- Basic Syntax
- Operators
- Declarations
- Functions
- Built-in Types
- Type Conversions
- Packages
- Control structures
- Arrays, Slices, Ranges
- Maps
- Structs
- Pointers
- Interfaces
- Embedding
- Errors
- Concurrency
- Printing
- Reflection
- Snippets
- Time
- Best Practice
Most example code taken from A Tour of Go, which is an excellent introduction to Go. If you're new to Go, do that tour. Seriously.
- Imperative language
- Statically typed
- Syntax tokens similar to C (but less parentheses and no semicolons) and the structure to Oberon-2
- Compiles to native code (no JVM)
- No classes, but structs with methods
- Interfaces
- No implementation inheritance. There's type embedding, though.
- Functions are first class citizens
- Functions can return multiple values
- Has closures
- Pointers, but not pointer arithmetic
- Built-in concurrency primitives: Goroutines and Channels
File hello.go:
package main
import "fmt"
func main() {
fmt.Println("Hello Go")
}$ go run hello.go
| Operator | Description |
|---|---|
+ |
addition |
- |
subtraction |
* |
multiplication |
/ |
quotient |
% |
remainder |
& |
bitwise and |
| |
bitwise or |
^ |
bitwise xor |
&^ |
bit clear (and not) |
<< |
left shift |
>> |
right shift |
| Operator | Description |
|---|---|
== |
equal |
!= |
not equal |
< |
less than |
<= |
less than or equal |
> |
greater than |
>= |
greater than or equal |
| Operator | Description |
|---|---|
&& |
logical and |
|| |
logical or |
! |
logical not |
| Operator | Description |
|---|---|
& |
address of / create pointer |
* |
dereference pointer |
<- |
send / receive operator (see 'Channels' below) |
Type goes after identifier!
var foo int // declaration without initialization
var foo int = 42 // declaration with initialization
var foo, bar int = 42, 1302 // declare and init multiple vars at once
var foo = 42 // type omitted, will be inferred
foo := 42 // shorthand, only in func bodies, omit var keyword, type is always implicit
const constant = "This is a constant"
// iota can be used for incrementing numbers, starting from 0
const (
_ = iota
a
b
c = 1 << iota
d
)
fmt.Println(a, b) // 1 2 (0 is skipped)
fmt.Println(c, d) // 8 16 (2^3, 2^4)// a simple function
func functionName() {}
// function with parameters (again, types go after identifiers)
func functionName(param1 string, param2 int) {}
// multiple parameters of the same type
func functionName(param1, param2 int) {}
// return type declaration
func functionName() int {
return 42
}
// Can return multiple values at once
func returnMulti() (int, string) {
return 42, "foobar"
}
var x, str = returnMulti()
// Return multiple named results simply by return
func returnMulti2() (n int, s string) {
n = 42
s = "foobar"
// n and s will be returned
return
}
var x, str = returnMulti2()func main() {
// assign a function to a name
add := func(a, b int) int {
return a + b
}
// use the name to call the function
fmt.Println(add(3, 4))
}
// Closures, lexically scoped: Functions can access values that were
// in scope when defining the function
func scope() func() int{
outer_var := 2
foo := func() int { return outer_var}
return foo
}
func another_scope() func() int{
// won't compile because outer_var and foo not defined in this scope
outer_var = 444
return foo
}
// Closures
func outer() (func() int, int) {
outer_var := 2
inner := func() int {
outer_var += 99 // outer_var from outer scope is mutated.
return outer_var
}
inner()
return inner, outer_var // return inner func and mutated outer_var 101
}func main() {
fmt.Println(adder(1, 2, 3)) // 6
fmt.Println(adder(9, 9)) // 18
nums := []int{10, 20, 30}
fmt.Println(adder(nums...)) // 60
}
// By using ... before the type name of the last parameter you can indicate that it takes zero or more of those parameters.
