MongoDB的集合(collection)可以看做关系型数据库的表,文档对象(document)可以看做关系型数据库的一条记录。但两者并不完全对等。表的结构是固定的,MongoDB集合并没有这个约束;另外,存入集合的文档对象甚至可以嵌入子文档,或者“子集合”。他们最终都可以用类似于BJSON的格式描述。我们今天就来分析MongoDB这一特性带来的独特数据管理方式。我们还是以samus驱动为例来分析,samus驱动支持两种方式访问数据库,基本方式和linq方式,基本方式在上篇以介绍过,linq方式我不想单独讲解应用实例,这篇我会用两种方式来对比介绍。
一、包含子文档的集合操作
有这么一个应用场景,某网站提供会员登录的功能,用户需要注册账号才能享受会员服务,但是注册者可能会因为用户资料表单输入项过大而放弃填写,因此用户信息分为主要资料和详细资料两项,初次注册只需要填写主要资料就行了。我们打算把详细信息设计为子文档存储。
1) linq方式实现
1. 新建数据描述类,描述用户信息
- /// <summary>
- /// 用户主要资料
- /// </summary>
- public class UserInfo
- {
- public string UserId { get; set; }
- public string UserName { get; set; }
- public string PassWord { get; set; }
- public Detail Detail { get; set; }
- }
- /// <summary>
- /// 用户详细资料
- /// </summary>
- public class Detail
- {
- public string Address { get; set; }
- public int Age { get; set; }
- public string Email { get; set; }
- }
2. 我们要新建一个用户业务操作类“UserBLL”。这个时候要让驱动知道UserInfo类描述了“用户资料”的字段信息,在GetMongo()方法实现了配置步骤,UserBLL完整代码如下:
- public class UserBLL
- {
- public string connectionString = "mongodb://localhost";
- public string databaseName = "myDatabase";
- private Mongo mongo;
- private MongoDatabase mongoDatabase;
- //注意这里泛型类型为“UserInfo”
- private MongoCollection<UserInfo> mongoCollection;
- public UserBLL()
- {
- mongo = GetMongo();
- mongoDatabase = mongo.GetDatabase(databaseName) as MongoDatabase;
- mongoCollection = mongoDatabase.GetCollection<UserInfo>() as MongoCollection<UserInfo>;
- mongo.Connect();
- }
- ~UserBLL()
- {
- mongo.Disconnect();
- }
- /// <summary>
- /// 配置Mongo,将类UserInfo映射到集合
- /// </summary>
- private Mongo GetMongo()
- {
- var config = new MongoConfigurationBuilder();
- config.Mapping(mapping =>
- {
- mapping.DefaultProfile(profile =>
- {
- profile.SubClassesAre(t => t.IsSubclassOf(typeof(UserInfo)));
- });
- mapping.Map<UserInfo>();
- });
- config.ConnectionString(connectionString);
- return new Mongo(config.BuildConfiguration());
- }
- }
3. 接着,在“UserBLL”类中定义一个方法“InsertSomeData()”来插入一些数据:
- /// <summary>
- /// 插入一些数据
- /// </summary>
- public void InsertSomeData()
- {
- UserInfo userInfo1 = new UserInfo()
- {
- UserId = "1001",
- UserName = "张三",
- PassWord = "123456"
- };
- mongoCollection.Save(userInfo1);
- UserInfo userInfo2 = new UserInfo()
- {
- UserId = "1002",
- UserName = "李四",
- PassWord = "123456",
- Detail = new Detail()
- {
- Address = "湖北",
- Age = 20,
- Email = "lisi@163.com"
- }
- };
- mongoCollection.Save(userInfo2);
- UserInfo userInfo3 = new UserInfo()
- {
- UserId = "1003",
- UserName = "王五",
- PassWord = "123456",
- Detail = new Detail()
- {
- Address = "广东",
- Age = 20,
- Email = "wangwu@163.com"
- }
- };
- mongoCollection.Save(userInfo3);
- UserInfo userInfo4 = new UserInfo()
- {
- UserId = "1004",
- UserName = "赵六",
- PassWord = "123456",
- Detail = new Detail()
- {
- Address = "湖北"
- }
- };
- mongoCollection.Save(userInfo4);
- }
