MongoDB 聚合框架类比MYSQL

迷南。 2022-08-06 01:05 99阅读 0赞

[MongoDB 2.1][]引入了聚合框架，可以替代MapReduce用于一般的聚合操作。如果你有看过相关文档，应该已经注意到了这个新特性。本文就主要介绍一下MongoDB 2.1中的聚合框架。

Pipeline语法简介

MongoDB聚合就是把一系列特殊操作符作用于一个集合。一个操作符就是一个拥有单个属性的JavaScript对象，其属性即操作符名称，其值是一个可选对象：

支持的操作符命名有：$project, $match, $limit, $skip, $unwind, $group, and $sort,它们每个都有其各自的选项集。一系列操作符就称为管道（Pipeline）：

当在执行一个Pipeline时，MongoDB会互相传递操作符。”传递”在此处借用了其在Linux中的含义：一个操作符的输出会成为接下来操作符的输入。而每个操作符的结果会是文档的一个新的集合。所以MongoDB会如下所示来执行前面的管道：

你可以给一个管道随意添加任意多的操作符，甚至是在两个不同的位置两次添加相同操作符：

这也就解释了为何一个管道不写成简单的JavaScript对象，而是一个对象集：在一个对象中，同一个操作符不能出现两次：

要在一个MongoDB集合上执行管道，则要在集合上使用aggregate（）函数：

提示：如果你使用Node.js，本地适配器（从v0.9.9.2开始）和ODM（从v3.1.0开始）都是支持新聚合框架的。例如，想要在MongoDB模型上执行之前的Pipeline，你只需要写如下代码：

聚合框架的主要好处是MongoDB在执行它时省却了JavaScript引擎的开销。直接以C++实现使得它执行起来速度是非常快的。相比较于经典SQL聚合，聚合框架的主要限制就是它被局限于一个单一集合。也就是说，你不能应用类似连接的操作在数个集合上进行MongoDB聚合。除此之外，它还是非常之强大。

在本文，我还将举例说明Pipeline操作符的威力，并与它们SQL中的同类进行比较。

选择，重命名，组合

可以使用$project 操作符来选择或是重命名集合中的属性，这与SQL中SELECT语句的使用是类似的

$project 操作符还可以使用任意支持的表达式操作符（$and, $or, $gt, $lt, $eq, $add, $mod, $substr, $toLower, $toUpper, $dayOfWeek, $hour, $cond, $ifNull, to name a few） 来创建组合字段以及子文档。

