http://ajax.googleapis.com/ajax/services/language/translate?v=1.0&langpair=" + HttpUtility.UrlEncode(langPair) + "&q=" + HttpUtility.UrlEncode(sourceText);
            HttpWebRequest request = (HttpWebRequest)WebRequest.Create(url);
            request.Method = "GET";
            request.Referer = "http://www.my-ajax-site.com";
            try
            {
                HttpWebResponse response = (HttpWebResponse)request.GetResponse();
                Stream responseStream = response.GetResponseStream();
                StreamReader reader = new StreamReader(responseStream, Encoding.GetEncoding("UTF-8"));
                string responseStr = reader.ReadToEnd();
                ResponseResult readConfig = (ResponseResult)JavaScriptConvert.DeserializeObject(responseStr, typeof(ResponseResult));
                if (readConfig.responseStatus == "200")
                {
                    result = readConfig.responseData.translatedText;
                }
                else
                {
                    result = readConfig.responseStatus;
                }
            }
            catch (Exception Ex)
            {
                result = "err:" + Ex.Message;
            }
            return result;
        }
        #endregion
        #region Google 翻译: Post方式获取翻译
        /// <summary>
        /// Google 翻译: Post方式获取翻译
        /// </summary>
        /// <param name="sourceText"></param>
        /// <param name="fromType"></param>
        /// <param name="toType"></param>
        /// <returns></returns>
        private static string GoogleTranslater_PostMethod(string sourceText, string fromType, string toType)
        {
            string fromLan = fromType.ToLower() == "zh-cn" ? "zh" : "en";
            string toLan = toType.ToLower() == "zh-cn" ? "zh" : "en";
            HttpWebRequest requestScore = (HttpWebRequest)WebRequest.Create("http://translate.google.com/translate_t#");
            StringBuilder postContent = new StringBuilder();
            Encoding myEncoding = Encoding.UTF8;
            postContent.Append(HttpUtility.UrlEncode("hl", myEncoding));
            postContent.Append("=");
            postContent.Append(HttpUtility.UrlEncode("en", myEncoding));
            postContent.Append("&");
            postContent.Append(HttpUtility.UrlEncode("ie", myEncoding));
            postContent.Append("=");
            postContent.Append(HttpUtility.UrlEncode("UTF-8", myEncoding));
            postContent.Append("&");
            postContent.Append(HttpUtility.UrlEncode("sl", myEncoding));
            postContent.Append("=");
            postContent.Append(HttpUtility.UrlEncode(fromLan, myEncoding));
            postContent.Append("&");
            postContent.Append(HttpUtility.UrlEncode("text", myEncoding));
            postContent.Append("=");
            postContent.Append(HttpUtility.UrlEncode(sourceText, myEncoding));
            postContent.Append("&");
            postContent.Append(HttpUtility.UrlEncode("tl", myEncoding));
            postContent.Append("=");
            postContent.Append(HttpUtility.UrlEncode(toLan, myEncoding));

            byte[] data = Encoding.ASCII.GetBytes(postContent.ToString());
            requestScore.UserAgent = "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.2; SV1; .NET CLR 1.1.4322; .NET CLR 2.0.50727)";
            requestScore.Method = "Post";
            //requestScore.ContentType = "application/x-www-form-urlencoded;charset=gb2312";
            requestScore.ContentLength = data.Length;
            requestScore.KeepAlive = true;
            requestScore.Timeout = (6 * 60 * 1000);
            requestScore.ProtocolVersion = HttpVersion.Version10;

