This is an implementation of Red-Black tree written by Golang which does not support duplicate keys
.
With a healthy Go language installed, simply run go get github.com/HuKeping/rbtree
All you have to do is to implement a comparison function Less() bool
for your Item
which will be store in the Red-Black tree, here are some examples.
int
items.package main
import (
"fmt"
"github.com/HuKeping/rbtree"
)
func main() {
rbt := rbtree.New()
m := 0
n := 10
for m < n {
rbt.Insert(rbtree.Int(m))
m++
}
m = 0
for m < n {
if m%2 == 0 {
rbt.Delete(rbtree.Int(m))
}
m++
}
// 1, 3, 5, 7, 9 were expected.
rbt.Ascend(rbt.Min(), Print)
}
func Print(item rbtree.Item) bool {
i, ok := item.(rbtree.Int)
if !ok {
return false
}
fmt.Println(i)
return true
}
string
items.package main
import (
"fmt"
"github.com/HuKeping/rbtree"
)
func main() {
rbt := rbtree.New()
rbt.Insert(rbtree.String("Hello"))
rbt.Insert(rbtree.String("World"))
rbt.Ascend(rbt.Min(), Print)
}
func Print(item rbtree.Item) bool {
i, ok := item.(rbtree.String)
if !ok {
return false
}
fmt.Println(i)
return true
}
struct
items.package main
import (
"fmt"
"github.com/HuKeping/rbtree"
"time"
)
type Var struct {
Expiry time.Time `json:"expiry,omitempty"`
ID string `json:"id",omitempty`
}
// We will order the node by `Time`
func (x Var) Less(than rbtree.Item) bool {
return x.Expiry.Before(than.(Var).Expiry)
}
func main() {
rbt := rbtree.New()
var1 := Var{
Expiry: time.Now().Add(time.Second * 10),
ID: "var1",
}
var2 := Var{
Expiry: time.Now().Add(time.Second * 20),
ID: "var2",
}
var3 := Var{
Expiry: var2.Expiry,
ID: "var2-dup",
}
var4 := Var{
Expiry: time.Now().Add(time.Second * 40),
ID: "var4",
}
var5 := Var{
Expiry: time.Now().Add(time.Second * 50),
ID: "var5",
}
rbt.Insert(var1)
rbt.Insert(var2)
rbt.Insert(var3)
rbt.Insert(var4)
rbt.Insert(var5)
tmp := Var{
Expiry: var4.Expiry,
ID: "This field is not the key factor",
}
// var4 and var5 were expected
rbt.Ascend(rbt.Get(tmp), Print)
}
func Print(item rbtree.Item) bool {
i, ok := item.(Var)
if !ok {
return false
}
fmt.Printf("%+v\n", i)
return true
}