开源一个 Go 泛型数据容器库

https://github.com/lxzan/dao

DAO 是一个基于泛型的数据结构与算法库, 旨在补充标准库在数据容器和算法方面的不足, 简化业务开发.

Features

简单介绍下一些特色数据容器

Vector

为简化 CRUD 而开发的容器, 使用带接口约束泛型, 部分弥补了 Go 不支持泛型方法的不足

package main import ( "fmt" "github.com/lxzan/dao/vector" "time" ) type User struct { ID int64 Name string CreatedAt time.Time UpdatedAt time.Time } func (u *User) GetID() int64 { return u.ID } func main() { var docs = vector.NewFromDocs[int64, *User]( &User{ID: 1, Name: "ming"}, &User{ID: 3, Name: "hong"}, &User{ID: 3, Name: "hong"}, &User{ID: 4, Name: "mei"}, &User{ID: 4, Name: "mei"}, ) docs. Unique(). Filter(func(i int, v *User) bool { return v.ID%2 == 1 }). Range(func(i int, v *User) bool { fmt.Printf("id=%d, name=%s\n", v.ID, v.Name) return true }) } 

IndexedHeap

N 叉索引堆, 在普通堆的基础上拓展了更新和删除方法.

package main import ( "github.com/lxzan/dao/heap" "github.com/lxzan/dao/types/cmp" ) func main() { var h = heap.NewIndexedHeap[int, string](heap.Quadratic, cmp.Less[int]) h.Push(1, "") h.Push(3, "") h.Push(5, "") var node = h.Push(2, "") h.Push(4, "") h.Push(6, "") h.DeleteByIndex(node.Index()) for h.Len() > 0 { println(h.Pop().Key()) } } 

Deque

高性能双端队列, 基于数组实现, 使用数据索引模拟指针. Deque 内部不包含任何指针, 删除的槽位内存可复用, 不依赖 sync.Pool.

package main import ( "fmt" "github.com/lxzan/dao/deque" ) func main() { var q = deque.Deque[int]{} q.PushBack(1) q.PushBack(3) q.PushBack(5) q.PushBack(7) q.PushBack(9) for i := q.Front(); i != nil; i = q.Get(i.Next()) { fmt.Printf("%d ", i.Value()) } fmt.Printf("\n") for i := q.Back(); i != nil; i = q.Get(i.Prev()) { fmt.Printf("%d ", i.Value()) } } 

红黑树

高性能红黑树实现, 从 nginx 移植而来.

package main import ( "github.com/lxzan/dao/rbtree" ) func main() { var tree = rbtree.New[int, struct{}]() for i := 0; i < 100; i++ { tree.Set(i, struct{}{}) } var results = tree. NewQuery(). Left(func(key int) bool { return key >= 45 }). Right(func(key int) bool { return key <= 65 }). Limit(5). Offset(5). Order(rbtree.DESC). FindAll() for _, item := range results { println(item.Key) } }