Why is data serialization needed?

Data serialization is necessary to convert a data structure into a stream of bytes that can be easily transmitted over a network, written to a file, or stored in a database. The reverse process, deserialization, restores the data structure from this byte stream.

Main reasons for using serialization:

• Data transfer: Objects in a program's memory have a specific structure and references that cannot be directly sent over a network. Serialization converts them into a universal format.
• Data storage: Structured data can be conveniently saved in files or databases after serialization.
• Data exchange between different systems: Systems written in different languages or running on different platforms can exchange data using standardized serialization formats.
• Caching: Serialized objects can be easily stored in cache for quick access.
• Remote Procedure Call (RPC): Passing arguments and results between processes or machines often uses serialization.

Popular serialization formats include:

• JSON
• XML
• Protocol Buffers
• MsgPack
• Gob (Golang-specific)

Example of serialization in Go using JSON:

package main

import (
	"encoding/json"
	"fmt"
)

type Person struct {
	Name string `json:"name"` // Tag for mapping the Name field in JSON with key "name"
	Age  int    `json:"age"`
}

func main() {
	p := Person{Name: "Alice", Age: 30}

	// Serialization to JSON
	jsonData, err := json.Marshal(p)
	if err != nil {
		fmt.Println("Error marshalling:", err)
		return
	}
	fmt.Println(string(jsonData)) // {"name":"Alice","age":30}

	// Deserialization back from JSON
	var p2 Person
	err = json.Unmarshal(jsonData, &p2)
	if err != nil {
		fmt.Println("Error unmarshalling:", err)
		return
	}
	fmt.Printf("%+v\n", p2) // {Name:Alice Age:30}
}