归并文档

归并文档用的就是$group操作符 。

<table style="padding:0px; margin:0px auto; font-size:14px; font-family:微软雅黑; color:rgb(51,51,51); line-height:21px; border:1px dotted rgb(204,204,204); table-layout:fixed"> 
 <tbody style="padding:0px; margin:0px"> 
 <tr style="padding:0px; margin:0px"> 
 <td style="padding:0px; margin:0px; font-size:14px; font-family:微软雅黑; word-wrap:break-word"> //&nbsp;fastest&nbsp;way&nbsp; &gt;&nbsp;db.books.count();&nbsp; 3&nbsp; //&nbsp;if&nbsp;you&nbsp;really&nbsp;want&nbsp;to&nbsp;use&nbsp;aggregation&nbsp; &gt;&nbsp;db.books.aggregate([&nbsp; &nbsp;&nbsp;{&nbsp;$group:&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;//&nbsp;_id&nbsp;is&nbsp;required,&nbsp;so&nbsp;give&nbsp;it&nbsp;a&nbsp;constant&nbsp;value&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;//&nbsp;to&nbsp;group&nbsp;all&nbsp;the&nbsp;collection&nbsp;into&nbsp;one&nbsp;result&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;null,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;//&nbsp;increment&nbsp;nbBooks&nbsp;for&nbsp;each&nbsp;document&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;nbBooks:&nbsp;{&nbsp;$sum:&nbsp;1&nbsp;}&nbsp; &nbsp;&nbsp;}&nbsp;}&nbsp; ]);&nbsp; [&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;null,&nbsp;nbBooks:&nbsp;3&nbsp;}&nbsp; ]&nbsp; &nbsp;&nbsp; &gt;&nbsp;SELECT&nbsp;COUNT(*)&nbsp;FROM&nbsp;book;&nbsp; +----------+&nbsp; |&nbsp;COUNT(*)&nbsp;|&nbsp; +----------+&nbsp; |&nbsp;3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; +----------+&nbsp; //&nbsp;sample&nbsp;data&nbsp; &gt;&nbsp;db.books.find()&nbsp; [&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;147,&nbsp;title:&nbsp;"War&nbsp;and&nbsp;Peace",&nbsp;author_id:&nbsp;72347&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;148,&nbsp;title:&nbsp;"Anna&nbsp;Karenina",&nbsp;author_id:&nbsp;72347&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;149,&nbsp;title:&nbsp;"Pride&nbsp;and&nbsp;Prejudice",&nbsp;author_id:&nbsp;42345&nbsp;}&nbsp; ]&nbsp; &nbsp;&nbsp; #&nbsp;sample&nbsp;data&nbsp; &gt;&nbsp;SELECT&nbsp;*&nbsp;FROM&nbsp;book&nbsp; +-----+---------------------+-----------+&nbsp; |&nbsp;id&nbsp;&nbsp;|&nbsp;title&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;author_id&nbsp;|&nbsp; +-----+---------------------+-----------+&nbsp; |&nbsp;147&nbsp;|&nbsp;War&nbsp;and&nbsp;Peace&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;72347&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;148&nbsp;|&nbsp;Anna&nbsp;Karenina&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;72347&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;149&nbsp;|&nbsp;Pride&nbsp;and&nbsp;Prejudice&nbsp;|&nbsp;42345&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; +-----+---------------------+-----------+&nbsp; &gt;&nbsp;db.books.aggregate([&nbsp; &nbsp;&nbsp;{&nbsp;$group:&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;//&nbsp;group&nbsp;by&nbsp;author_id&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;"$author_id",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;//&nbsp;increment&nbsp;nbBooks&nbsp;for&nbsp;each&nbsp;document&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;nbBooks:&nbsp;{&nbsp;$sum:&nbsp;1&nbsp;}&nbsp; &nbsp;&nbsp;}&nbsp;}&nbsp; ]);&nbsp; [&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;72347,&nbsp;nbBooks:&nbsp;2&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;42345,&nbsp;nbBooks:&nbsp;1&nbsp;}&nbsp; ]&nbsp; &nbsp;&nbsp; &gt;&nbsp;SELECT&nbsp;author_id,&nbsp;COUNT(*)&nbsp; &nbsp;&nbsp;FROM&nbsp;book&nbsp; &nbsp;&nbsp;GROUP&nbsp;BY&nbsp;author_id;&nbsp; +-----------+----------+&nbsp; |&nbsp;author_id&nbsp;|&nbsp;COUNT(*)&nbsp;|&nbsp; +-----------+----------+&nbsp; |&nbsp;72347&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;42345&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; +-----------+----------+</td> 
 </tr> 
 </tbody> 
</table>

多操作符Pipeline

一个管道可能不止有一个操作符。以下就是$group操作符和$project的组合：

更为复杂的聚合

$group支持大量的聚合函数：$first, $last, $min, $max, $avg, $sum, $push, 以及$addToSet。可以查看MongoDB文档[http://docs.mongodb.org/manual/reference/aggregation][http_docs.mongodb.org_manual_reference_aggregation]

