How to change an unnormalized csv file to a complex json or java object

I have the following unnormalized csv file

user_id,nickname,joinDate,product_id,price
1,kmh,2023-07-24,P131,3000
1,kmh,2023-07-24,P132,4000
1,kmh,2023-07-24,P133,7000
1,kmh,2023-07-24,P134,9000
2,john,2023-07-24,P135,2500
2,john,2023-07-24,P136,6000
3,alice,2023-07-25,P137,4500
3,alice,2023-07-25,P138,8000

I'm going to change this to the following json format (or java object).

[
    {
        "user_id": 1,
        "nickname": "kmh",
        "joinDate": "2023-07-24",
        "orders": [
            {
                "product_id": "P131",
                "price": 3000
            },
            {
                "product_id": "P132",
                "price": 4000
            },
            {
                "product_id": "P133",
                "price": 7000
            },
            {
                "product_id": "P134",
                "price": 9000
            }
        ]
    },
    {
        "user_id": 2,
        "nickname": "john",
        "joinDate": "2023-07-24",
        "orders": [
            {
                "product_id": "P135",
                "price": 2500
            },
            {
                "product_id": "P136",
                "price": 6000
            }
        ]
    },
    {
        "user_id": 3,
        "nickname": "alice",
        "joinDate": "2023-07-25",
        "orders": [
            {
                "product_id": "P137",
                "price": 4500
            },
            {
                "product_id": "P138",
                "price": 8000
            }
        ]
    }
]

I've been searching for quite a long time and haven't found a library or tool that enables this .

I have so many different types of csv that I need tools or libraries to change all of these. Are there any libraries or tools that make this possible?

All you need is a way to parse the CSV to a Java object. You can do this manually, or by using an existing library.

For example, you can use Jackson with the CSV data format like this:

class MyRecord {
@JsonProperty("user_id")
private int userId;
private String nickname;
private LocalDate joinDate;
@JsonProperty("product_id")
private String productId;
// getters and setters
// a meaningful toString method
}
public class Main {
public static void main(String[] args) throws IOException {
String csv = "user_id,nickname,joinDate,product_id,price\n" +
"1,kmh,2023-07-24,P131,3000\n" +
"1,kmh,2023-07-24,P132,4000\n" +
"1,kmh,2023-07-24,P133,7000\n" +
"1,kmh,2023-07-24,P134,9000\n" +
"2,john,2023-07-24,P135,2500\n" +
"2,john,2023-07-24,P136,6000\n" +
"3,alice,2023-07-25,P137,4500\n" +
"3,alice,2023-07-25,P138,8000";
CsvSchema schema = CsvSchema.emptySchema().withHeader(); // uses CSV header to read the schema
ObjectMapper mapper = new CsvMapper().registerModule(new JavaTimeModule()); // to deserialise Java 8 LocalDate
MappingIterator<MyRecord> resultIterator = mapper.readerFor(MyRecord.class).with(schema).readValues(csv);
while (resultIterator.hasNext()) {
System.out.println(resultIterator.next());
}
resultIterator.close();
}
} 

Which prints:

MyRecord[userId=1, nickname='kmh', joinDate=2023-07-24, productId='P131', price=3000]
MyRecord[userId=1, nickname='kmh', joinDate=2023-07-24, productId='P132', price=4000]
MyRecord[userId=1, nickname='kmh', joinDate=2023-07-24, productId='P133', price=7000]
MyRecord[userId=1, nickname='kmh', joinDate=2023-07-24, productId='P134', price=9000]
MyRecord[userId=2, nickname='john', joinDate=2023-07-24, productId='P135', price=2500]
MyRecord[userId=2, nickname='john', joinDate=2023-07-24, productId='P136', price=6000]
MyRecord[userId=3, nickname='alice', joinDate=2023-07-25, productId='P137', price=4500]
MyRecord[userId=3, nickname='alice', joinDate=2023-07-25, productId='P138', price=8000]

For your case, you can deserialize the csv into JsonNode directly without creating POJO class. And then use JSON library Josson to transform the JSON by function group().

String csv = "user_id,nickname,joinDate,product_id,price\n" +
"1,kmh,2023-07-24,P131,3000\n" +
"1,kmh,2023-07-24,P132,4000\n" +
"1,kmh,2023-07-24,P133,7000\n" +
"1,kmh,2023-07-24,P134,9000\n" +
"2,john,2023-07-24,P135,2500\n" +
"2,john,2023-07-24,P136,6000\n" +
"3,alice,2023-07-25,P137,4500\n" +
"3,alice,2023-07-25,P138,8000";
ArrayNode arrayNode = Josson.createArrayNode();
CsvSchema schema = CsvSchema.emptySchema().withHeader();
try (MappingIterator<JsonNode> it = new CsvMapper().readerFor(JsonNode.class).with(schema).readValues(csv)) {
arrayNode.addAll(it.readAll());
}
Josson josson = Josson.create(arrayNode);
JsonNode grouped = josson.getNode(
"group(map(user_id, nickname, joinDate), map(product_id, price))" +
".map(**:key, orders:elements)");
System.out.println(grouped.toPrettyString());

Function group()

1. Group by "key" of {user_id, nickname, joinDate}
2. With "elements" of {product_id, price}

Functoin map()

1. Extract values inside object "key"
2. Add field "orders" copy from "elements"

Output

[ {
"user_id" : "1",
"nickname" : "kmh",
"joinDate" : "2023-07-24",
"orders" : [ {
"product_id" : "P131",
"price" : "3000"
}, {
"product_id" : "P132",
"price" : "4000"
}, {
"product_id" : "P133",
"price" : "7000"
}, {
"product_id" : "P134",
"price" : "9000"
} ]
}, {
"user_id" : "2",
"nickname" : "john",
"joinDate" : "2023-07-24",
"orders" : [ {
"product_id" : "P135",
"price" : "2500"
}, {
"product_id" : "P136",
"price" : "6000"
} ]
}, {
"user_id" : "3",
"nickname" : "alice",
"joinDate" : "2023-07-25",
"orders" : [ {
"product_id" : "P137",
"price" : "4500"
}, {
"product_id" : "P138",
"price" : "8000"
} ]
} ]