Struct in Go (Golang)

admin

5 years ago

This is the chapter 16 of the golang comprehensive tutorial series. Refer to this link for other chapters of the series – Golang Comprehensive Tutorial Series

Next Tutorial – Array
Previous Tutorial – Pointer

Now let’s check out the current tutorial. Below is the table of contents for current tutorial

Table of Contents

Overview

GO struct is named collection of data fields which can be of different types. Struct acts as a container that has different heterogeneous data types which together represents an entity. For example, different attributes are used to represent an employee in an organization. Employee can have

Name of string type

Age of int type

DOB of time.Time type

Salary of int type

.. and so on. A struct can be used to represent an employee

type employee struct {
    name   string
    age    int
    salary int
}

A struct in golang can be compared to a class in Object Oriented Languages

Declaring a struct type

Below is the format for declaring a struct

type struct_name struct {
    field_name1 field_type1
    field_name2 field_type2
    ...
}

In the above format, struct_name is the name of the struct. It has a field named field_name1 of type field_type1 and a field named field_name2 of type field_type2. This declares a new named struct type which acts as a blueprint. The type keyword is used to introduce a new type

Example

type point struct {
    x float64
    y float64
}

The above declaration declares a new struct named point which has two field x and y. Both fields are of float64 type.Once a new struct type is declared we can define new concrete struct variable from it as we will see in next section

Creating a struct variable

Declaring a struct just declares a named struct type. Creating a struct variable creates an instance of that struct with memory being initialized as well. We can create a empty struct variable without given any value to any of the field

emp := employee{}

In this case, all the fields in the struct are initialized with a default zero value of that field type.

We can also initialize the value for each struct field while creating a struct variable. There are two variations

Each field on the same line

emp := employee{name: "Sam", age: 31, salary: 2000}

Each field on different lines

emp := employee{
   name:   "Sam",
   age:    31,
   salary: 2000,
}

It is also ok to initialize only some of the fields with value. The field which are not initialized with value will get the default zero value of their type

emp := employee{
   name: "Sam",
   age: 31,
}

In above case salary will get default value of zero since it is not initialized

Let’s see a working code illustrating above points:

package main

import "fmt"

type employee struct {
    name   string
    age    int
    salary int
}

func main() {
    emp1 := employee{}
    fmt.Printf("Emp1: %+v\n", emp1)

    emp2 := employee{name: "Sam", age: 31, salary: 2000}
    fmt.Printf("Emp2: %+v\n", emp2)

    emp3 := employee{
        name:   "Sam",
        age:    31,
        salary: 2000,
    }
    fmt.Printf("Emp3: %+v\n", emp3)

    emp4 := employee{
        name: "Sam",
        age:  31,
    }
    fmt.Printf("Emp4: %+v\n", emp4)
}

Output

Emp1: {name: age:0 salary:0}
Emp2: {name:Sam age:31 salary:2000}
Emp3: {name:Sam age:31 salary:2000}
Emp4: {name:Sam age:31 salary:0}

For above program

We first declare an employee struct.

emp1’s fields are all initialized with default zero value of its type i.e name with “”, age and salary with 0.

emp2 has been initialized with all fields on the same line. Its fields are correctly printed with their value

emp3’s has been initialized with all fields on different lines. Its fields are correctly printed with their value

emp4’s salary field is initialized with default zero value of 0. While other other two fields are correctly printed with their value.

It is to be noted that in the initialization of a struct, every new line with in curly braces has to end with a comma. So below initialization will raise error as

"salary" : 2000

doesn’t end with a comma.

emp := employee{
  name:   "Sam",
  age:    31,
  salary: 2000
}

This will be fine

emp := employee{
  name:   "Sam",
  age:    31,
  salary: 2000}

Without field names

struct can also be initialized without specifying the field names. But in this case, all values for each of the field has to be provided in sequence

emp := employee{"Sam", 31, 2000}

A compiler error will be raised if all values are not provided when field name is not used.