<table style="padding:0px; margin:0px auto; font-size:14px; font-family:微软雅黑; color:rgb(51,51,51); line-height:21px; border:1px dotted rgb(204,204,204); table-layout:fixed"> 
 <tbody style="padding:0px; margin:0px"> 
 <tr style="padding:0px; margin:0px"> 
 <td style="padding:0px; margin:0px; font-size:14px; font-family:微软雅黑; word-wrap:break-word"> //&nbsp;sample&nbsp;data&nbsp; &gt;&nbsp;db.reviews.find();&nbsp; [&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;"455",&nbsp;bookId:&nbsp;"974147",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;date:&nbsp;new&nbsp;Date("2012-07-10"),&nbsp;score:&nbsp;1&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;"456",&nbsp;bookId:&nbsp;"345335",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;date:&nbsp;new&nbsp;Date("2012-07-12"),&nbsp;score:&nbsp;5&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;"457",&nbsp;bookId:&nbsp;"345335",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;date:&nbsp;new&nbsp;Date("2012-07-13"),&nbsp;score:&nbsp;2&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;"458",&nbsp;bookId:&nbsp;"974147",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;date:&nbsp;new&nbsp;Date("2012-07-16"),&nbsp;score:&nbsp;3&nbsp;}&nbsp; ]&nbsp; &nbsp;&nbsp; #&nbsp;sample&nbsp;data&nbsp; &gt;&nbsp;SELECT&nbsp;*&nbsp;FROM&nbsp;review;&nbsp; +-----+---------+--------------+-------+&nbsp; |&nbsp;id&nbsp;&nbsp;|&nbsp;book_id&nbsp;|&nbsp;date&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;score&nbsp;|&nbsp; +-----+---------+--------------+-------+&nbsp; |&nbsp;455&nbsp;|&nbsp;974147&nbsp;&nbsp;|&nbsp;"2012-07-10"&nbsp;|&nbsp;1&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;456&nbsp;|&nbsp;345335&nbsp;&nbsp;|&nbsp;"2012-07-12"&nbsp;|&nbsp;5&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;457&nbsp;|&nbsp;345335&nbsp;&nbsp;|&nbsp;"2012-07-13"&nbsp;|&nbsp;2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;458&nbsp;|&nbsp;974147&nbsp;&nbsp;|&nbsp;"2012-07-16"&nbsp;|&nbsp;3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; +-----+---------+--------------+-------+&nbsp; &gt;&nbsp;db.reviews.aggregate([&nbsp; &nbsp;&nbsp;{&nbsp;$group:&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;"$bookId",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;avgScore:&nbsp;&nbsp;{&nbsp;$avg:&nbsp;"$score"&nbsp;},&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;maxScore:&nbsp;&nbsp;{&nbsp;$max:&nbsp;"$score"&nbsp;},&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;nbReviews:&nbsp;{&nbsp;$sum:&nbsp;1&nbsp;}&nbsp; &nbsp;&nbsp;}&nbsp;}&nbsp; ]);&nbsp; [&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;345335,&nbsp;avgScore:&nbsp;3.5,&nbsp;maxScore:&nbsp;5,&nbsp;nbReviews:&nbsp;2&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp;_id:&nbsp;974147,&nbsp;avgScore:&nbsp;3,&nbsp;maxScore:&nbsp;3,&nbsp;nbReviews:&nbsp;2&nbsp;}&nbsp; ]&nbsp; &nbsp;&nbsp; &gt;&nbsp;SELECT&nbsp;book_id,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;AVG(score)&nbsp;as&nbsp;avg_score,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;MAX(score)&nbsp;as&nbsp;max_score,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;COUNT(*)&nbsp;as&nbsp;nb_reviews&nbsp; &nbsp;&nbsp;FROM&nbsp;review&nbsp; &nbsp;&nbsp;GROUP&nbsp;BY&nbsp;book_id&nbsp;;&nbsp; +---------+------------+----------+------------+&nbsp; |&nbsp;book_id&nbsp;|&nbsp;avg_score&nbsp;|&nbsp;max_score&nbsp;|&nbsp;nb_reviews&nbsp;|&nbsp; +---------+------------+----------+------------+&nbsp; |&nbsp;345335&nbsp;&nbsp;|&nbsp;3.5&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;5&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;974147&nbsp;&nbsp;|&nbsp;2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;3&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;2&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; +---------+------------+----------+------------+</td> 
 </tr> 
 </tbody> 
</table>

条件

你可以对集合加以限制，使其被查询对象处理，再传递给$match操作符 。至于你是将此操作符置于$group操作符之前还是之后，也就决定着它在SQL中的同等角色是WHERE还是HAVING。

