> 欢迎大家关注公众号「JAVA前线」查看更多精彩分享文章，主要包括源码分析、实际应用、架构思维、职场分享、产品思考等等，同时欢迎大家加我微信「java_front」一起交流学习
# 1 文章概述 ## 1.1 基本概念 MongoDB是一种非关系型数据库，数据最终存储为BSON（Binary JSON）类型。MongoDB包含三个重要逻辑概念：数据库、集合和文档，与关系型数据库相关概念映射如下图： ![01 MongoDB与关系型数据库.jpg](https://p3-juejin.byteimg.com/tos-cn-i-k3u1fbpfcp/9efa60b9391742f4be0446bcf35fd2ca~tplv-k3u1fbpfcp-watermark.image?)
## 1.2 数据准备 ### 1.2.1 逻辑设计 bookdb数据库包含library与book两个集合，相关字段说明如下图： ![02 MongoDB逻辑设计.jpg](https://p3-juejin.byteimg.com/tos-cn-i-k3u1fbpfcp/6554a5bcecf14cd5ad12006663ef377f~tplv-k3u1fbpfcp-watermark.image?)
### 1.2.2 新增数据 ```text // 创建数据库 use bookdb // 新增图书馆 db.library.insertMany([ { "_id": 1, "library_name": "图书馆_1" }, { "_id": 2, "library_name": "图书馆_2" }, { "_id": 3, "library_name": "图书馆_3" } ]) // 新增图书 db.book.insertMany([ { "_id": 1, "book_type": 1, "book_name": "图书_1", "book_price": 10, "book_count": 100, "borrowers": [{ "name": "张三", "id": 100 }, { "name": "李四", "id": 200 }], "library_id": 1 }, { "_id": 2, "book_type": 2, "book_name": "图书_2", "book_price": 20, "book_count": 100, "borrowers": [{ "name": "张三", "id": 100 }], "library_id": 2 }, { "_id": 3, "book_type": 1, "book_name": "图书_3", "book_price": 30, "book_count": 100, "borrowers": [{ "name": "张三", "id": 100 }, { "name": "王五", "id": 300 }], "library_id": 2 } ]) ```
# 2 聚合概念 ## 2.1 基本语法 MongoDB聚合语法第一个参数是管道：文档在上个管道处理完后传递给下个管道，第二个参数是选项：设置聚合操作特性 ```text db.collection.aggregate(pipeline, options) ```
## 2.2 聚合管道常用聚合管道有以下九种类型： $project：对文档进行投影 $limit：出管道内前N个文档 $skip：跳过管道内前N个文档 $sort：对文档进行排序 $out：输出管道中文档 $match：对文档进行筛选 $unwind：铺平文档中的数组字段 $lookup：对文档进行查询 $group：对文档进行分组
# 3 聚合实例 ## 3.1 project 1表示展示某字段 0表示不展示某字段借阅人编号和姓名拆分成两个数组 ```text db.book.aggregate({ $project: { "_id": 0, "book_name": 1, "borrowerIds": "$borrowers.id", "borrowerNames": "$borrowers.name" } }) --------------------------------------------------------- // 1 { "book_name": "图书_1", "borrowerIds": [ 100, 200 ], "borrowerNames": [ "张三", "李四" ] } // 2 { "book_name": "图书_2", "borrowerIds": [ 100 ], "borrowerNames": [ "张三" ] } // 3 { "book_name": "图书_3", "borrowerIds": [ 100, 300 ], "borrowerNames": [ "张三", "王五" ] } ```
## 3.2 limit 只展示一个投影结果 ```text db.book.aggregate([ { $project: { "_id": 0, "book_name": 1, "borrowerIds": "$borrowers.id", "borrowerNames": "$borrowers.name" } }, { $limit: 1 } ]) --------------------------------------------------------- // 1 { "book_name": "图书_1", "borrowerIds": [ 100, 200 ], "borrowerNames": [ "张三", "李四" ] } ```
## 3.3 skip 跳过一个且只展示一个投影结果 ```text db.book.aggregate([ { $project: { "_id": 0, "book_name": 1, "borrowerIds": "$borrowers.id", "borrowerNames": "$borrowers.name" } }, { $skip: 1 }, { $limit: 1 } ]) --------------------------------------------------------- // 1 { "book_name": "图书_2", "borrowerIds": [ 100 ], "borrowerNames": [ "张三" ] } ```
## 3.4 sort ```text db.book.aggregate([ { $project: { "_id": 1, "book_name": 1, "library_id": 1 } }, { $sort: { "library_id": - 1, // 降序 "_id": 1 // 升序 } } ]) --------------------------------------------------------- // 1 { "_id": 2, "book_name": "图书_2", "library_id": 2 } // 2 { "_id": 3, "book_name": "图书_3", "library_id": 2 } // 3 { "_id": 1, "book_name": "图书_1", "library_id": 1 } ```
MongoDB内存排序有100M限制，如果排序数据过多需要设置选项`allowDiskUse=true`，表示数据可以写入临时文件进行排序 ```text db.book.aggregate([ { $project: { "_id": 1, "book_name": 1, "library_id": 1 } }, { $sort: { "library_id": - 1, "_id": 1 } } ], { allowDiskUse: true }) ```