// The function is invoked like any other function except we can pass as many arguments as we want.
func adder(args ...int) int {
total := 0
for _, v := range args { // Iterates over the arguments whatever the number.
total += v
}
return total
}bool
string
int int8 int16 int32 int64
uint uint8 uint16 uint32 uint64 uintptr
byte // alias for uint8
rune // alias for int32 ~= a character (Unicode code point) - very Viking
float32 float64
complex64 complex128All Go's predeclared identifiers are defined in the builtin package.
var i int = 42
var f float64 = float64(i)
var u uint = uint(f)
// alternative syntax
i := 42
f := float64(i)
u := uint(f)- Package declaration at top of every source file
- Executables are in package
main - Convention: package name == last name of import path (import path
math/rand=> packagerand) - Upper case identifier: exported (visible from other packages)
- Lower case identifier: private (not visible from other packages)
func main() {
// Basic one
if x > 10 {
return x
} else if x == 10 {
return 10
} else {
return -x
}
// You can put one statement before the condition
if a := b + c; a < 42 {
return a
} else {
return a - 42
}
// Type assertion inside if
var val interface{} = "foo"
if str, ok := val.(string); ok {
fmt.Println(str)
}
} // There's only `for`, no `while`, no `until`
for i := 1; i < 10; i++ {
}
for ; i < 10; { // while - loop
}
for i < 10 { // you can omit semicolons if there is only a condition
}
for { // you can omit the condition ~ while (true)
}
for index, value := range pow {
}
// use break/continue on current loop
// use break/continue with label on outer loop
here:
for i := 0; i < 2; i++ {
for j := i + 1; j < 3; j++ {
if i == 0 {
continue here
}
fmt.Println(j)
if j == 2 {
break
}
}
}
there:
for i := 0; i < 2; i++ {
for j := i + 1; j < 3; j++ {
if j == 1 {
continue
}
fmt.Println(j)
if j == 2 {
break there
}
}
} // switch statement
switch operatingSystem {
case "darwin":
fmt.Println("Mac OS Hipster")
// cases break automatically, no fallthrough by default
case "linux":
fmt.Println("Linux Geek")
default:
// Windows, BSD, ...
fmt.Println("Other")
}
// as with for and if, you can have an assignment statement before the switch value
switch os := runtime.GOOS; os {
case "darwin": ...
}
// you can also make comparisons in switch cases
number := 42
switch {
case number < 42:
fmt.Println("Smaller")
case number == 42:
fmt.Println("Equal")
case number > 42:
fmt.Println("Greater")
}
// cases can be presented in comma-separated lists
var char byte = '?'
switch char {
case ' ', '?', '&', '=', '#', '+', '%':
fmt.Println("Should escape")
}var a [10]int // declare an int array with length 10. Array length is part of the type!
a[3] = 42 // set elements
i := a[3] // read elements
// declare and initialize
var a = [2]int{1, 2}
a := [2]int{1, 2} //shorthand
a := [...]int{1, 2} // elipsis -> Compiler figures out array lengthvar a []int // declare a slice - similar to an array, but length is unspecified
var a = []int {1, 2, 3, 4} // declare and initialize a slice (backed by the array given implicitly)
a := []int{1, 2, 3, 4} // shorthand
chars := []string{0:"a", 2:"c", 1: "b"} // ["a", "b", "c"]
var a [][]str // This declares a slice of slices of strings. It's a two-dimensional slice where each element is a slice of strings.
a := make([][]string, len(anotherSlices)) // Initialize a 2D slice with the size of the dataSlice
var b = a[lo:hi] // creates a slice (view of the array) from index lo to hi-1
var b = a[1:4] // slice from index 1 to 3
var b = a[:3] // missing low index implies 0
var b = a[3:] // missing high index implies len(a)
// both a and b point to the same array. when change b, a is changed corresponding.
a = append(a, 17, 3) // append items to slice a
c := append(a, b...) // concatenate slices a and b
// append clone a to new array which is pointed by a new result slice
// create a slice with make
a = make([]byte, 5, 5) // first arg length, second capacity
a = make([]byte, 5) // capacity is optional
// create a slice from an array
x := [3]string{"Лайка", "Белка", "Стрелка"}
s := x[:] // a slice referencing the storage of xlen(a) gives you the length of an array/a slice. It's a built-in function, not a attribute/method on the array.
// loop over an array/a slice
for i, e := range a {
// i is the index, e the element
}
// if you only need e:
for _, e := range a {
// e is the element
}
// ...and if you only need the index
for i := range a {
}
// In Go pre-1.4, you'll get a compiler error if you're not using i and e.
