避免在 C# 中重复方法实现

Question

我有几个类，我需要解析 json 对象。我看到这个 json 对象的初始循环在所有类中都几乎相同，除了子方法。

例如在Class1.cs中

private static void FindObject(JToken token)
{
  switch (token.Type)
  {
      case JTokenType.Array:          
         JArray array = token as JArray;
         array.ForEach(a => FindObject(a));
         break;           
      case JTokenType.String:
         token.Replace(GetNewImgTag(token.ToString()));
         break;
      case JTokenType.Object:
         token.Children().ForEach(t => FindObject(t));
         break;
       case JTokenType.Property:
         JProperty prop = token as JProperty;

         if (prop.Value.Type == JTokenType.Array)
         {
             FindObject(prop.Value);
             return;
         }

         prop.Value = GetNewImgTag(prop.Value.ToString());
         break;
      default:
         throw new NotImplementedException(token.Type + " is not defined");
  }    
}

private static JToken GetNewImgTag(string text)
{
   ...
}

和 Class 2.cs 是

private static void FindObject(JToken token)
{
  switch (token.Type)
  {
      case JTokenType.Array:          
         JArray array = token as JArray;
         array.ForEach(a => FindObject(a));
         break;           
      case JTokenType.String:
         token.Replace(ReplaceLinks(token.ToString()));
         break;
      case JTokenType.Object:
         token.Children().ForEach(t => FindObject(t));
         break;
       case JTokenType.Property:
         JProperty prop = token as JProperty;

         if (prop.Value.Type == JTokenType.Array)
         {
             FindObject(prop.Value);
             return;
         }

         prop.Value = ReplaceLinks(prop.Value.ToString());
         break;
      default:
         throw new NotImplementedException(token.Type + " is not defined");
  }    
}

private static JToken ReplaceLinks(string text)
{
   ...
}

如果你比较这两个类，FindObject() 几乎相同，除了子方法调用。我需要在几个班级中实现这一点。我正在尝试避免这种多重重复方法的创建。

谁能提出更好的设计方法？

我在这里看到了类似的帖子，但我无法将此代表应用于我的场景。

Avoiding repetitive code in multiple similar methods (C#)

Answer 1

一种直接的方法是识别不同的部分，并将其作为您传递给单独函数的委托。

这是一个工作示例。

public static class MyTokenReaderUtilities
{
     public static void ConvertEachProperty(JToken token, Func<string, JToken> convertString)
     {
        switch (token.Type)
        {
            case JTokenType.Array:          
               JArray array = token as JArray;
              array.ForEach(a => ConvertEachProperty(a, convertString));
              break;           
           case JTokenType.String:
             token.Replace(convertString(token.ToString()));
           break;
          case JTokenType.Object:
              token.Children().ForEach(t => ConvertEachProperty(t, convertString));
              break;
          case JTokenType.Property:
              JProperty prop = token as JProperty;

              if (prop.Value.Type == JTokenType.Array)
              {
                ConvertEachProperty(prop.Value, convertString);
                return;
              }
              prop.Value = convertString(prop.Value.ToString());
              break;
          default:
             throw new NotImplementedException(token.Type + " is not defined");
       }  
     }
}

现在，在第 1 课中：

private static void FindObject(JToken token)
{
   MyTokenReaderUtilities.ConvertEachProperty(token, GetNewImgTag);
}

在第 2 课中：

private static void FindObject(JToken token)
{
   MyTokenReaderUtilities.ConvertEachProperty(token, ReplaceLinks);
}