## 3.5 out 投影结果输出到新集合 ```text db.book.aggregate([ { $project: { "_id": 0, "book_name": 1, "borrowerIds": "$borrowers.id", "borrowerNames": "$borrowers.name" } }, { $out: "newCollection" } ]) db.newCollection.find() --------------------------------------------------------- // 1 { "_id": ObjectId("62bec0636f9c37787b9590b9"), "book_name": "图书_1", "borrowerIds": [ 100, 200 ], "borrowerNames": [ "张三", "李四" ] } // 2 { "_id": ObjectId("62bec0636f9c37787b9590ba"), "book_name": "图书_2", "borrowerIds": [ 100 ], "borrowerNames": [ "张三" ] } // 3 { "_id": ObjectId("62bec0636f9c37787b9590bb"), "book_name": "图书_3", "borrowerIds": [ 100, 300 ], "borrowerNames": [ "张三", "王五" ] } ```
## 3.6 match where book_name = xxx ```text db.book.aggregate([ { $match: { "book_name": "图书_2" } }, { $project: { "_id": 1, "book_name": 1, "library_id": 1 } } ]) --------------------------------------------------------- // 1 { "_id": 2, "book_name": "图书_2", "library_id": 2 } ```
where library_id = 2 and price > 15 ```text db.book.aggregate([ { $match: { $and: [ { "library_id": 2 }, { "book_price": { $gt: 25 } } ] } }, { $project: { "_id": 1, "book_name": 1, "library_id": 1, "book_price": 1 } } ]) --------------------------------------------------------- // 1 { "_id": 3, "book_name": "图书_3", "book_price": 30, "library_id": 2 } ```
## 3.7 unwind 文档按照借阅人数组铺平 includeArrayIndex表示索引 ```text db.book.aggregate([ { $unwind: { path: "$borrowers", includeArrayIndex: "idx" } } ]) --------------------------------------------------------- // 1 { "_id": 1, "book_type": 1, "book_name": "图书_1", "book_price": 10, "book_count": 100, "borrowers": { "name": "张三", "id": 100 }, "library_id": 1, "idx": NumberLong("0") } // 2 { "_id": 1, "book_type": 1, "book_name": "图书_1", "book_price": 10, "book_count": 100, "borrowers": { "name": "李四", "id": 200 }, "library_id": 1, "idx": NumberLong("1") } // 3 { "_id": 2, "book_type": 2, "book_name": "图书_2", "book_price": 20, "book_count": 100, "borrowers": { "name": "张三", "id": 100 }, "library_id": 2, "idx": NumberLong("0") } // 4 { "_id": 3, "book_type": 1, "book_name": "图书_3", "book_price": 30, "book_count": 100, "borrowers": { "name": "张三", "id": 100 }, "library_id": 2, "idx": NumberLong("0") } // 5 { "_id": 3, "book_type": 1, "book_name": "图书_3", "book_price": 30, "book_count": 100, "borrowers": { "name": "王五", "id": 300 }, "library_id": 2, "idx": NumberLong("1") } ```
## 3.8 lookup 查询图书馆有哪些图书 lookup可以实现连表查询 MongoDB 3.4之前聚合语法： - from：待关联集合【book】 - localField: 本集合关联键【library】 - foreignField：待关联键【book】 - as：待关联集合数据【book】 ```text db.library.aggregate([ { $lookup: { from: "book", localField: "_id", foreignField: "library_id", as: "books_info" } } ]) --------------------------------------------------------- // 1 { "_id": 1, "library_name": "图书馆_1", "books_info": [ { "_id": 1, "book_type": 1, "book_name": "图书_1", "book_price": 10, "book_count": 100, "borrowers": [ { "name": "张三", "id": 100 }, { "name": "李四", "id": 200 } ], "library_id": 1 } ] } // 2 { "_id": 2, "library_name": "图书馆_2", "books_info": [ { "_id": 2, "book_type": 2, "book_name": "图书_2", "book_price": 20, "book_count": 100, "borrowers": [ { "name": "张三", "id": 100 } ], "library_id": 2 }, { "_id": 3, "book_type": 1, "book_name": "图书_3", "book_price": 30, "book_count": 100, "borrowers": [ { "name": "张三", "id": 100 }, { "name": "王五", "id": 300 } ], "library_id": 2 } ] } // 3 { "_id": 3, "library_name": "图书馆_3", "books_info": [ ] } ```
MongoDB 3.4之后聚合语法： - from：待关联集合【book】 - let：声明本集合字段在管道使用 - pipeline：操作管道 ```text db.library.aggregate([ { $lookup: { from: "book", let: { "lid": "$_id" }, pipeline: [ { $match: { $expr: { $and: [ { $eq: ["$$lid", "$library_id"] } ] } } } ], as: "books_info" } } ]) --------------------------------------------------------- // 1 { "_id": 1, "library_name": "图书馆_1", "books_info": [ { "_id": 1, "book_type": 1, "book_name": "图书_1", "book_price": 10, "book_count": 100, "borrowers": [ { "name": "张三", "id": 100 }, { "name": "李四", "id": 200 } ], "library_id": 1 } ] } // 2 { "_id": 2, "library_name": "图书馆_2", "books_info": [ { "_id": 2, "book_type": 2, "book_name": "图书_2", "book_price": 20, "book_count": 100, "borrowers": [ { "name": "张三", "id": 100 } ], "library_id": 2 }, { "_id": 3, "book_type": 1, "book_name": "图书_3", "book_price": 30, "book_count": 100, "borrowers": [ { "name": "张三", "id": 100 }, { "name": "王五", "id": 300 } ], "library_id": 2 } ] } // 3 { "_id": 3, "library_name": "图书馆_3", "books_info": [ ] } ```
新增价格大于20查询条件 ```text db.library.aggregate([ { $lookup: { from: "book", let: { "lid": "$_id" }, pipeline: [ { $match: { $expr: { $and: [ { $eq: ["$$lid", "$library_id"] }, { $gt: ["$book_price", 20] } ] } } } ], as: "books_info" } } ]) --------------------------------------------------------- // 1 { "_id": 1, "library_name": "图书馆_1", "books_info": [ ] } // 2 { "_id": 2, "library_name": "图书馆_2", "books_info": [ { "_id": 3, "book_type": 1, "book_name": "图书_3", "book_price": 30, "book_count": 100, "borrowers": [ { "name": "张三", "id": 100 }, { "name": "王五", "id": 300 } ], "library_id": 2 } ] } // 3 { "_id": 3, "library_name": "图书馆_3", "books_info": [ ] } ```
## 3.9 group ### 3.9.1 简单统计 _id：图书类型作为分组键 count：每个类型有多少种书 ```text db.book.aggregate([ { $group: { _id: "$book_type", count: { $sum: 1 } } } ]) --------------------------------------------------------- // 1 { "_id": 2, "count": 1 } // 2 { "_id": 1, "count": 2 } ```
### 3.9.2 复杂统计 _id：图书类型作为分组键 type_count：每个类型有多少种书 type_book_count：每个类型有多少本书 minTotalPrice：每个类型总价最小值 maxTotalPrice：每个类型总价最大值 totalPrice：每个类型总价 avgPrice：每个类型平均价 ```text db.book.aggregate([ { $group: { _id: "$book_type", type_count: { $sum: 1 }, type_book_count: { $sum: "$book_count" }, minTotalPrice: { $min: { $multiply: ["$book_price", "$book_count"] } }, maxTotalPrice: { $max: { $multiply: ["$book_price", "$book_count"] } }, totalPrice: { $sum: { $multiply: ["$book_price", "$book_count"] } }, avgPrice: { $avg: "$book_price" } } } ]) --------------------------------------------------------- // 1 { "_id": 2, "type_count": 1, "type_book_count": 100, "minTotalPrice": 2000, "maxTotalPrice": 2000, "totalPrice": 2000, "avgPrice": 20 } // 2 { "_id": 1, "type_count": 2, "type_book_count": 200, "minTotalPrice": 1000, "maxTotalPrice": 3000, "totalPrice": 4000, "avgPrice": 20 } ```
### 3.9.3 空分组键 _id：空分组键表示统计全量数据 ```text db.book.aggregate([ { $group: { _id: null, type_count: { $sum: 1 }, type_book_count: { $sum: "$book_count" }, minTotalPrice: { $min: { $multiply: ["$book_price", "$book_count"] } }, maxTotalPrice: { $max: { $multiply: ["$book_price", "$book_count"] } }, totalPrice: { $sum: { $multiply: ["$book_price", "$book_count"] } }, avgPrice: { $avg: "$book_price" } } } ]) --------------------------------------------------------- // 1 { "_id": null, "type_count": 3, "type_book_count": 300, "minTotalPrice": 1000, "maxTotalPrice": 3000, "totalPrice": 6000, "avgPrice": 20 } ```
# 4 文章总结第一介绍了MongoDB与关系型数据库关系，并且准备本文需要的测试数据，第二介绍了聚合语法和聚合管道相关概念，第三通过实例介绍了如何使用聚合操作，希望本文对大家有所帮助。
> 欢迎大家关注公众号「JAVA前线」查看更多精彩分享文章，主要包括源码分析、实际应用、架构思维、职场分享、产品思考等等，同时欢迎大家加我微信「java_front」一起交流学习