MongoDB pipeline operators

Friends who are familiar with the Linux operating system should know that there are pipes in Linux, which can be used to process data conveniently. MongoDB version 2.2 also introduces a new data aggregation framework. A document can pass through a pipeline composed of multiple nodes. Each node has its own special functions, such as document grouping, document filtering, etc. Each node will accept a series of documents. , do some type conversion on these documents, and then pass the converted documents to the next node, and the last node will return the results to the client. In this article, we will first look at some basic pipeline operators.

$match

$match can be used to filter documents. After the filtering is completed, we can filter the obtained Let’s do data aggregation on the document subset. The query operators we introduced before can be used in $match. For example, get all the documents whose author is "Du Fu" in the collection, as follows:

db.sang_collect.aggregate({$match:{author:"杜甫"}})

We are here In actual use, it is best to put $match in front of the pipeline, which can reduce the workload of the subsequent pipeline. At the same time, we can also use the index when executing $match before projection and grouping.

$project

Basic usage

$project can be used to extract the desired field, as follows:

db.sang_collect.aggregate({$project:{title:1,_id:0}})

1 means that the field is required, 0 means If you do not need this field, you can also rename the returned field, for example, change title to articleTitle, as follows:

db.sang_collect.aggregate({$project:{"articleTitle":"$title"}})

However, there is a problem that needs attention here. If there is an index on the original field, the field after renaming There will be no index on it, so it is best to use the index before renaming.

Mathematical expression

Mathematical expression can be used to add, subtract, multiply, and divide a set of values ​​to modulo. For example, my data structure is as follows:

{
    "_id" : ObjectId("59f841f5b998d8acc7d08863"),
    "orderAddressL" : "ShenZhen",
    "prodMoney" : 45.0,
    "freight" : 13.0,
    "discounts" : 3.0,
    "orderDate" : ISODate("2017-10-31T09:27:17.342Z"),
    "prods" : [ 
        "可乐", 
        "奶茶"
    ]
}

The total cost of the order Add freight to the cost of the product, and the query is as follows:

db.sang_collect.aggregate({$project:{totalMoney:{$add:["$prodMoney","$freight"]}}})

The actual payment cost is the total cost minus the discount, as follows:

db.sang_collect.aggregate({$project:{totalPay:{$subtract:[{$add:["$prodMoney","$freight"]},"$discounts"]}}})

Let’s do three more nonsensical operations, such as calculating prodMoney, freight and discounts The product of:

db.sang_collect.aggregate({$project:{test1:{$multiply:["$prodMoney","$freight","$discounts"]}}})

Another example is to find the quotient of $prodMoney and $freight, as follows:

db.sang_collect.aggregate({$project:{test1:{$pide:["$prodMoney","$freight"]}}})

Another example is to use $freight to modulo $prodMoney, as follows:

db.sang_collect.aggregate({$project:{test1:{$mod:["$prodMoney","$freight"]}}})

Both addition and multiplication can accept multiple arguments, the rest accept two arguments.

Date expression

Date expression can extract the year, month, day, week, hour, minute, second and other information from a date type, as follows:

db.sang_collect.aggregate({$project:{"年份":{$year:"$orderDate"},"月份":{$month:"$orderDate"},"一年中第几周":{$week:"$orderDate"},"日期":{$dayOfMonth:"$orderDate"},"星期":{$dayOfWeek:"$orderDate"},"一年中第几天":{$dayOfYear:"$orderDate"},"时":{$hour:"$orderDate"},"分":{$minute:"$orderDate"},"秒":{$second:"$orderDate"},"毫秒":{$millisecond:"$orderDate"},"自定义格式化时间":{$dateToString:{format:"%Y年%m月%d %H:%M:%S",date:"$orderDate"}}}})

The execution results are as follows:

{
    "_id" : ObjectId("59f841f5b998d8acc7d08861"),
    "年份" : 2017,
    "月份" : 10,
    "一年中第几周" : 44,
    "日期" : 31,
    "星期" : 3,
    "一年中第几天" : 304,
    "时" : 9,
    "分" : 27,
    "秒" : 17,
    "毫秒" : 342,
    "自定义格式化时间" : "2017年10月31 09:27:17"
}

$dayOfWeek returns the week, 1 represents Sunday, 7 represents Saturday, $week represents the week of the year, starting from 0. $dateToString is a feature in MongoDB3.0+. There are also the following formatting characters:

##%MMinute (2 digits, zero padded)00-59%SSecond (2 digits, zero padded)00-60##%L%j##%wDay of week (1-Sunday, 7-Saturday)1-7%UWeek of year (2 digits, zero padded)00-53

字符串表达式

字符串表达式中有字符串的截取、拼接、转大写、转小写等操作，比如我截取orderAddressL前两个字符返回，如下：

db.sang_collect.aggregate({$project:{addr:{$substr:["$orderAddressL",0,2]}}})

再比如我将orderAddressL和orderDate拼接后返回：

db.sang_collect.aggregate({$project:{addr:{$concat:["$orderAddressL",{$dateToString:{format:"--%Y年%m月%d",date:"$orderDate"}}]}}})

结果如下：

{
    "_id" : ObjectId("59f841f5b998d8acc7d08861"),
    "addr" : "NanJing--2017年10月31"
}

再比如我将orderAddressL全部转为小写返回：

db.sang_collect.aggregate({$project:{addr:{$toLower:"$orderAddressL"}}})

再比如我将orderAddressL全部转为大写返回：

db.sang_collect.aggregate({$project:{addr:{$toUpper:"$orderAddressL"}}})

逻辑表达式

想要比较两个数字的大小，可以使用$cmp操作符，如下：

db.sang_collect.aggregate({$project:{test:{$cmp:["$freight","$discounts"]}}})

如果第一个参数大于第二个参数返回正数，第一个参数小于第二个则返回负数，也可以利用$strcasecmp来比较字符串（中文无效）：

db.sang_collect.aggregate({$project:{test:{$strcasecmp:[{$dateToString:{format:"..%Y年%m月%d",date:"$orderDate"}},"$orderAddressL"]}}})

至于我们之前介绍的$eq/$ne/$gt/$gte/$lt/$lte等操作符在这里一样是适用的。另外还有$and、$or、$not等表达式可用，以$and为例，如下：

db.sang_collect.aggregate({$project:{test:{$and:[{"$eq":["$freight","$prodMoney"]},{"$eq":["$freight","$discounts"]}]}}})

$and中的每个参数都为true时返回true，$or则表示参数中有一个为true就返回true，$not则会对它的参数的值取反，如下：

db.sang_collect.aggregate({$project:{test:{$not:{"$eq":["$freight","$prodMoney"]}}}})

另外还有两个流程控制语句，如下：

db.sang_collect.aggregate({$project:{test:{$cond:[false,"trueExpr","falseExpr"]}}})

$cond第一个参数如果为true，则返回trueExpr，否则返回falseExpr.

db.sang_collect.aggregate({$project:{test:{$ifNull:[null,"replacementExpr"]}}})

$ifNull第一个参数如果为null，则返回replacementExpr，否则就返回第一个参数。

以上内容就是MongoDB管道操作符部分介绍，希望能帮助到大家。

MongoDB固定集合详解

node.js操作MongoDB的代码案例分享

phpstudy如何扩展MongoDB

Character	Meaning	Value range
%Y	Year (4 digits, zero padded)	0000-9999
%m	Month (2 digits, zero padded)	01-12
%d	Day of Month (2 digits, zero padded )	01-31
%H	Hour (2 digits, zero padded, 24-hour clock)	00 -23
Millisecond (3 digits , zero padded)	000-999
Day of year (3 digits, zero padded)	001- 366

The above is the detailed content of MongoDB pipeline operators. For more information, please follow other related articles on the PHP Chinese website!