4. 定义一个查找数据的方法“Select”,它将查找用户详细信息中,地址在湖北的全部用户:
- <summary>
- /// 查询详细资料地址为湖北的用户信息
- /// </summary>
- public List<UserInfo> Select()
- {
- return mongoCollection.Linq().Where(x => x.Detail.Address == "湖北").ToList();
- }
5. 还定义一个删除数据的方法,将删除集合全部数据:
- /// <summary>
- /// 删除全部用户信息
- /// </summary>
- public void DeleteAll()
- {
- mongoCollection.Remove(x => true);
- }
6. 在Main方法中添加如下代码:
- static void Main(string[] args)
- {
- UserBLL userBll = new UserBLL();
- userBll.InsertSomeData();
- var users = userBll.Select();
- foreach (var user in users)
- {
- Console.WriteLine(user.UserName + "是湖北人");
- };
- userBll.DeleteAll();
- }
7. ***执行程序,打印如下信息:
李四是湖北人
赵六是湖北人
1) 普通实现
普通方式实现不想多讲,直接贴代码,看看与linq方式有什么区别:
- class Program
- {
- static void Main(string[] args)
- {
- UserBLL userBll = new UserBLL();
- userBll.InsertSomeData();
- var users = userBll.Select();
- foreach (var user in users)
- {
- Console.WriteLine(user["UserName"].ToString() + "是湖北人");
- };
- userBll.DeleteAll();
- Console.ReadLine();
- }
- }
- public class UserBLL
- {
- public string connectionString = "mongodb://localhost";
- public string databaseName = "myDatabase";
- public string collectionName = "UserInfo";
- private Mongo mongo;
- private MongoDatabase mongoDatabase;
- private MongoCollection<Document> mongoCollection;
- public UserBLL()
- {
- mongo = new Mongo(connectionString);
- mongoDatabase = mongo.GetDatabase(databaseName) as MongoDatabase;
- mongoCollection = mongoDatabase.GetCollection<Document>(collectionName) as MongoCollection<Document>;
- mongo.Connect();
- }
- ~UserBLL()
- {
- mongo.Disconnect();
- }
- /// <summary>
- /// 插入一些数据
- /// </summary>
- public void InsertSomeData()
- {
- Document userInfo1 = new Document();
- userInfo1["UserId"] = "1001";
- userInfo1["UserName"] = "张三";
- userInfo1["PassWord"] = "123456";
- mongoCollection.Save(userInfo1);
- Document userInfo2 = new Document();
- userInfo2["UserId"] = "1002";
- userInfo2["UserName"] = "李四";
- userInfo2["PassWord"] = "123456";
- //子文档
- var userInfo2Detail = new Document();
- userInfo2Detail["Address"] = "湖北";
- userInfo2Detail["Age"] = 20;
- userInfo2Detail["Email"] = "lisi@163.com";
- userInfo2["Detail"] = userInfo2Detail;
- mongoCollection.Save(userInfo2);
- Document userInfo3 = new Document();
- userInfo3["UserId"] = "1003";
- userInfo3["UserName"] = "王五";
- userInfo3["PassWord"] = "123456";
- var userInfo3Detail = new Document();
- userInfo3Detail["Address"] = "广东";
- userInfo3Detail["Age"] = 20;
- userInfo3Detail["Email"] = "wangwu@163.com";
- userInfo3["Detail"] = userInfo3Detail;
- mongoCollection.Save(userInfo3);
- Document userInfo4 = new Document();
- userInfo4["UserId"] = "1004";
- userInfo4["UserName"] = "赵六";
- userInfo4["PassWord"] = "123456";
- var userInfo4Detail = new Document();
- userInfo4Detail["Address"] = "湖北";
- userInfo4["Detail"] = userInfo4Detail;
