从 json 数组连接自定义对象的字符串属性

Question

这里是问题的简要说明。 从服务器我得到 json 数组，它看起来类似于：

    [
  {
    "messagenumber": "0529069f-a403-4eea-a955-430a10995745",
    "message": "A",
    "partnumber": 1,
    "total": 3
  },
  {
    "messagenumber": "0529069f-a403-4eea-a955-430a10995745",
    "message": "B",
    "partnumber": 2,
    "total": 3
  },
  {
    "messagenumber": "0529069f-a403-4eea-a955-430a10995745",
    "message": "C",
    "partnumber": 3,
    "total": 3
  },
  {
    "messagenumber": "52e7d68d-462b-46b9-8eec-f289bcdf7b06",
    "message": "AA",
    "partnumber": 1,
    "total": 2
  }......
]

它表示消息部分的列表。属于同一组的消息部分被赋予相同的messagenumber，partnumber代表订单号。最后，在客户端上，我必须组合属于同一组的所有消息部分并呈现完整消息。从上面的示例中，单个消息的结果将如下所示：

var msg = new FullMessage();
msg.MessageNumber = "0529069f-a403-4eea-a955-430a10995745";
msg.Message = "ABC"; //here we have full message created out of 3 parts
msg.PartNumber = 1; //not important 
msg.Total = 1 //also not important

最终结果应该是一个 FullMessage 对象的集合。 此外，应该忽略部分消息（并非所有部分都存在）。

这可以使用 LINQ 语句优雅地完成吗？

感谢您的任何建议。

Answer 1

这应该做你想做的：

var myList = new List<FullMessage>();
myList.GroupBy(msg => msg.MessageNumber)
      .Where(grouping => grouping.Count() == grouping.First().Total)
      .Select(grouping => grouping.OrderBy(x => x.order))
      .Select(grouping => grouping.Select(x => x.Message)
      .Aggregate((x, y) => x + y)); 

它的作用如下：

1. 通过MessageNumber 对消息进行分组
2. 过滤掉所有部分消息
3. 按PartNumber 对每个分组的所有消息进行排序，以确保所有消息的顺序正确
4. 浏览所有分组并选择所有消息
5. 聚合消息

结果是串联字符串的枚举。

Answer 2

下面的示例代码：

void Main()
{
    // Testing mocked data
    var data = new List<ServerMessagePart> {
        new ServerMessagePart{
            Messagenumber = "0529069f-a403-4eea-a955-430a10995745",
            Message= "A",
            Partnumber= 1,
            Total = 3
        },
        new ServerMessagePart{
            Messagenumber = "0529069f-a403-4eea-a955-430a10995745",
            Message= "C",
            Partnumber= 3,
            Total = 3
        },
        new ServerMessagePart{
            Messagenumber = "0529069f-a403-4eea-a955-430a10995745",
            Message= "B",
            Partnumber= 2,
            Total = 3
        },
        new ServerMessagePart{
            Messagenumber = "52e7d68d-462b-46b9-8eec-f289bcdf7b06",
            Message= "AA",
            Partnumber= 1,
            Total = 2
        }
    };

    // Compiled messages
    var messages = new List<ServerMessage>();

    // Process server partial messages - group parsts by message number
    var partialMessages = data.GroupBy(x => x.Messagenumber);
    foreach(var messageParts in partialMessages) {

        // Get expected parts number
        var expected = messageParts.Max(x=>x.Total);

        // Skip incompleted messages
        if(messageParts.Count() < expected) {
            continue;
        }

        // Sort messages and compile the message
        var message = messageParts.OrderBy(x => x.Partnumber).Select(x => x.Message).Aggregate((x,y) => x + y);

        // Final message
        messages.Add(new ServerMessage{
            Messagenumber = messageParts.First().Messagenumber,
            Message = message
        });
    }

    // Final messages
}

class ServerMessage {
    public string Messagenumber {get; set;}
    public string Message {get; set;}
}

class ServerMessagePart {
    public string Messagenumber {get; set;}
    public string Message {get; set;}
    public int Partnumber {get; set;}
    public int Total {get; set;}
}

Answer 3

假设消息看起来像

public class FullMessage
{
    public string MessageNumber { get; set; }
    public string Message { get; set; }
    public int PartNumber { get; set; }
    public int Total { get; set; }
}

这是我可能的解决方案的想法：

1. 按消息编号分组
2. 按部件号对组中的每条消息进行排序
3. 使用Aggregate 在新的FullMessage 对象中组合消息部分并计算总数（不确定是否需要计算，所以这是可选的）

结果是这个 LINQ 查询：

IEnumerable<FullMessage> combinedMessages = messages.GroupBy(m => m.MessageNumber)
    .Select(g => g.OrderBy(m => m.PartNumber)
    .Aggregate(new FullMessage() { MessageNumber = g.Key }, (r, m) =>
    {
        r.Message += m.Message;
        //Not sure if counting up is required in your code
        //so the following line is optional
        r.Total += m.Total;
        return r;
    }));

Answer 4

这是我的：（已更新以删除部分内容）

var messages = new Message[] {
    new Message("52e7d68d-462b-46b9-8eec-f289bcdf7b06", "BB", 2, 2),
    new Message("0529069f-a403-4eea-a955-430a10995745", "A", 1, 3),
    new Message("0529069f-a403-4eea-a955-430a10995745", "B", 2, 3),
    new Message("0529069f-a403-4eea-a955-430a10995745", "C", 3, 3),
    new Message("52e7d68d-462b-46b9-8eec-f289bcdf7b06", "AA", 1, 2),
};

    var grouped = (from m in messages
                  group m by m.messagenumber into g
                  select new
                  {
                    messagenumber = g.Key,
                    message = String.Join("", g.OrderBy(dr =>dr.partnumber)
                                               .Select(m=>m.message)),
                    count = g.Count(),
                    total = g.First().total
                  }).Where(c=>c.count == c.total);

Answer 5

试试这个代码：

var input = new List<FullMessage>();

var result = input.GroupBy(m => new { m.MessageNumber, m.Total })
    .Where(g => g.Count() == g.Key.Total)
    .Select(g => new FullMessage
    {
        MessageNumber = g.Key.MessageNumber,
        Message = g.OrderBy(i => i.PartNumber).Aggregate("", (c, n) => c + n.Message)
    });