// Go 1.4 introduced a variable-free form, so that you can do this
for range time.Tick(time.Second) {
// do it once a sec
}m := make(map[string]int)
m["key"] = 42
fmt.Println(m["key"])
delete(m, "key")
elem, ok := m["key"] // test if key "key" is present and retrieve it, if so
// map literal
var m = map[string]Vertex{
"Bell Labs": {40.68433, -74.39967},
"Google": {37.42202, -122.08408},
}
// map complicate: key is a slice of 26 digits, value is a slice of strings
m := make(map[[26]int][]string)
// iterate over map content
for key, value := range m {
}There are no classes, only structs. Structs can have methods.
// A struct is a type. It's also a collection of fields
// Declaration
type Vertex struct {
X, Y float64
}
// Creating
var v = Vertex{1, 2}
var v = Vertex{X: 1, Y: 2} // Creates a struct by defining values with keys
var v = []Vertex{{1,2},{5,2},{5,5}} // Initialize a slice of structs
// Accessing members
v.X = 4
// You can declare methods on structs. The struct you want to declare the
// method on (the receiving type) comes between the the func keyword and
// the method name. The struct is copied on each method call(!)
func (v Vertex) Abs() float64 {
return math.Sqrt(v.X*v.X + v.Y*v.Y)
}
// Call method
v.Abs()
// For mutating methods, you need to use a pointer (see below) to the Struct
// as the type. With this, the struct value is not copied for the method call.
func (v *Vertex) add(n float64) {
v.X += n
v.Y += n
}Anonymous structs:
Cheaper and safer than using map[string]interface{}.
point := struct {
X, Y int
}{1, 2}p := Vertex{1, 2} // p is a Vertex
q := &p // q is a pointer to a Vertex
r := &Vertex{1, 2} // r is also a pointer to a Vertex
// The type of a pointer to a Vertex is *Vertex
var s *Vertex = new(Vertex) // new creates a pointer to a new struct instance// interface declaration
type Animal interface {
Move() string
}
// instead, types implicitly satisfy an interface if they implement all required methods
type horse struct {}
func NewHorse() Animal {
return &Horse{}
}
// implement all required methods
func (h *horse) Move() string {
return "run"
}
// using
var pet Animal = NewHorse()
pet.Move()
[reference](https://github.com/uber-go/guide/blob/master/style.md)
// verify Interface Compliance
// The statement var _ Provider = (*GgOauthHandler)(nil) will fail to compile if *GgOauthHandler ever stops matching the Provider interface.
type GgOauthHandler struct {
}
var _ Provider = (*GgOauthHandler)(nil)There is no subclassing in Go. Instead, there is interface and struct embedding.
// ReadWriter implementations must satisfy both Reader and Writer
type ReadWriter interface {
Reader
Writer
}
// Server exposes all the methods that Logger has
type Server struct {
Host string
Port int
*log.Logger
}
// initialize the embedded type the usual way
server := &Server{"localhost", 80, log.New(...)}
// methods implemented on the embedded struct are passed through
server.Log(...) // calls server.Logger.Log(...)
// the field name of the embedded type is its type name (in this case Logger)
var logger *log.Logger = server.LoggerThere is no exception handling. Instead, functions that might produce an error just declare an additional return value of type error. This is the error interface:
// The error built-in interface type is the conventional interface for representing an error condition,
// with the nil value representing no error.
type error interface {
Error() string
}Here's an example:
func sqrt(x float64) (float64, error) {
if x < 0 {
return 0, errors.New("negative value")
}
return math.Sqrt(x), nil
}
func main() {
val, err := sqrt(-1)
if err != nil {
// handle error
fmt.Println(err) // negative value
return
}
// All is good, use `val`.
fmt.Println(val)
}Goroutines are lightweight threads (managed by Go, not OS threads). go f(a, b) starts a new goroutine which runs f (given f is a function).
// just a function (which can be later started as a goroutine)
func doStuff(s string) {
}
func main() {
// using a named function in a goroutine
go doStuff("foobar")
// using an anonymous inner function in a goroutine
go func (x int) {
// function body goes here
}(42)
}ch := make(chan int) // create a channel of type int
ch <- 42 // Send a value to the channel ch.
v := <-ch // Receive a value from ch
// Non-buffered channels block. Read blocks when no value is available, write blocks until there is a read.