Let’s see a program

package main

import "fmt"

type employee struct {
    name   string
    age    int
    salary int
}

func main() {
    emp := employee{"Sam", 31, 2000}
    fmt.Printf("Emp: %+v\n", emp)

    //emp = employee{"Sam", 31}
}

Output

Emp2: {name:Sam age:31 salary:2000}

Uncomment the line

emp = employee{"Sam", 31}

in the above program, and it will raise compiler error

too few values in employee literal

Accessing and Setting Struct Fields

Structs fields can be accessed using the dot operator. Below is the format for getting the value

n := emp.name

Similarly a value can be assigned to a struct field too.

emp.name = "some_new_name"

package main

import "fmt"

type employee struct {
    name   string
    age    int
    salary int
}

func main() {
    emp := employee{name: "Sam", age: 31, salary: 2000}

    //Accessing a struct field
    n := emp.name
    fmt.Printf("Current name is: %s\n", n)

    //Assigning a new value
    emp.name = "John"
    fmt.Printf("New name is: %s\n", emp.name)
}

Output

Current name is: Sam
New name is: John

Pointer to a struct

There are two ways of creating a pointer to the struct

Using the & operator

Using the new keyword

Let’s looks at each of above method one by one.

Using the & operator

The & operator can be used to get the pointer to a struct variable.

emp := employee{name: "Sam", age: 31, salary: 2000}
empP := &emp

struct pointer can also be directly created as well

empP := &employee{name: "Sam", age: 31, salary: 2000}

Let’s look at a program

package main

import "fmt"

type employee struct {
    name   string
    age    int
    salary int
}

func main() {
    emp := employee{name: "Sam", age: 31, salary: 2000}
    empP := &emp
    fmt.Printf("Emp: %+v\n", empP)
    empP = &employee{name: "John", age: 30, salary: 3000}
    fmt.Printf("Emp: %+v\n", empP)
}

Output

Emp: &{name:Sam age:31 salary:2000}
Emp: &{name:John age:30 salary:3000}

Using the new keyword

Using the new() keyword will:

Create the struct

Initialize all the field to the zero default value of their type

Return the pointer to the newly created struct

This will return a pointer

empP := new(employee)

Pointer address can be print using the %p format modifier

fmt.Printf("Emp Pointer: %p\n", empP)

Deference operator ‘*’ can be used to print the value at the pointer.

fmt.Printf("Emp Value: %+v\n", *empP)

It will print

Emp Value: {name: age:0 salary:0}

When not using the dereference pointer but using the format identifier %+v, then ampersand will be appended before the struct indicating that is a pointer.

fmt.Printf("Emp Value: %+v\n", empP)

It will print

Emp Value: &{name: age:0 salary:0}

Let’s see full program denoting above points

package main

import "fmt"

type employee struct {
    name   string
    age    int
    salary int
}

func main() {
    empP := new(employee)
    fmt.Printf("Emp Pointer Address: %p\n", empP)
    fmt.Printf("Emp Pointer: %+v\n", empP)
    fmt.Printf("Emp Value: %+v\n", *empP)
}

Output

Emp Pointer Address: 0xc000130000
Emp Pointer: &{name: age:0 salary:0}
Emp Value: {name: age:0 salary:0}

Print a Struct Variable

There are two ways to print all struct variables including all its key and values.

Using the fmt package

Printing the struct in JSON form using the json/encoding package. This also allows pretty print of a struct as well.

Let’s see the two ways in which we can print the instance of the employee struct.

Using the fmt package

fmt.Printf() function can be used to print a struct. Different format identifiers can be used to print a struct in different ways. Let’s see how different format identifiers can be used to print a struct in different formats.

