CODEWITHSUNDEEP
jsonjson-ld
Franz Kafka
Franz Kafka

Reputation: 844

Map one JSON schema to a different JSON schema

I have a bunch of JSON files, thousands of different schemas. Using GenSON (the Python JSON schema generator), I managed to create schema files for each of the input files. Now, what I'd like to do is standardize all these different files to one defined schema. Here's an example:

Input

{
     "name": "Bob Odenkirk",
     "title": "Software Engineer",
     "location": {
         "locality": "San Francisco",
         "region": "CA",
         "country": "United States"
     },
     "age": 62,
     "status": "Active"
}

Output

{
     "names": ["Bob Odenkirk"],
     "occupations": ["Software Engineer"],
     "locations": ["San Francisco, CA"]    
}

Essentially, I am looking for a language agnostic method (i.e., I don't care what programming language is used) of defining how an input JSON file should be parsed to an output JSON file.

Upvotes: 7

Views: 12120

Answers (4)

omkaartg
omkaartg

Reputation: 2787

I think the best, fastest, easiest way to parse many JSON files together is using python.
I was doing something similar to your project and ran into the same problem. I found this site which teaches how to use python to actually parse JSON files together. Turns out there is a library on python called json(use pip to download json dependencies) which enables JSON file processing. If you already have a python editor, This method would be easier and faster then using Jolt Check This website for more info: https://code.tutsplus.com/tutorials/how-to-work-with-json-data-using-python--cms-25758.
You can also use JS, which is again faster than Jolt. this is the website: https://learn.microsoft.com/en-us/scripting/javascript/reference/json-parse-function-javascript . It is very easy as you can use JSON.parse() function

Upvotes: 0

Milo S
Milo S

Reputation: 4586

Jolt Spec 

[
  // First build the "city, state" string for location
  {
    "operation": "modify-default-beta",
    "spec": {
      "location": {
        "locConcat": "=concat(@(1,locality),', ',@(1,region))"
      }
    }
  },
  // Then map the fields as needed to positions in an output json
  {
    "operation": "shift",
    "spec": {
      "name": "name[0]",
      "title": "occupations[0]",
      "location": {
        "locConcat": "locations[0]"
      }
    }
  }
]

Upvotes: 2

Kalaikumar Thangasamy
Kalaikumar Thangasamy

Reputation: 564

I am not sure is your expecting like below. Long time back I have created flat object and output format object. It will return output format object with data filled.

var input = {
     "name": "Bob Odenkirk",
     "title": "Software Engineer",
     "location": {
         "locality": "San Francisco",
         "region": "CA",
         "country": "United States"
     },
     "age": 62,
     "status": "Active"
};

var outputFormat = {
    "name": "name",
  "occupations": "title",
  "locations": "location.locality, location.region"
};

var flatInput = {};

function generateFlatInput(input, parent){
     for (var prop in input) {
        if(input.hasOwnProperty(prop) && typeof input[prop] === 'object')
            flatInput = generateFlatInput(input[prop], parent + prop + '.');
        else
         flatInput[parent + prop] =  input[prop];
    }

    return flatInput;
}

function generateOutput(input, outputFormat, delimiter){
     input = generateFlatInput(input, '');

     for (var prop in outputFormat) {
        var fields = outputFormat[prop].split(delimiter);
      var fieldValue = [];
      for(i = 0; i < fields.length; i++){
                    if(!input.hasOwnProperty(fields[i].trim())) continue;

          fieldValue.push(input[fields[i].trim()]);
      }
      outputFormat[prop] = fieldValue.join(delimiter);
   }

   return outputFormat;
}

console.log(generateOutput(input, outputFormat, ', '));

https://jsfiddle.net/u2yyuguk/1/

Upvotes: 1

mikep
mikep

Reputation: 3895

The url https://github.com/bazaarvoice/jolt#jolt says that Jolt may be what you're looking for.

Jolt

JSON to JSON transformation library written in Java where the "specification" for the transform is itself a JSON document.

Useful For

Transforming JSON data from ElasticSearch, MongoDb, Cassandra, etc before sending it off to the world

Extracting data from a large JSON documents for your own consumption

Upvotes: 6

Related Questions