            Stream stream = requestScore.GetRequestStream();
            stream.Write(data, 0, data.Length);
            stream.Close();
            string content = string.Empty;
            try
            {
                System.Net.ServicePointManager.Expect100Continue = false;
                HttpWebResponse responseSorce = (HttpWebResponse)requestScore.GetResponse();
                StreamReader reader = new StreamReader(responseSorce.GetResponseStream());
                content = reader.ReadToEnd();
                responseSorce.Close();
                reader.Dispose();
                stream.Dispose();
            }
            catch (WebException ex)
            {
                HttpWebResponse responseSorce = (HttpWebResponse)ex.Response;//得到请求网站的详细错误提示
                StreamReader reader = new StreamReader(responseSorce.GetResponseStream());
                content = reader.ReadToEnd();
                responseSorce.Close();
                reader.Dispose();
                stream.Dispose();
            }
            finally
            {
                requestScore.Abort();
            }
            string reg = @"<(?<HtmlTag>[\w]+)[^>]*\s[iI][dD]=(?<Quote>[""']?)result_box(?(Quote)\k<Quote>)[""']?[^>]*>((?<Nested><\k<HtmlTag>[^>]*>)|</\k<HtmlTag>>(?<-Nested>)|.*?)*</\k<HtmlTag>>";
            //string reg = @"<(span) id=result_box [^>]*>.*?</\1>";//匹配出翻译内容
            Regex r = new Regex(reg);
            MatchCollection mcItem = r.Matches(content);
            string result = ConvertHtmlToText(mcItem[0].Value);
            return result;
        }
        /// <summary>
        /// 将HTML转换为纯文本
        /// </summary>
        /// <param name="source"></param>
        /// <returns></returns>
        public static string ConvertHtmlToText(string source)
        {
            // 代码的实现的思路是:
            //a、先将html文本中的所有空格、换行符去掉(因为html中的空格和换行是被忽略的)
            //b、将<head>标记中的所有内容去掉
            //c、将<script>标记中的所有内容去掉
            //d、将<style>标记中的所有内容去掉
            //e、将td换成空格,tr,li,br,p 等标记换成换行符
            //f、去掉所有以“<>”符号为头尾的标记去掉。
            //g、转换&,&nbps;等转义字符换成相应的符号
            //h、去掉多余的空格和空行
            string result;
            //remove line breaks,tabs
            result = source.Replace("\r", " ");
            result = result.Replace("\n", " ");
            result = result.Replace("\t", " ");
            //remove the header
            result = Regex.Replace(result, "(<head>).*(</head>)", string.Empty, RegexOptions.IgnoreCase);
            result = Regex.Replace(result, @"<( )*script([^>])*>", "<script>", RegexOptions.IgnoreCase);
            result = Regex.Replace(result, @"(<script>).*(</script>)", string.Empty, RegexOptions.IgnoreCase);
            //remove all styles
            result = Regex.Replace(result, @"<( )*style([^>])*>", "<style>", RegexOptions.IgnoreCase); //clearing attributes
            result = Regex.Replace(result, "(<style>).*(</style>)", string.Empty, RegexOptions.IgnoreCase);
            //insert tabs in spaces of <td> tags
            result = Regex.Replace(result, @"<( )*td([^>])*>", " ", RegexOptions.IgnoreCase);
            //insert line breaks in places of <br> and <li> tags
            result = Regex.Replace(result, @"<( )*br( )*>", "\r", RegexOptions.IgnoreCase);
            result = Regex.Replace(result, @"<( )*li( )*>", "\r", RegexOptions.IgnoreCase);
            //insert line paragraphs in places of <tr> and <p> tags
            result = Regex.Replace(result, @"<( )*tr([^>])*>", "\r\r", RegexOptions.IgnoreCase);
            result = Regex.Replace(result, @"<( )*p([^>])*>", "\r\r", RegexOptions.IgnoreCase);
            //remove anything thats enclosed inside < >
            result = Regex.Replace(result, @"<[^>]*>", string.Empty, RegexOptions.IgnoreCase);
            //replace special characters:
            result = Regex.Replace(result, @"&", "&", RegexOptions.IgnoreCase);
            result = Regex.Replace(result, @" ", " ", RegexOptions.IgnoreCase);
            result = Regex.Replace(result, @"<", "<", RegexOptions.IgnoreCase);
            result = Regex.Replace(result, @">", ">", RegexOptions.IgnoreCase);
            result = Regex.Replace(result, @"&(.{2,6});", string.Empty, RegexOptions.IgnoreCase);
            //remove extra line breaks and tabs
            result = Regex.Replace(result, @" ( )+", " ");
            result = Regex.Replace(result, "(\r)( )+(\r)", "\r\r");
            result = Regex.Replace(result, @"(\r\r)+", "\r\n");
            return result;
        }
        #endregion
        #region 微软翻译
        /// <summary>
        /// 微软翻译API :  语言类型:"zh-cn", "en"
        /// </summary>
        /// <param name="orgStr">翻译原文</param>
        /// <param name="fromType">原文语言类型</param>
        /// <param name="toType">目标语言类型</param>
        /// <returns></returns>
        public static string MircsoftTanslater(string orgStr, string fromType, string toType)
        {
            string content = string.Empty;
            string appId = "56E164FED4017D272E06AD7E16778536251CA5CB";
            string text = orgStr;// "Translate this for me";
            string from = fromType;// "en";
            string to = toType;// "zh-cn";

