Skip to content

Latest commit

 

History

History

03-Structs

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
src
src
 
 
go.mod
go.mod
 
 
readme.md
readme.md
 
 

readme.md

Structs

Struct Basics

  • A Data Structure used widely throughout Go
  • A collection of different properties that are related
  • More flexible properties than just a single value type (e.g. string or number)
  • Similar to Python Dictionary, but it is not an Object
  • The type of each field inside the struct can be different
  • Can be used to represent complex data structures
  • We can embbed one struct inside another struct to build complex data structures
    • We can reuse structs in any way we want
    • This allows Composition of different data structures
    • When embedding another struct, we can also skip the fieldName and only use the struct type
      • This will have the effect of creating the fieldName the same name as the struct name
      • If the fieldName is different than the struct name, then it needs to be explicit
// Declaring new structs
type contactInfo struct {
    email   string
    zipCode int
}

type person struct {
    firstName string
    lastName  string
    // Embedded struct
    contact   contactInfo
}
  • Now, we can start making use of the struct in the main() function
  • The order of the fields can be used when defining a struct instance
    • However, this is not the recommended approach
func main() {
    // Creating a value of type person
    // Using field order: This is not the recommended approach
    alex := person{
        "Alex",        // firstName
        "Anderson",    // lastName
        // Embedded struct
        // When embedding another struct, we can also skip the fieldName
        // We can just use the struct type
        // This has the effect of creating the fieldName the same name as the struct name
        contactInfo{   // contactInfo
            "[email protected]", // email
            12345                // zipCode
        }
    }

    // Show the struct
    fmt.Println(alex)
}
  • Instead of relying on the order of fields, use named fields
    • This is the recommended approach
func main() {
    // Creating a value of type person
    // Using named fields: This is the recommended approach
    maria := person{
        firstName: "Maria",
        lastName:  "Anderson",
        // Embedded struct
        contact: contactInfo{
            email:   "[email protected]",
            zipCode: 98765
        }
    }

    // Show the struct
    fmt.Println(maria)
}
  • We can also declare then assign later
    • The struct's fields would be assigned their zero-value by default before a new value is assigned to it
Data Type Zero-Value
struct null
string ""
int 0
float64 0
bool false 
func main() {
    // Declare variable:
    // Fields default to null-value
    var julie person

    // To print a struct with key:value format, use %+v
    fmt.Printf("julie before assignment: %+v\n", julie)

    // Assigning / Re-assigning values
    julie.firstName = "Julie"
    julie.lastName = "Arkorius"
    julie.contact.email = "[email protected]"
    julie.contact.zipCode = 12345

    // Check again
    fmt.Printf("julie after assignment: %+v\n", julie)
}

Struct With Receiver Function

  • Similar to Types, we can define receiver functions to be attached to Structs as well
  • Note: The code below has some issues because of Pointers. We will review that below
// Receiver function for "person"
func (p person) print() {
    fmt.Printf("%+v\n", p)
}

// This function has some issues because of pointer
func (p person) updateFirstName(newFirstName string) {
    p.firstName = newFirstName
}

func main() {
    // Defining a new person variable
    jim := person{
        firstName: "Jim",
        lastName:  "Patterson",
        contact: contactInfo{
            email:    "[email protected]",
            zipCode: 98765,
        },
    }

    // Calling a receiver function
    jim.print()
    jim.updateFirstName("Jimmy")
    jim.print()

    // Why did it not update?
    // Because of how Pointers work
}

Pointers and Values

  • Previously, we saw that calling jim.updateFirstName() did not update jim.firstName field
  • Why did it not update the first name to "Jimmy"?
    • Because of Pointers in Go
    • A struct variable is a pointer type
    • Points to a value stored in memory
  • By default, Go is a pass-by-value language
    • With the way jim.updateFirstName() is defined, the value stored in jim is copied and stored in a new location in memory, which is then used by updateFirstName()
    • However, the original variable jim is still pointing to the old value
    • So, we are not updating the original struct but a copy of it in a different memory location
  • Posible Solution: We could re-assign back to the original variable
    • However, this is not preferred because it duplicates values in memory, thus inneficient
    • Also, if the struct is large, we have to move a large data 2 times in memory
        1. Original -> Copy
        1. Modify Copy
        1. Copy -> Original
  • Actual Solution: To solve this issue, we need to use Pointers to force Go to pass-by-reference and update the original

Using Pointer

  • We use the &varName to create a reference (memory address) to varName
// JimPtr contains the memory address pointing to jim
jimPtr := &jim
  • We use the pointer type to pass to the function that wants to update the original variable
// Before
// jim.updateFirstName("Jimmy")