开发嵌入式数组

如果集合中的文件包含数组，那么你就可以使用操作符将这些数组分散到几个特定的文档。

<table style="padding:0px; margin:0px auto; font-size:14px; font-family:微软雅黑; color:rgb(51,51,51); line-height:21px; border:1px dotted rgb(204,204,204); table-layout:fixed"> 
 <tbody style="padding:0px; margin:0px"> 
 <tr style="padding:0px; margin:0px"> 
 <td style="padding:0px; margin:0px; font-size:14px; font-family:微软雅黑; word-wrap:break-word"> //&nbsp;sample&nbsp;data&nbsp; &gt;&nbsp;db.articles.find();&nbsp; [&nbsp; &nbsp;&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;12351254,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;title:&nbsp;"Space&nbsp;Is&nbsp;Getting&nbsp;Closer",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;tags:&nbsp;["science",&nbsp;"space",&nbsp;"iss"]&nbsp; &nbsp;&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;22956492,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;title:&nbsp;"Computer&nbsp;Solves&nbsp;Rubiks&nbsp;Cube",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;tags:&nbsp;["computing",&nbsp;"science"]&nbsp; &nbsp;&nbsp;}&nbsp; ]&nbsp; &nbsp;&nbsp; #&nbsp;sample&nbsp;data&nbsp; &gt;&nbsp;SELECT&nbsp;*&nbsp;FROM&nbsp;article;&nbsp; +------------+---------------------------+&nbsp; |&nbsp;id&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;title&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; +----------+-----------------------------+&nbsp; |&nbsp;12351254&nbsp;|&nbsp;Space&nbsp;Is&nbsp;Getting&nbsp;Closer&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;22956492&nbsp;|&nbsp;Computer&nbsp;Solves&nbsp;Rubiks&nbsp;Cube&nbsp;|&nbsp; +------------+---------------------------+&nbsp; &gt;&nbsp;SELECT&nbsp;*&nbsp;FROM&nbsp;tag;&nbsp; +-----+------------+-----------+&nbsp; |&nbsp;id&nbsp;&nbsp;|&nbsp;article_id&nbsp;|&nbsp;name&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; +-----+------------+-----------+&nbsp; |&nbsp;534&nbsp;|&nbsp;12351254&nbsp;&nbsp;&nbsp;|&nbsp;science&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;535&nbsp;|&nbsp;12351254&nbsp;&nbsp;&nbsp;|&nbsp;space&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;536&nbsp;|&nbsp;12351254&nbsp;&nbsp;&nbsp;|&nbsp;iss&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;816&nbsp;|&nbsp;22956492&nbsp;&nbsp;&nbsp;|&nbsp;computing&nbsp;|&nbsp; |&nbsp;817&nbsp;|&nbsp;22956492&nbsp;&nbsp;&nbsp;|&nbsp;science&nbsp;&nbsp;&nbsp;|&nbsp; +-----+------------+-----------+&nbsp; &gt;&nbsp;db.articles.aggregate([&nbsp; &nbsp;&nbsp;{&nbsp;$unwind:&nbsp;"$tags"&nbsp;}&nbsp; ]);&nbsp; [&nbsp; &nbsp;&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;12351254,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;title:&nbsp;"Space&nbsp;Is&nbsp;Getting&nbsp;Closer",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;tags:&nbsp;"science"&nbsp; &nbsp;&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;12351254,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;title:&nbsp;"Space&nbsp;Is&nbsp;Getting&nbsp;Closer",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;tags:&nbsp;"space"&nbsp; &nbsp;&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;22956492,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;title:&nbsp;"Computer&nbsp;Solves&nbsp;Rubiks&nbsp;Cube",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;tags:&nbsp;"computing"&nbsp; &nbsp;&nbsp;},&nbsp; &nbsp;&nbsp;{&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;_id:&nbsp;22956492,&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;title:&nbsp;"Computer&nbsp;Solves&nbsp;Rubiks&nbsp;Cube",&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;tags:&nbsp;"science"&nbsp; &nbsp;&nbsp;}&nbsp; ]&nbsp; &nbsp;&nbsp; &gt;&nbsp;SELECT&nbsp;article.id,&nbsp;article.title,&nbsp;tag.name&nbsp; &nbsp;&nbsp;FROM&nbsp;article&nbsp;LEFT&nbsp;JOIN&nbsp;tag&nbsp; &nbsp;&nbsp;ON&nbsp;article.id&nbsp;=&nbsp;tag.article_id;&nbsp; +------------+-----------------------------+-----------+&nbsp; |&nbsp;article.id&nbsp;|&nbsp;article.title&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;tag.name&nbsp;&nbsp;|&nbsp; +------------+-----------------------------+-----------+&nbsp; |&nbsp;12351254&nbsp;&nbsp;&nbsp;|&nbsp;Space&nbsp;Is&nbsp;Getting&nbsp;Closer&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;science&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;12351254&nbsp;&nbsp;&nbsp;|&nbsp;Space&nbsp;Is&nbsp;Getting&nbsp;Closer&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp;space&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;|&nbsp; |&nbsp;22956492&nbsp;&nbsp;&nbsp;|&nbsp;Computer&nbsp;Solves&nbsp;Rubiks&nbsp;Cube&nbsp;|&nbsp;computing&nbsp;|&nbsp; |&nbsp;22956492&nbsp;&nbsp;&nbsp;|&nbsp;Computer&nbsp;Solves&nbsp;Rubiks&nbsp;Cube&nbsp;|&nbsp;science&nbsp;&nbsp;&nbsp;|&nbsp; +------------+-----------------------------+-----------+</td> 
 </tr> 
 </tbody> 
</table>

聚合开发数组

聚合框架真正的威力是在你将$unwind传送给$group时才得以体现的。这与在SQL中使用LEFT JOIN…GROUP BY是类似的。

结论

想象下可以用这个功能来做些什么呢?一个接着一个的传输操作符可以进行归并，排序，限定等操作。在MongoDB自带文档中有个很具代表性的例子，它是用两个连续的$group操作符来组成一个管道。而在SQL数据库中只能用子查询才能做到这一点。

如果你所用的MapReduce功能足够简单，则可将你的MongoDB代码重构为聚合框架，执行起来会更快。

[MongoDB 2.1]: http://www.searchdatabase.com.cn/showcontent_67311.htm
[http_docs.mongodb.org_manual_reference_aggregation]: http://docs.mongodb.org/manual/reference/aggregation