            string uri = "http://api.microsofttranslator.com/v2/Http.svc/Translate?appId=" + appId + "&text=" + System.Web.HttpUtility.UrlEncode(text) + "&from=" + from + "&to=" + to;
            HttpWebRequest httpWebRequest = (HttpWebRequest)WebRequest.Create(uri);
            WebResponse response = null;
            try
            {
                response = httpWebRequest.GetResponse();
                StreamReader reader = new StreamReader(response.GetResponseStream());
                content = reader.ReadToEnd();//"<string xmlns=\"http://schemas.microsoft.com/2003/10/Serialization/\">Hello, China</string>" 
                content = content.Replace("<string xmlns=\"http://schemas.microsoft.com/2003/10/Serialization/\">", "");
                content = content.Replace("</string>", "");
                response.Close();
                reader.Dispose();
            }
            catch (WebException e)
            {
                content = ProcessWebException(e, "Failed to translate");
            }
            finally
            {
                if (response != null)
                {
                    response.Close();
                    response = null;
                }
            }
            return content;
        }
        private static string ProcessWebException(WebException e, string message)
        {
            string result = string.Empty;
            result = string.Format("{0}: {1}", message, e.ToString());
            // Obtain detailed error information
            string strResponse = string.Empty;
            using (HttpWebResponse response = (HttpWebResponse)e.Response)
            {
                using (Stream responseStream = response.GetResponseStream())
                {
                    using (StreamReader sr = new StreamReader(responseStream, System.Text.Encoding.ASCII))
                    {
                        strResponse = sr.ReadToEnd();
                    }
                }
            }
            result = string.Format("Http status code={0}, error message={1}", e.Status, strResponse);
            return result;
        }
        #endregion
    }
    /// <summary>
    /// 翻译返回类
    /// </summary>
    public class ResponseResult
    {
        public ResponseData responseData { get; set; }
        public string responseDetails { get; set; }
        public string responseStatus { get; set; }
    }
    /// <summary>
    ///  
    /// </summary>
    public class ResponseData
    {
        public string translatedText { get; set; }
    }
}


用云栖社区APP,舒服~

【云栖快讯】诚邀你用自己的技术能力来用心回答每一个问题,通过回答传承技术知识、经验、心得,问答专家期待你加入!  详情请点击

网友评论

清山
文章502篇 | 关注3
关注
预置千套精美模板,覆盖上百行业,模板随意切换,支持6种语言,支持企业备案 查看详情
是基于语音识别、语音合成、自然语言理解等技术,为企业在多种实际应用场景下,赋予产品“能听、会... 查看详情
为您提供简单高效、处理能力可弹性伸缩的计算服务,帮助您快速构建更稳定、安全的应用,提升运维效... 查看详情
阿里云总监课正式启航

阿里云总监课正式启航