Let’s first create an instance of employee

emp := employee{name: "Sam", age: 31, salary: 2000}

%v – It will print only values. Field name will not be printed. This is the default way of printing a struct. Eg

fmt.Printf("%v", emp)  -  {Sam 31 2000}

%+v – It will print both field and value. Eg

fmt.Printf("%+v", emp) - {name:Sam age:31 salary:2000}

fmt.Println() function can also be used to print a struct. Since %v is the default for fmt.Printlin() function, hence output will be same as using %v for fmt.Printf()

fmt.Println(emp) - {Sam 31 2000}

Let’s see a working program too

package main

import "fmt"

type employee struct {
    name   string
    age    int
    salary int
}

func main() {
    emp := employee{name: "Sam", age: 31, salary: 2000}
    fmt.Printf("Emp: %v\n", emp)
    fmt.Printf("Emp: %+v\n", emp)
    fmt.Printf("Emp: %#v\n", emp)
    fmt.Println(emp)
}

Output

Emp: {Sam 31 2000}
Emp: {name:Sam age:31 salary:2000}
Emp: main.employee{name:"Sam", age:31, salary:2000}
{Sam 31 2000}

Printing the struct in JSON form

Second method is to print the struct in the JSON format. Marshal and MarshalIndent function of encoding/json package can be used to print a struct in JSON format. Here is the difference

Marshal – Below is the signature of the Marshal function. This function returns the JSON encoding of v by traversing the value recursively

Marshal(v interface{}) ([]byte, error)

MarshalIndent– Below is the signature of the MarshalIndent function. It is similar to Marshal function but applies Indent to format the output. So it can be used to pretty print a struct

MarshalIndent(v interface{}, prefix, indent string) ([]byte, error)

It is to be noted that both Marshal and MarshalIndent function can only access the exported fields of a struct, which means that only the capitalized fields can be accessed and encoded in JSON form.

package main

import (
    "encoding/json"
    "fmt"
    "log"
)

type employee struct {
    Name   string
    Age    int
    salary int
}

func main() {
    emp := employee{Name: "Sam", Age: 31, salary: 2000}
    //Marshal
    empJSON, err := json.Marshal(emp)
    if err != nil {
        log.Fatalf(err.Error())
    }
    fmt.Printf("Marshal funnction output %s\n", string(empJSON))

    //MarshalIndent
    empJSON, err = json.MarshalIndent(emp, "", "  ")
    if err != nil {
        log.Fatalf(err.Error())
    }
    fmt.Printf("MarshalIndent funnction output %s\n", string(empJSON))
}

Output:

Marshal funnction output {"Name":"Sam","Age":31}

MarshalIndent funnction output {
  "Name": "Sam",
  "Age": 31
}

The salary field is not printed in the output because it begins with a lowercase letter and is not exported. The Marshal function output is not formatted while the MarshalIndent function output is formatted.

golang also allows the JSON encoded struct key name to be different by the use of struct meta fields as will see in the next section.

Struct Field Meta or Tags

A struct in go also allows adding metadata to its fields. These meta fields can be used to encode decode into different forms, doing some forms of validations on struct fields, etc. So basically any meta information can be stored with fields of a struct and can be used by any package or library for different purposes.

Below is the format for attaching a meta-data. Meta-data is a string literal i.e it is enclosed in backquotes

type strutName struct{
   fieldName type `key:value key2:value2`
}

Now for our use case, we will add JSON tags to employee struct as below. Marshal function will use the key name specified in the tags

type employee struct {
    Name   string `json:"n"`
    Age    int    `json:"a"`
    Salary int    `json:"s"`
}

Let’s see full program

package main

import (
    "encoding/json"
    "fmt"
    "log"
)

type employee struct {
    Name   string `json:"n"`
    Age    int    `json:"a"`
    Salary int    `json:"s"`
}

func main() {
    emp := employee{Name: "Sam", Age: 31, Salary: 2000}
    //Converting to jsonn
    empJSON, err := json.MarshalIndent(emp, '', '  ')
    if err != nil {
        log.Fatalf(err.Error())
    }
    fmt.Println(string(empJSON))
}

Output:

{
  "n": "Sam",
  "a": 31,
  "s": 2000
}

The key name in the output is same as specified in the json meta tags.

