ChoETL 嵌套 JSON 到 CSV

Question

我需要将 json 转换为 csv。问题是我不能 select 在嵌套 json 结构中我需要的一切。 json 文件示例：

{
    "system": {
        "created": "2021-08-01T13:33:37.123Z",
        "by": "web"
    },
    "location": {
        "id": 100,
        "country": "DE"
    },
    "order": [
        {
            "OrderID": 22,
            "OrderName": "Soda",
            "OrderArticles": [
                {
                    "Size": 33,
                    "ProductName": "Coke",
                    "ProductId": "999"
                },
                {
                    "Size": 66,
                    "ProductName": "Fanta",
                    "ProductId": "888"
                },
                {
                    "Size": 50,
                    "ProductName": "Pepsi",
                    "ProductId": "444"
                }
            ],
            "ProcessedId": 1001,
            "Date": "2021-08-02"
        },
        {
            "OrderID": 23,
            "OrderName": "Beverage",
            "OrderArticles": [
                {
                    "Size": 44,
                    "ProductName": "Coke",
                    "ProductId": "999"
                }
            ],
            "ProcessedId": 1002,
            "Date": "2021-08-03"
        }
    ]
}

这是我想要的输出：

created;by;id;country;OrderID;OrderName;Size;ProductName;ProductId
2021-08-01T13:33:37.123Z;web;100;DE;22;Soda;33;Coke;999
2021-08-01T13:33:37.123Z;web;100;DE;22;Soda;66;Fanta;888
2021-08-01T13:33:37.123Z;web;100;DE;22;Soda;50;Pepsi;444
2021-08-01T13:33:37.123Z;web;100;DE;23;Beverage;44;Coke;999

我可以自己获取 created 和 by 值以及 OrderArticles 的值。我只是不知道如何让他们在一起。这是我用来获取结果但分为 2 个不同结果的代码：

using (var r = new ChoJSONReader(inBlob).WithJSONPath("$..order[*]").AllowComplexJSONPath(true))
{
    return (r.SelectMany(r1 => ((dynamic[])r1.OutputArticles).Select(r2 => new
    {
        r1.OrderID,
        r1.OrderName,
        r1.Size,
        r1.ProductName,
        r1.ProductId
    })));
}



using (var r = new ChoJSONReader(inBlob).WithJSONPath("$").AllowComplexJSONPath(true))
{
    return (r.Select(r1 => new
    {
        r1.system.created,
        r1.system.by
    }));
}

Answer 1

这是我的解决方案。

这是我的数据模型：

using System.Text.Json.Serialization;

namespace JsonToCSV.Models;

// Root myDeserializedClass = JsonSerializer.Deserialize<Root>(myJsonResponse);
public class System
{
    [JsonPropertyName("created")]
    public string Created { get; set; }

    [JsonPropertyName("by")]
    public string By { get; set; }
}

public class Location
{
    [JsonPropertyName("id")]
    public int Id { get; set; }

    [JsonPropertyName("country")]
    public string Country { get; set; }
}

public class OrderArticle
{
    [JsonPropertyName("Size")]
    public int Size { get; set; }

    [JsonPropertyName("ProductName")]
    public string ProductName { get; set; }

    [JsonPropertyName("ProductId")]
    public string ProductId { get; set; }
}

public class Order
{
    [JsonPropertyName("OrderID")]
    public int OrderID { get; set; }

    [JsonPropertyName("OrderName")]
    public string OrderName { get; set; }

    [JsonPropertyName("OrderArticles")]
    public List<OrderArticle> OrderArticles { get; set; }

    [JsonPropertyName("ProcessedId")]
    public int ProcessedId { get; set; }

    [JsonPropertyName("Date")]
    public string Date { get; set; }
}

public class Root
{
    [JsonPropertyName("system")]
    public System System { get; set; }

    [JsonPropertyName("location")]
    public Location Location { get; set; }

    [JsonPropertyName("order")]
    public List<Order> Orders { get; set; }
}

这是业务逻辑（如果你愿意，我可以用 LINQ 替换它）：

using System.Text.Json;
using JsonToCSV.Models;

var dataAsText = File.ReadAllText("data.json");

var data = JsonSerializer.Deserialize<Root>(dataAsText);

var csv = new List<string> { "created;by;id;country;OrderID;OrderName;Size;ProductName;ProductId" };

foreach (var order in data.Orders)
{
    foreach (var orderArticle in order.OrderArticles)
    {
        csv.Add(String.Format("{0};{1};{2};{3};{4};{5};{6};{7};{8}",
            data.System.Created,
            data.System.By,
            data.Location.Id,
            data.Location.Country,
            order.OrderID,
            order.OrderName,
            orderArticle.Size,
            orderArticle.ProductName,
            orderArticle.ProductId
        ));
    }    
}

File.WriteAllLines("data.csv", csv);

创建 .csv 文件，内容为：

created;by;id;country;OrderID;OrderName;Size;ProductName;ProductId
2021-08-01T13:33:37.123Z;web;100;DE;22;Soda;33;Coke;999
2021-08-01T13:33:37.123Z;web;100;DE;22;Soda;66;Fanta;888
2021-08-01T13:33:37.123Z;web;100;DE;22;Soda;50;Pepsi;444
2021-08-01T13:33:37.123Z;web;100;DE;23;Beverage;44;Coke;999

