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 similar to Java/C/C++, but less parantheses and no semicolons
- 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
- Go 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
- Type goes after identifier!
var foo int // declaration without initializationvar foo int = 42 // declaration with initializationvar foo, bar int = 42, 1302 // declare and init multiple vars at oncevar foo = 42 // type omitted, will be inferredfoo := 42 // shorthand, only in func bodies, omit var keyword, type is always implicitconst constant = "This is a constant"
// 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: Functions can access values that were in scope when defining the
// function
// adder returns an anonymous function with a closure containing the variable sum
func adder() func(int) int {
sum := 0
return func(x int) int {
sum += x // sum is declared outside, but still visible
return sum
}
}func main() {
fmt.Println(adder(1, 2, 3)) // return 6
fmt.Println(adder(9, 9)) // return 18
}
// 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 complex128
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)
if x > 0 {
return x
} else {
return -x
}
// You can put one statement before the condition
if a := b + c; a < 42 {
return a
} else {
return a - 42
} // 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)
} // 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": ...
}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
a := [2]int{1, 2}
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
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)
// 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
}var m map[string]int
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},
}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 int
}
// Creating
var v = Vertex{1, 2}
var v = Vertex{x: 1, y: 2} // Creates a struct by defining values with keys
// 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
}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 Awesomizer interface {
Awesomize() string
}
// types do *not* declare to implement interfaces
type Foo struct {}
// instead, types implicitly satisfy an interface if they implement all required methods
func (foo Foo) Awesomize() string {
return "Awesome!"
}TODO
There is no exception handling. Functions that might produce an error just declare an additional return value of type Error. This is the Error interface:
type error interface {
Error() string
}A function that might return an error:
func doStuff() (int, error) {
}
func main() {
result, error := doStuff()
if (error != nil) {
// handle error
} else {
// all is good, use result
}
}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 if a value already has been written but not 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(c) // 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")
}
}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)
// }
nice