/// <summary> /// html转换为纯文本 /// </summary> /// <param name="source"></param> /// <returns></returns> private static string HtmlToPlainText(string source) { string result; //remove line breaks,tabs result = source.Replace(" ", " "); result = result.Replace(" ", " "); result = result.Replace(" ", " "); //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( )*>", " ", RegexOptions.IgnoreCase); result = Regex.Replace(result, @"<( )*li( )*>", " ", RegexOptions.IgnoreCase); //insert line paragraphs in places of <tr> and <p> tags result = Regex.Replace(result, @"<( )*tr([^>])*>", " ", RegexOptions.IgnoreCase); result = Regex.Replace(result, @"<( )*p([^>])*>", " ", 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, @"'", "'", 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, "( )( )+( )", " "); result = Regex.Replace(result, @"( )+", " "); return result; }