// After
jimPtr.updateFirstName("Jimmy")
  • This requires the receiver function definition to change as well: It needs to take a Pointer-type as argument instead
Syntax Decription
varName *type A Pointer-type variable that points to a value of the type (i.e. contains the reference/memory address to the value of the type) 
// updateFirstNamePointer takes a Pointer-type that points to a value of type person
func (ptrPers *person) updateFirstNamePointer(newFirstName string) {
    ...
}
  • Within the receiver function definition, we de-reference the pointer (i.e. get back the value it is pointing at) using the *varName or (*varName) syntax
Syntax Decription
*varName or (*varName) Operator to access the value that exists at the memory address (pointer) 
// Using pointer as receiver allows us to pass-by-reference
func (ptrPers *person) updateFirstNamePointer(newFirstName string) {
    // De-reference the pointer type
    // This means updating the value at the reference/pointer
    (*ptrPers).firstName = newFirstName
}
  • However, Go can automatically de-reference a pointer as well
  • So the following also works
// Using pointer as receiver allows us to pass-by-reference
func (ptrPers *person) updateFirstNamePointer(newFirstName string) {
    // Go can automatically de-reference the pointer
    ptrPers.firstName = newFirstName
}
  • Now, in main(), we can make use of this receiver function
  • We use the & sign to create a reference
Syntax Decription
&varName or (&varName) Operator to get the reference (memory address) that points to variable 
func main() {
    jim := person{
        firstName: "Jim",
        lastName:  "Patterson",
        contact: contactInfo{
            email:    "[email protected]",
            zipCode: 98765,
        },
    }

    // Checking jim before calling a receiver function
    jim.print()

    // Updating jim: Passing By Reference
    (&jim).updateFirstNamePointer("Jimmy")

    // Checking jim after calling a receiver function
    jim.print()
}
  • However, with Go, it is possible to substitute a pointer with its root variable
  • So the following still work, even if the receiver requires a pointer type
// Updating jim
jim.updateFirstNamePointer("Big ol'Jim")
jim.print()
  • TLDR
    • Make sure to define the parameter(s) of the receiver function to be of Pointer-type varName *type
    • Optional: Within the receiver function, we de-reference the pointer with *varName or *(varName) to access/update the value at the Pointer-type's address
      • But Go can automatically de-reference, so using just varName also works
    • Optional: When calling the receiver function (in main), pass a memory reference to it using &varName or &(varName)
      • But Go can automatically reference, so using just varName also works

About Pointers

Syntax Decription
&varName or (&varName) Operator to get the reference (memory address) that points to variable's value: This is Referencing a Pointer
*varName or (*varName) Operator to access the value that exists at the memory address/pointer &varName: This is De-referencing a Pointer
varName The original variable that we asigned the value to: Turns into its equivalent memory address reference using &
*type A Pointer type that point to a memory address whose value is the type: &varName is of type *type 
var jon person = person{firstName: "Jon", lastName: "Leukippos"}
var ptrJon *person = &jon
var jonVal person = *ptrJon
var jonCopy person = jon
varName varType Memory Address (Reference) &varName Value (Dereference) *ptrVarName Pointer *type
jon person &jon same as ptrJon

0x1F4CB3		 *ptrJon same as jonVal

{firstName: "Jon", lastName: "Leukippos"}

Same value as jonCopy but different objects: Original vs Copy		 *person

Memory Address type referencing to a value of type person

Pointer Gotchas

  • Go is typically a Pass-By-Value language
  • However, Slices seems to by default passed-by-reference (Pointer) for better performance
    • NOTE: Arrays are passed-by-value though!!
  • In reality though, Go is always passed by value, but slices use a trick
  • When we create a slice, Go internally creates 2 separate data structure:
    • The Slice Data Structure
      • Pointer to the underlying array -> Points to a different address in memory
      • Capacity
      • Length
    • An underlying Array
      • Contains the actual list of items
  • When we modify the array, the Slice Data Structure gets moved around by value, but it is still pointing to the same underlying Array
    • The Slice Data Structure is passed around by value
    • But it always point to the underlying array
    • The Slice Data Structure gets modified (by value) when passed around
    • But the reference to the undelying array data structure remains
func main() {
    mySlice := []string{"Hi", "there", "how", "are", "you"}
    myArray := [4]string{"This", "is", "an", "Array"}

    updateSlice(mySlice)
    // Works because passed by reference
    // Updating on the argument = updating the same memory address

    updateArray(myArray)
    // Does not work because passed by value
    // Updating on the argument = updating a different copied value
    // Need to use pointer to update the same variable

    fmt.Println(mySlice)
    fmt.Println(myArray)
}

// Helper Functions
func updateSlice(s []string) {
    s[0] = "Bye"
}

func updateArray(arr [4]string) {
    arr[0] = "What"
}

List Of Data Structures That Are Passed BY REFERENCE (REFERENCE TYPES)

No need to use pointers with these

  • Slice
  • Map
  • Channel
  • Pointer
  • Function

List Of Data Structures That Are Passed BY VALUE (VALUES TYPES)

Use pointers to change the underlying values for these in functions

  • int
  • float
  • string
  • bool
  • struct
  • Array