Golang语言开发的开源NoSQL数据库——TieDot简介
简介
项目开源地址:https://github.com/HouzuoGuo/tiedot
发起者留下了他的Twitter,貌似姓郭,是个美籍华人
项目简介中,有关于对性能的描述,有人用此数据库抓取了维基百科,保存5900万数据,共73G。
安装
配置好Go环境后,运行
go get github.com/HouzuoGuo/tiedot
入门
使用有2种方式,使用HTTP做接口,适用任何语言;使用嵌入式,使用Go语言,这里介绍使用Go语言,数据库的嵌入模式,足以应付百万请求/天了。
项目自带了演示代码,通过下面的命令执行
./tiedot -mode=example
性能评估命令
./tiedot -mode=bench # 对40万数据,进行增删改查 ./tiedot -mode=bench2 # 还没仔细看
项目中,example.go 文件是Go语言的使用示例
// It is very important to initialize random number generator seed! rand.Seed(time.Now().UTC().UnixNano()) // Create and open database创建并打开数据库 dir := "/tmp/MyDatabase" os.RemoveAll(dir) defer os.RemoveAll(dir) myDB, err := db.OpenDB(dir) if err != nil { panic(err) } // Create two collections Feeds and Votes创建2张表 // "2" means collection data and indexes are divided into two halves, allowing concurrent access from two threads if err := myDB.Create("Feeds", 2); err != nil { panic(err) } if err := myDB.Create("Votes", 2); err != nil { panic(err) } // What collections do I now have?查询都有哪些表 for name := range myDB.StrCol { fmt.Printf("I have a collection called %s\n", name) } // Rename collection "Votes" to "Points"把表"Votes"重命名为"Points" if err := myDB.Rename("Votes", "Points"); err != nil { panic(err) } // Drop (delete) collection "Points"删除表"Points" if err := myDB.Drop("Points"); err != nil { panic(err) } // Scrub (repair and compact) "Feeds"修复并压缩表"Feeds" myDB.Scrub("Feeds") // ****************** Document Management ****************** // Start using a collection使用表"Feeds" feeds := myDB.Use("Feeds") // Insert document (document must be map[string]interface{})插入数据 docID, err := feeds.Insert(map[string]interface{}{ "name": "Go 1.2 is released", "url": "golang.org"}) if err != nil { panic(err) } // Read document根据id查询数据 var readBack interface{} feeds.Read(docID, &readBack) // pass in document's physical ID fmt.Println(readBack) // Update document (document must be map[string]interface{})改数据 err = feeds.Update(docID, map[string]interface{}{ "name": "Go is very popular", "url": "google.com"}) if err != nil { panic(err) } // Delete document删除数据 feeds.Delete(docID) // Delete document feeds.Delete(123) // An ID which does not exist does no harm // ****************** Index Management ******************索引管理 // Secondary indexes assist in many types of queries // Create index (path leads to document JSON attribute)建索引 if err := feeds.Index([]string{"author", "name", "first_name"}); err != nil { panic(err) } if err := feeds.Index([]string{"Title"}); err != nil { panic(err) } if err := feeds.Index([]string{"Source"}); err != nil { panic(err) } // What indexes do I have on collection A?查询有哪些索引 for path := range feeds.SecIndexes { fmt.Printf("I have an index on path %s\n", path) } // Remove index删索引 if err := feeds.Unindex([]string{"author", "name", "first_name"}); err != nil { panic(err) } // ****************** Queries ******************查询表 // Let's prepare a number of docments for a start feeds.Insert(map[string]interface{}{"Title": "New Go release", "Source": "golang.org", "Age": 3}) feeds.Insert(map[string]interface{}{"Title": "Kitkat is here", "Source": "google.com", "Age": 2}) feeds.Insert(map[string]interface{}{"Title": "Good Slackware", "Source": "slackware.com", "Age": 1}) queryStr := `[{"eq": "New Go release", "in": ["Title"]}, {"eq": "slackware.com", "in": ["Source"]}]` var query interface{} json.Unmarshal([]byte(queryStr), &query)查询条件 queryResult := make(map[uint64]struct{}) // query result (document IDs) goes into map keys保存查询结果的变量 if err := db.EvalQuery(query, feeds, &queryResult); err != nil {执行查询 panic(err) } // Query results are physical document IDs打印查询结果 for id := range queryResult { fmt.Printf("Query returned document ID %d\n", id) } // To use the document itself, simply read it back for id := range queryResult { feeds.Read(id, &readBack) fmt.Printf("Query returned document %v\n", readBack) } // Gracefully close database关闭数据库 myDB.Close()
郑重声明:本站内容如果来自互联网及其他传播媒体,其版权均属原媒体及文章作者所有。转载目的在于传递更多信息及用于网络分享,并不代表本站赞同其观点和对其真实性负责,也不构成任何其他建议。