代码优化

keroroqingwa · May 5, 2020 · e85b205 · e85b205
1 parent 5f562d2
commit e85b205
Show file tree

Hide file tree

Showing 2 changed files with 51 additions and 19 deletions.
diff --git a/src/CatBookApp.Application/BookSearches/Captures/BiqugeCapture.cs b/src/CatBookApp.Application/BookSearches/Captures/BiqugeCapture.cs
@@ -9,7 +9,7 @@
 namespace CatBookApp.BookSearches.Captures
 {
     /// <summary>
-    /// 抓取来源：笔趣阁
+    /// 抓取来源：笔趣阁 https://www.xsbiquge.com
     /// </summary>
     public class BiqugeCapture : IBookCapture
     {
@@ -23,34 +23,23 @@ public List<BookInfoDto> GetBooks(string q, int pn)
         {
             HtmlDocument doc = new HtmlDocument();
             string url = string.Empty;
-            //这里请求三次是因为。。。 调试过就知道，你就当做是错误重试吧 (′゜ω。‵)
+            //这里请求两次是因为。。。 调试过就知道，你就当做是错误重试吧 (′゜ω。‵)
             try
             {
                 try
                 {
                     //no.1
-                    url = $"https://www.xxbiquge.com/search.php?keyword={q}&page={pn}&p={pn - 1}";
+                    url = $"https://www.xsbiquge.com/search.php?keyword={q}&page={pn}&p={pn - 1}";
                     HtmlWeb webClient = new HtmlWeb();
                     doc = webClient.Load(url);
                 }
                 catch
                 {
-                    try
-                    {
-                        //no.2
-                        url = $"http://zhannei.baidu.com/cse/search?s=8823758711381329060&q={q}&page={pn}&p={pn - 1}";
-                        Thread.Sleep(1000 * 1);
-                        var html = Utils.HttpHelper.Get(url);
-                        doc.LoadHtml(html);
-                    }
-                    catch
-                    {
-                        //no.3
-                        url = $"http://zhannei.baidu.com/cse/search?s=3654077655350271938&q={q}&page={pn}&p={pn - 1}";
-                        Thread.Sleep(1000 * 2);
-                        HtmlWeb webClient = new HtmlWeb();
-                        doc = webClient.Load(url);
-                    }
+                    //no.2
+                    url = $"https://www.xxbiquge.com/search.php?keyword={q}&page={pn}&p={pn - 1}";
+                    Thread.Sleep(1000 * 1);
+                    var html = Utils.HttpHelper.Get(url);
+                    doc.LoadHtml(html);
                 }
             }
             catch (Exception ex)
@@ -136,6 +125,7 @@ public BookChapterDto GetBookChapters(string bookLink)
             var bookChapter = new BookChapterDto()
             {
                 BookName = doc.DocumentNode.SelectSingleNode("//div[@id='info']/h1").InnerText.Trim(),
+                BookLink = bookLink,
                 Author = nodes[0].InnerText.Replace(nodes[0].InnerText.Split('：')[0] + "：", string.Empty).Trim(),
                 Status = nodes[1].InnerText.Replace(nodes[1].InnerText.Split('：')[0] + "：", string.Empty).Replace(",加入书架,直达底部", string.Empty),
                 Last_Update_Time = nodes[2].InnerText.Replace(nodes[2].InnerText.Split('：')[0] + "：", string.Empty),

diff --git a/src/CatBookApp.Application/BookSearches/Captures/BookCaptureHelper.cs b/src/CatBookApp.Application/BookSearches/Captures/BookCaptureHelper.cs
@@ -31,8 +31,50 @@ public static string ClearSensitiveCharacter(string input)
                 input = input.Replace(item.ToString(), "");
             }
             //
+            input = HtmlToTxt(input);
 
             return input;
         }
+
+        /// <summary>
+        /// 对字符串进行检查和替换其中的特殊字符
+        /// </summary>
+        /// <param name="strHtml"></param>
+        /// <returns></returns>
+        public static string HtmlToTxt(string strHtml)
+        {
+            string[] aryReg ={
+                        @"<script[^>]*?>.*?</script>",
+                        @"<(\/\s*)?!?((\w+:)?\w+)(\w+(\s*=?\s*(([""'])(\\[""'tbnr]|[^\7])*?\7|\w+)|.{0})|\s)*?(\/\s*)?>",
+                        //@"([\r\n])[\s]+",
+                        @"&(quot|#34);",
+                        @"&(amp|#38);",
+                        @"&(lt|#60);",
+                        @"&(gt|#62);",
+                        @"&(nbsp|#160);",
+                        @"&(iexcl|#161);",
+                        @"&(cent|#162);",
+                        @"&(pound|#163);",
+                        @"&(copy|#169);",
+                        @"&#(\d+);",
+                        @"-->",
+                        @"<!--.*\n"
+                        };
+
+            //string newReg = aryReg[0];
+            string strOutput = strHtml;
+            for (int i = 0; i < aryReg.Length; i++)
+            {
+                Regex regex = new Regex(aryReg[i], RegexOptions.IgnoreCase);
+                strOutput = regex.Replace(strOutput, string.Empty);
+            }
+
+            strOutput.Replace("<", "");
+            strOutput.Replace(">", "");
+            strOutput.Replace("\r\n", "");
+
+
+            return strOutput;
+        }
     }
 }