Anonymous Fields in a Struct

A struct can have anonymous fields as well, meaning a field having no name. The type will become the field name. In below example, string will be the field name as well

type employee struct {
    string
    age    int
    salary int
}

The anonymous field can also be accessed and assigned a value

package main

import "fmt"

type employee struct {
    string
    age    int
    salary int
}

func main() {
    emp := employee{string: "Sam", age: 31, salary: 2000}
    //Accessing a struct field
    n := emp.string
    fmt.Printf("Current name is: %s\n", n)
    //Assigning a new value
    emp.string = "John"
    fmt.Printf("New name is: %s\n", emp.string)
}

Output

Current name is: Sam
New name is: John

Nested Struct

A struct can have another struct nested in it. Let’s see an example of a nested struct. In below employee struct has address struct nested it in.

package main

import "fmt"

type employee struct {
    name    string
    age     int
    salary  int
    address address
}

type address struct {
    city    string
    country string
}

func main() {
    address := address{city: "London", country: "UK"}
    emp := employee{name: "Sam", age: 31, salary: 2000, address: address}
    fmt.Printf("City: %s\n", emp.address.city)
    fmt.Printf("Country: %s\n", emp.address.country)
}

Output

City: London
Country: UK

Notice how nested struct fields are accessed.

emp.address.city
emp.address.country

Anonymous nested struct fields

The nested struct field can also be anonymous. Also, in this case, nested struct’s fields are directly accessed. So below is valid

emp.city
emp.country

It is also to be noted that below is still valid in this case

emp.address.city
emp.address.country

Let’s see a program

package main

import "fmt"

type employee struct {
	name   string
	age    int
	salary int
	address
}

type address struct {
	city    string
	country string
}

func main() {
	address := address{city: "London", country: "UK"}

	emp := employee{name: "Sam", age: 31, salary: 2000, address: address}

	fmt.Printf("City: %s\n", emp.address.city)
	fmt.Printf("Country: %s\n", emp.address.country)

	fmt.Printf("City: %s\n", emp.city)
	fmt.Printf("Country: %s\n", emp.country)
}

Output

City: London
Country: UK
City: London
Country: UK

Notice in above program that city field of address struct can be accessed in two ways

emp.city
emp.address.city

Similar for the country field of the address struct.

Exported and UnExported fields of a struct

Go doesn’t have any public, private or protected keyword. The only mechanism to control the visibility outside the package is using the capitalized and non-capitalized formats

Capitalized Identifiers are exported. The capital letter indicates that this is an exported identifier and is available outside the package.

Non-capitalized identifiers are not exported. The lowercase indicates that the identifier is not exported and will only be accessed from within the same package.

So any struct which starts with a capital letter is exported to other packages. Similarly any struct field which starts with capital is exported otherwise not. Let’s see an example that shows exporting and non-exporting of structs and struct fields. See model.go and test.go below. Both belong to the main package.

Structure
- Struct Person is exported
- Struct company is non-exported

Structure’s Field
- Person struct field Name is exported
- Person struct field age is not exported but Name is exported

model.go

package main

import "fmt"

//Person struct
type Person struct {
    Name string
    age  int
}

type company struct {
}

Let’s write a file test.go in same main package. See below.

test.go

package main

import "fmt"

//Test function
func Test() {
    //STRUCTURE IDENTIFIER
    p := &Person{
        Name: "test",
        age:  21,
    }
    fmt.Println(p)
    c := &company{}
    fmt.Println(c)
    
    //STRUCTURE'S FIELDS
    fmt.Println(p.Name)
    fmt.Println(p.age)
}

On running this file, it is able to access all exported and un-exported fields in model.go as both lies in the same package main. There is no compilation error and it gives below output

Output:

&{test 21}
&{}
test
21

Let’s move the above file model.go to a different package named model. Now notice the output on running ‘go build’. It gives compilation errors. All the compilation error are because test.go in main package to not able to refer to un-exported fields of model.go in model package

model.go

package model

//Person struct
type Person struct {
	Name string
	age  int
}

type company struct {
}

test.go

package main

import (
	"fmt"
        //This will path of your model package
	"/model"
)