// Create a buffered channel. Writing to a buffered channels does not block if less than <buffer size> unread values have been written.
ch := make(chan int, 100)
close(ch) // closes the channel (only sender should close)
// read from channel and test if it has been closed
v, ok := <-ch
// if ok is false, channel has been closed
// Read from channel until it is closed
for i := range ch {
fmt.Println(i)
}
// select blocks on multiple channel operations, if one unblocks, the corresponding case is executed
func doStuff(channelOut, channelIn chan int) {
select {
case channelOut <- 42:
fmt.Println("We could write to channelOut!")
case x := <- channelIn:
fmt.Println("We could read from channelIn")
case <-time.After(time.Second * 1):
fmt.Println("timeout")
}
}-
A send to a nil channel blocks forever
var c chan string c <- "Hello, World!" // fatal error: all goroutines are asleep - deadlock!
-
A receive from a nil channel blocks forever
var c chan string fmt.Println(<-c) // fatal error: all goroutines are asleep - deadlock!
-
A send to a closed channel panics
var c = make(chan string, 1) c <- "Hello, World!" close(c) c <- "Hello, Panic!" // panic: send on closed channel
-
A receive from a closed channel returns the zero value immediately
var c = make(chan int, 2) c <- 1 c <- 2 close(c) for i := 0; i < 3; i++ { fmt.Printf("%d ", <-c) } // 1 2 0
fmt.Println("Hello, 你好, नमस्ते, Привет, ᎣᏏᏲ") // basic print, plus newline
p := struct { X, Y int }{ 17, 2 }
fmt.Println( "My point:", p, "x coord=", p.X ) // print structs, ints, etc
s := fmt.Sprintln( "My point:", p, "x coord=", p.X ) // print to string variable
fmt.Printf("%d hex:%x bin:%b fp:%f sci:%e",17,17,17,17.0,17.0) // c-ish format
s2 := fmt.Sprintf( "%d %f", 17, 17.0 ) // formatted print to string variable
hellomsg := `
"Hello" in Chinese is 你好 ('Ni Hao')
"Hello" in Hindi is नमस्ते ('Namaste')
` // multi-line string literal, using back-tick at beginning and endA type switch is like a regular switch statement, but the cases in a type switch specify types (not values) which are compared against the type of the value held by the given interface value.
func do(i interface{}) {
switch v := i.(type) {
case int:
fmt.Printf("Twice %v is %v\n", v, v*2)
case string:
fmt.Printf("%q is %v bytes long\n", v, len(v))
default:
fmt.Printf("I don't know about type %T!\n", v)
}
}
func main() {
do(21)
do("hello")
do(true)
}Go programs can embed static files using the "embed" package as follows:
package main
import (
"embed"
"log"
"net/http"
)
// content holds the static content (2 files) for the web server.
//go:embed a.txt b.txt
var content embed.FS
func main() {
http.Handle("/", http.FileServer(http.FS(content)))
log.Fatal(http.ListenAndServe(":8080", nil))
}package main
import (
"fmt"
"net/http"
)
// define a type for the response
type Hello struct{}
// let that type implement the ServeHTTP method (defined in interface http.Handler)
func (h Hello) ServeHTTP(w http.ResponseWriter, r *http.Request) {
fmt.Fprint(w, "Hello!")
}
func main() {
var h Hello
http.ListenAndServe("localhost:4000", h)
}
// Here's the method signature of http.ServeHTTP:
// type Handler interface {
// ServeHTTP(w http.ResponseWriter, r *http.Request)
// }Both of these are valid RFC3339 times:
"2015-09-15T14:00:12-00:00"
"2015-09-15T14:00:13Z"If you can (e.g. a non-shared resource that does not need to be passed as reference), use a value. By the following reasons:
- Your code will be nicer and more readable, avoiding pointer operators and null checks.
- Your code will be safer against Null Pointer panics.
- Your code will be often faster: yes, faster! Why?
Reason 1: you will allocate less items in the heap. Allocating/deallocating from stack is immediate, but allocating/deallocating on Heap may be very expensive (allocation time + garbage collection). You can see some basic numbers here: http://www.macias.info/entry/201802102230_go_values_vs_references.md
Reason 2: especially if you store returned values in slices, your memory objects will be more compacted in memory: looping a slice where all the items are contiguous is much faster than iterating a slice where all the items are pointers to other parts of the memory. Not for the indirection step but for the increase of cache misses.
Myth breaker: a typical x86 cache line are 64 bytes. Most structs are smaller than that. The time of copying a cache line in memory is similar to copying a pointer.
Only if a critical part of your code is slow I would try some micro-optimization and check if using pointers improves somewhat the speed, at the cost of less readability and mantainability.