C# JUST Library for JSON to JSON transformation

아티클
1/17/2024

Introduction

JUST Stands for JSON Under Simple Transformation.
The article describes how to use the JUST.NET library to transform JSON documents.

Background

XSLT is a very popular way of transforming XML documents using a simple transformation language. More and more applications are now using JSON as a data format because it is much simpler and less bulky than XML. However, there isn't a very easy way of transforming JSON documents.

We have created a library in .NET which enables the transformation of JSON documents using very a simple transformation language. This is an attempt to create an XSLT parallel for JSON. This article describes how to use that library.

Install the NuGet Package

Pull the latest JUST.NET from https://www.nuget.org
*
Install-Package JUST
*

Using the code

Once you download the NuGet package, you create a simple console application. Below is a simple C# code snippet that you can use to transform your JSON:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Threading.Tasks;
using JUST;
using System.IO;
namespace JUST.Test
{
    public class  Program
    {
        public static  void Main(string[] args)
        {
            string input = File.ReadAllText("Input.json");
            string transformer = File.ReadAllText("Transformer.json"); ;
            string transformedString = JsonTransformer.Transform(transformer, input);
            Console.WriteLine(transformedString);
            
        }
    }
}

Here Input.json is the input JSON document while Transformer.json is the JSON document that transforms the input JSON.

Using JUST to transform JSON

JUST is a transformation language just like XSLT. It includes functions which are used inside the transformer JSON to transform the input JSON in the desired output JSON. This section describes the various functions present in JUST and how they can be used to transform your JSON. Every JUST function starts with "#" character.

valueof

This function is used to extract the value of a given property. The value is extracted using the JSON path of the property. For more information on how to use JSON path refer to:

http://www.newtonsoft.com/json/help/html/QueryJsonSelectTokenJsonPath.htm

Consider the input:-
*
*

{
  "menu": {  
    "popup": {
      "menuitem": [
       {
          "value": "Open",
          "onclick": "OpenDoc()"
        },
        {
          "value": "Close",
          "onclick": "CloseDoc()"
        }
      ]
    }
  }
}

Transformer:

{
  "result": {
    "Open": "#valueof($.menu.popup.menuitem[?(@.value=='Open')].onclick)",
    "Close": "#valueof($.menu.popup.menuitem[?(@.value=='Close')].onclick)"
  }
}

Output:*
*

{
   "result":{"Open":"OpenDoc()","Close":"CloseDoc()"}
}

if condition

This condition is used to evaluate and if-else condition.
if condition(condition expression, evaluation expression, true result, false result).

All four of the parameters can be 'valueof' expressions or constants.

Consider the input:

{
  "menu": {
    "id" : "github",
    "repository" : "JUST"
  }
}

Transformer:

{
  "ifconditiontesttrue": "#ifcondition(#valueof($.menu.id),github,#valueof($.menu.repository),fail)",
  "ifconditiontestfalse": "#ifcondition(#valueof($.menu.id),xml,#valueof($.menu.repository),fail)"
}

Output:-

{
   "ifconditiontesttrue":"JUST",
   "ifconditiontestfalse":"fail"
}

string and math functions

At the moment only the basic and often used string and math functions are provided in the library.

lastindexof(input string,search string)
firstindexof(input string,search string)
substring(input string,start index,length)
concat(string 1,string 2)
add(value 1,value 2)
subtract(value 1,value 2)
multiply(value 1,value 2)
divide(value 1,values 2)

Consider the input:

{
  "stringref": "thisisandveryunuasualandlongstring",
  "numbers": [ "1", "2", "3", "4", "5" ]
}

Transformer:

{
  "stringresult": {
    "lastindexofand": "#lastindexof(#valueof($.stringref),and)",
    "firstindexofand": "#firstindexof(#valueof($.stringref),and)",
    "substring": "#substring(#valueof($.stringref),9,11)",
    "concat": "#concat(#valueof($.menu.id.file),#valueof($.menu.value.Window))"
  },
  "mathresult": {
    "add": "#add(#valueof($.numbers[0]),3)",
    "subtract": "#subtract(#valueof($.numbers[4]),#valueof($.numbers[0]))",
    "multiply": "#multiply(2,#valueof($.numbers[2]))",
    "divide": "#divide(9,3)"
  }
}

Output:

{"stringresult":
   {
    "lastindexofand":"21",
    "firstindexofand":"6",
    "substring":"veryunuasua",
    "concat":""
   },
   "mathresult":
   {
    "add":"4",
    "subtract":"4",
    "multiply":"6",
    "devide":"3"
   }
}

Aggregate functions

The following aggregate functions are provided for single dimensional arrays:

concatall(array)
sum(array)
average(array)
min(array)
max(array)

Consider the input:

{
  "d": [ "one",  "two", "three" ],
  "numbers": [ "1", "2", "3", "4", "5" ]
}

Transformer:

{
  "conacted": "#concatall(#valueof($.d))",
  "sum": "#sum(#valueof($.numbers))",
  "avg": "#average(#valueof($.numbers))",
  "min": "#min(#valueof($.numbers))",
  "max": "#max(#valueof($.numbers))"
}

Output:

{
    "conacted":"onetwothree",
    "sum":"15",
    "avg":"3",
    "min":"1",
    "max":"5"
}

Aggregate functions for multidimensional arrays

These functions are essentially the same as the above ones, the only difference being that you can also provide a path to point to a particular element inside the array.

concatallatpath(array,path)
sumatpath(array,path)
averageatpath(array,path)
minatpath(array,path)
maxatpath(array,path)

Consider the input:

{
   "x": [
    {
      "v": {
        "a": "a1,a2,a3",
        "b": "1",
        "c": "10"
      }
    },
    {
      "v": {
        "a": "b1,b2",
        "b": "2",
        "c": "20"
      }
    },
    {
      "v": {
        "a": "c1,c2,c3",
        "b": "3",
        "c": "30"
      }
    }
  ]
}

Transformer:

{
  "arrayconacted": "#concatallatpath(#valueof($.x),$.v.a)",
  "arraysum": "#sumatpath(#valueof($.x),$.v.c)",
  "arrayavg": "#averageatpath(#valueof($.x),$.v.c)",
  "arraymin": "#minatpath(#valueof($.x),$.v.b)",
  "arraymax": "#maxatpath(#valueof($.x),$.v.b)"
}

Output:

{
    "arrayconacted":"a1,a2,a3b1,b2c1,c2,c3",
    "arraysum":"60",
    "arrayavg":"20",
    "arraymin":"1",
    "arraymax":"3"
}

Bulk functions

All the above functions set property values to predefined properties in the output JSON. However, in some cases, we don't know what our output will look like as it depends on the input. Bulk functions are provided for this purpose. They correspond with the template-match functions in XSLT.

Bulk functions by law have to be the first property of the JSON object. All bulk functions are represented as array elements of the property '#'. These are the bulk functions provided as of now:

copy(path)
replace(path)
delete(path)

Consider the input:

{
  "tree": {
    "branch": {
      "leaf": "green",
      "flower": "red",
      "bird": "crow",
      "extra": { "twig":"birdnest" }
    },
    "ladder": {"wood": "treehouse" }
  }
}

Transformer:

{
  "#": [ "#copy($)",   "#delete($.tree.branch.bird)", "#replace($.tree.branch.extra,#valueof($.tree.ladder))" ],
  "othervalue" : "othervalue"
}

Output:
*
*

{
   "othervalue":"othervalue",
   "tree":{
    "branch":{
     "leaf":"green",
     "flower":"red",
     "extra":{
      "wood":"treehouse"
     }
    },
    "ladder":{
     "wood":"treehouse"
    }
  }
}

Array Looping

In some cases, we don't want to copy the entire array to the destination JSON. We might want to transform the array into a different format or have some special logic for each element while setting the destination JSON.
For these cases, we would use array looping.