Answer 2

由于需要system.created、system.by、location.id、location.country字段，必须从root加载整个json，然后组合预期的csv

的对象

这是工作示例（采用最新的 nuget 包）

方法一：（使用动态模型）

StringBuilder csv = new StringBuilder();
using (var r = new ChoJSONReader("*** YOUR JSON FILE PATH ***")
       .JsonSerializationSettings(s => s.DateParseHandling = DateParseHandling.None)
      )
{
    using (var w = new ChoCSVWriter(csv)
        .WithDelimiter(";")
        .WithFirstLineHeader())
    {
        w.Write(r.SelectMany(root =>
            ((Array)root.order).Cast<dynamic>()
            .SelectMany(order => ((Array)order.OrderArticles).Cast<dynamic>()
            .Select(orderarticle => new
            {
                root.system.created,
                root.system.by,
                root.location.id,
                order.OrderID,
                order.OrderName,
                orderarticle.Size,
                orderarticle.ProductName,
                orderarticle.ProductId,
            })
                )
            )
        );
    }
}
Console.WriteLine(csv.ToString());

输出：

created;by;id;OrderID;OrderName;Size;ProductName;ProductId
2021-08-01T01:33:37.123Z;web;100;22;Soda;33;Coke;999
2021-08-01T01:33:37.123Z;web;100;22;Soda;66;Fanta;888
2021-08-01T01:33:37.123Z;web;100;22;Soda;50;Pepsi;444
2021-08-01T01:33:37.123Z;web;100;23;Beverage;44;Coke;999

方法二：使用POCO模型

定义与输入匹配的 POCO 对象 JSON

public class System
{
    [JsonProperty("created")]
    public string Created { get; set; }

    [JsonProperty("by")]
    public string By { get; set; }
}

public class Location
{
    [JsonProperty("id")]
    public int Id { get; set; }

    [JsonProperty("country")]
    public string Country { get; set; }
}

public class OrderArticle
{
    [JsonProperty("Size")]
    public int Size { get; set; }

    [JsonProperty("ProductName")]
    public string ProductName { get; set; }

    [JsonProperty("ProductId")]
    public string ProductId { get; set; }
}

public class Order
{
    [JsonProperty("OrderID")]
    public int OrderID { get; set; }

    [JsonProperty("OrderName")]
    public string OrderName { get; set; }

    [JsonProperty("OrderArticles")]
    public List<OrderArticle> OrderArticles { get; set; }

    [JsonProperty("ProcessedId")]
    public int ProcessedId { get; set; }

    [JsonProperty("Date")]
    public string Date { get; set; }
}

public class OrderRoot
{
    [JsonProperty("system")]
    public System System { get; set; }

    [JsonProperty("location")]
    public Location Location { get; set; }

    [JsonProperty("order")]
    public List<Order> Orders { get; set; }
}

然后使用下面的代码加载 json 并以预期格式输出 CSV

StringBuilder csv = new StringBuilder();
using (var r = new ChoJSONReader<OrderRoot>("*** YOUR JSON FILE PATH ***")
    .UseJsonSerialization()
    )
{
    using (var w = new ChoCSVWriter(csv)
        .WithDelimiter(";")
        .WithFirstLineHeader())
    {
        w.Write(r.SelectMany(root =>
            root.Orders
            .SelectMany(order => order.OrderArticles
            .Select(orderarticle => new
            {
                created = root.System.Created,
                by = root.System.By,
                id = root.Location.Id,
                order.OrderID,
                order.OrderName,
                orderarticle.Size,
                orderarticle.ProductName,
                orderarticle.ProductId,
            })
                )
            )
        );
    }
}
Console.WriteLine(csv.ToString());

方法 3：简化动态模型方法

StringBuilder csv = new StringBuilder();
using (var r = new ChoJSONReader("*** YOUR JSON FILE PATH ***")
    .WithField("created", jsonPath: "$..system.created", isArray: false, valueConverter: o => ((DateTime)o).ToString("yyyy-MM-ddThh:mm:ss.fffZ"))
    .WithField("by", jsonPath: "$..system.by", isArray: false)
    .WithField("id", jsonPath: "$..location.id", isArray: false)
    .WithField("country", jsonPath: "$..location.country", isArray: false)
    .WithField("OrderID")
    .WithField("OrderName")
    .WithField("Size")
    .WithField("ProductName")
    .WithField("ProductId")
    .Configure(c => c.FlattenNode = true)
    )
{
    using (var w = new ChoCSVWriter(csv)
        .WithDelimiter(";")
        .WithFirstLineHeader())
    {
        w.Write(r);
    }
}
Console.WriteLine(csv.ToString());

方法 4：更简化的动态模型方法

StringBuilder csv = new StringBuilder();
using (var r = new ChoJSONReader("*** YOUR JSON FILE PATH ***")
    .Configure(c => c.FlattenNode = true)
    .JsonSerializationSettings(s => s.DateParseHandling = DateParseHandling.None)
    )
{
    using (var w = new ChoCSVWriter(csv)
        .WithDelimiter(";")
        .WithFirstLineHeader()
        .Configure(c => c.IgnoreDictionaryFieldPrefix = true)
        )
    {
        w.Write(r);
    }
}
Console.WriteLine(csv.ToString());

样本fiddle：https://dotnetfiddle.net/VCezp8

ChoETL 嵌套 JSON 到 CSV

ChoETL nested JSON to CSV

c#

json

linq

csv

choetl