//Test function
func main() {
	//STRUCTURE IDENTIFIER
	p := &model.Person{
		Name: "test",
		age:  21,
	}
	fmt.Println(p)
	c := &model.company{}
	fmt.Println(c)

	//STRUCTURE'S FIELDS
	fmt.Println(p.Name)
	fmt.Println(p.age)
}

Output:

cannot refer to unexported name model.company
p.age undefined (cannot refer to unexported field or method age)

Struct Equality

The first thing to know before considering struct equality is weather if all struct fields types are comparable or not

Some of the comparable types as defined by go specification are

boolean
numeric
string,
pointer
channel
interface types
structs – if all it’s field type is comparable
array – if the type of value of array element is comparable

Some of the types which are not comparable as per go specification and which cannot be used as a key in a map are.

Slice
Map
Function

So two struct will be equal if first all their field types are comparable and all the corresponding field values are equal. Let’s see an example

package main

import "fmt"

type employee struct {
    name   string
    age    int
    salary int
}

func main() {
    emp1 := employee{name: "Sam", age: 31, salary: 2000}
    emp2 := employee{name: "Sam", age: 31, salary: 2000}
    if emp1 == emp2 {
        fmt.Println("emp1 annd emp2 are equal")
    } else {
        fmt.Println("emp1 annd emp2 are not equal")
    }
}

Output

emp1 annd emp2 are equal

If the struct field type are not comparable then there will be compilation error on checking struct equality using the == operator.

package main
import "fmt"
type employee struct {
    name        string
    age         int
    salary      int
    departments []string
}
func main() {
    emp1 := employee{name: "Sam", age: 31, salary: 2000, departments: []string{"CS"}}
    emp2 := employee{name: "Sam", age: 31, salary: 2000, departments: []string{"EC"}}
    if emp1 == emp2 {
        fmt.Println("emp1 annd emp2 are equal")
    } else {
        fmt.Println("emp1 annd emp2 are not equal")
    }
}

Above program will raise compilation error as employee struct contains a field deparments which is a slice of string. slice is not a comparable type and hence the compilation error.

invalid operation: emp1 == emp2 (struct containing []string cannot be compared)

Struct are value types

A struct is value type in go. So a struct variable name is not a pointer to the struct in fact it denotes the entire struct. A new copy of the struct will be created when

A struct variable is assigned to another struct variable.

A struct variable is passed as an argument to a function.

Let’s see above point with another example

package main

import "fmt"

type employee struct {
    name   string
    age    int
    salary int
}

func main() {
    emp1 := employee{name: "Sam", age: 31, salary: 2000}
    fmt.Printf("Emp1 Before: %v\n", emp1)

    emp2 := emp1

    emp2.name = "John"
    fmt.Printf("Emp1 After assignment: %v\n", emp1)
    fmt.Printf("Emp2: %v\n", emp2)

    test(emp1)
    fmt.Printf("Emp1 After Test Function Call: %v\n", emp1)
}

func test(emp employee) {
    emp.name = "Mike"
    fmt.Printf("Emp in Test function: %v\n", emp)
}

Output

Emp1 Before: {Sam 31 2000}
Emp1 After assignment: {Sam 31 2000}
Emp2: {John 31 2000}
Emp in Test function: {Mike 31 2000}
Emp1 After Test Function Call: {Sam 31 2000}

In above example,

we assigned the emp1 to emp2 and we then changed name emp2 to have a different value. After that when we print emp1, we see that it hasn’t changed. This is because when we assign emp1 to emp2, a copy is created and changing emp2 doesn’t have any effect on emp1

We passed emp1 to the test function and then again changed its name field in the test function. After that when we print emp1, we see that it hasn’t changed. The reason is same, when emp1 is passed as an argument to test function a copy of emp1 is created.

Conclusion

This is all about struct in golang. In this article, we learned different ways of initializing a struct, pointer to struct, different ways of printing, about anonymous fields, etc. I hope you have liked this article. Please share the feedback/improvements/mistakes in the comments.

Next Tutorial – Array
Previous Tutorial – Pointer