These are the functions provided for this purpose:-

loop(path) - path is the path of the array to loop
currentvalue()
currentindex()
lastindex()
lastvalue()
currentvalueatpath(path) - here path denotes the path inside the array
lastvalueatpath(path) - here path denotes the path inside the array

Consider the input:

{
  "tree": {
    "branch": {
      "leaf": "green",
      "flower": "red",
      "bird": "crow",
      "extra": { "twig":  "birdnest" }
    },
    "ladder": { "wood":  "treehouse" }
  },
  "numbers": [ "1", "2", "3", "4" ],
  "arrayobjects": [
    {"country": {"name": "norway","language": "norsk"}},
    {
      "country": {
        "name": "UK",
        "language": "english"
      }
    },
    {
      "country": {
        "name": "Sweden",
        "language": "swedish"
      }
    }]
}

Transformer:

{
  "iteration": {
    "#loop($.numbers)": {
      "CurrentValue": "#currentvalue()",
      "CurrentIndex": "#currentindex()",
      "IsLast": "#ifcondition(#currentindex(),#lastindex(),yes,no)",
      "LastValue": "#lastvalue()"
    }
  },
  "iteration2": {
    "#loop($.arrayobjects)": {
      "CurrentValue": "#currentvalueatpath($.country.name)",
      "CurrentIndex": "#currentindex()",
      "IsLast": "#ifcondition(#currentindex(),#lastindex(),yes,no)",
      "LastValue": "#lastvalueatpath($.country.language)"
    }
  },
  "othervalue": "othervalue"
}

Output:

{
"iteration":[
   {"CurrentValue":"1","CurrentIndex":"0","IsLast":"no","LastValue":"4"},
   {"CurrentValue":"2","CurrentIndex":"1","IsLast":"no","LastValue":"4"},
   {"CurrentValue":"3","CurrentIndex":"2","IsLast":"no","LastValue":"4"},
   {"CurrentValue":"4","CurrentIndex":"3","IsLast":"yes","LastValue":"4"}
  ],
   "iteration2":[
   {"CurrentValue":"norway","CurrentIndex":"0","IsLast":"no","LastValue":"swedish"},
   {"CurrentValue":"UK","CurrentIndex":"1","IsLast":"no","LastValue":"swedish"},
   {"CurrentValue":"Sweden","CurrentIndex":"2","IsLast":"yes","LastValue":"swedish"}
  ],
"othervalue":"othervalue"
}

Calling Custom functions

You can make your own custom functions in C# and call them from your transformer JSON.
A custom function has to reside inside a public class and has to be a public static method.
A custom function is called using the following syntax:-
#customfunction(dll name, FQN for the static function, argument1.......,argument)

Consider the following input:

{
  "tree": {
    "branch": {
      "leaf": "green",
      "flower": "red",
      "bird": "crow"
    }
  }
}

Transformer:

{
  "Season": "#customfunction(JUST.NET.Test,JUST.NET.Test.Season.findseason,#valueof($.tree.branch.leaf),#valueof($.tree.branch.flower))"
}

Custom function:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
namespace JUST.NET.Test
{
    public class  Season
    {
        public static  string findseason(string leafColour, string flowerColour)
        {
            if (leafColour == "green" && flowerColour == "red")
                return "summer";
            else
                return "winter";
        }
    }
}

Output:

{"Season": "summer"}

Link to test source code

We have made a GitHub repository which contains a C# project and has various transformation scenarios on an input JSON file.
https://github.com/WorkMaze/JUST.net.

다음을 통해 공유

C# JUST Library for JSON to JSON transformation

Introduction

Background

Install the NuGet Package

Using the code

Using JUST to transform JSON

valueof

if condition

string and math functions

Aggregate functions

Aggregate functions for multidimensional arrays

Bulk functions

Array Looping

Calling Custom functions

Link to test source code

추가 리소스