- mongoCollection.Save(userInfo4);
- }
- /// <summary>
- /// 查询详细资料地址为湖北的用户信息
- /// </summary>
- public IEnumerable<Document> Select()
- {
- return mongoCollection.Find(new Document { { "Detail.Address", "湖北" } }).Documents;
- }
- /// <summary>
- /// 删除全部用户信息
- /// </summary>
- public void DeleteAll()
- {
- mongoCollection.Remove(new Document { });
- }
- }
***,我们通过这段代码输出全部用户资料信息的BJSON格式:
- /// <summary>
- /// 打印数据BJSON
- /// </summary>
- public void PrintBJSON()
- {
- string BJSON = string.Empty;
- foreach (var documet in mongoCollection.FindAll().Documents)
- {
- BJSON += documet.ToString();
- }
- Console.WriteLine(BJSON);
- }
- 结果如下:
- { "UserId": "1001", "UserName": "张三", "PassWord": "123456", "_id": "4d80ec1ab8a4731338000001" }
- { "UserId": "1002", "UserName": "李四", "PassWord": "123456", "Detail": { "Address": "湖北", "Age": 20, "Email": "lisi@163.com" }, "_id": "4d80ec1ab8a4731338000002" }
- { "UserId": "1003", "UserName": "王五", "PassWord": "123456", "Detail": { "Address": "广东", "Age": 20, "Email": "wangwu@163.com" }, "_id": "4d80ec1ab8a4731338000003" }
- { "UserId": "1004", "UserName": "赵六", "PassWord": "123456", "Detail": { "Address": "湖北" }, "_id": "4d80ec1ab8a4731338000004" }
#p#
二、包含“子集合”的集合操作
同样举个例子:有一个学校人事管理系统要统计班级和学生的信息,现在定义了一个“班级集合”,这个集合里面的学生字段是一个“学生集合”,包含了本班全部学生。
1) linq方式实现
基础配置我就不多说了,数据类定义如下:
- /// <summary>
- /// 班级信息
- /// </summary>
- public class ClassInfo
- {
- public string ClassName { get; set; }
- public List<Student> Students { get; set; }
- }
- /// <summary>
- /// 学生信息
- /// </summary>
- public class Student
- {
- public string Name { get; set; }
- public int Age { get; set; }
- }
查询叫“张三”的学生在哪个班级,以及他的详细信息:
(这里其实是ToList后在内存中查的,linq方式直接查询好像驱动不支持。)
- public List<ClassInfo> Select()
- {
- return mongoCollection.Linq().ToList().Where(x => x.Students.Exists(s => s.Name == "张三")).ToList();
- }
1) 普通实现
查询叫“张三”的学生在哪个班级,以及他的详细信息:
- public List<Document> Select()
- {
- var mongocollection = mongoDatabase.GetCollection("ClassInfo");
- return mongocollection.Find(new Document { { "Students.Name", "张三" } }).Documents.ToList();
- }
打印数据的BJSON:
- { "_id": "4d814bae5c5f000000005f63", "ClassName": "1001", "Students": [ { "Name": "张三", "Age": 10 }, { "Name": "李四", "Age": 0 } ] }
- { "_id": "4d814bae5c5f000000005f64", "ClassName": "1002", "Students": [ ] }
- { "_id": "4d814bae5c5f000000005f65", "ClassName": "1003", "Students": [ { "Name": "王五", "Age": 11 }, { "Name": "赵六", "Age": 9 } ] }
三、小结
通过本节例子我们发现,MongoDB有它独特的文档结构可以描述数据对象之间的一些关系特征。它虽然没有关系型数据库多表符合查询那样强大的表间查询方式,但也可以通过文档结构描述更灵活的关系特性,可以这么说,关系型数据库能做的,MongoDB基本上也可以做到。甚至有些关系数据库不容易做到的,MongoDB也可以轻松做到,比如,描述数据类的继承关系等
原文链接:http://www.cnblogs.com/lipan/archive/2011/03/17/1986616.html
【编辑推荐】
- MongoDB学习笔记(一) MongoDB介绍及安装
- MongoDB学习笔记(二) 通过samus驱动实现基本数据操作
- MongoDB学习笔记(三) 在MVC模式下通过Jqgrid表格操作MongoDB数据
- MongoDB,无模式文档型数据库简介
- Visual Studio 2010下编译调试MongoDB源码