$group 使用 $month 不在字符串化日期上工作

Question

我正在使用 mLab，它似乎不想让我使用 ISODate("2016-06-16T00:00:00.201Z") 所以我最终不得不将我在 mLab 中的所有日期作为字符串行 "transaction_date": "2016-06-16T00:00:00.201Z".

字符串化的日期似乎适合我的数据范围：

$match {
   "transaction_date" : {$gte: "2016-04-01"}, 
   "transaction_date" : {$lt: "2016-08-01"}, 
       }

但是，当我尝试做最后一组以按月份对结果求和时，它不喜欢使用日期。

$group {
   "revenue_month" : {$month : "transaction_date"},
   "net_revenue" : {$sum: "item_net_total_value"
       }

为什么我会收到 'unknown group operator $month' 消息？谢谢

下面是完整的脚本：

db.datawarehouse.aggregate(

// Pipeline
[
// Stage 1 - Isolate correct document types
{
  $match: {
  "object_class" : "Goods & Services Transaction", 
  "object_type": "Transaction", 
  "object_category" : "Revenue",
  "object_origin_category" : "Bookkeeping"
  }
},

// Stage 2 - Restrict data set by date range
{
  $match: {
  "transaction_date" : {$gte: "2016-04-01"}, 
  "transaction_date" : {$lt: "2016-08-01"}, 
  }
},

// Stage 3 - For duplicate handling, sort by invoice number in descending creation date order
{
  $sort: {
  "transaction_reference":1, 
  "object_creation_date": -1
  }
},

// Stage 4 - For duplicate handing use $first to use latest version of all documents
{
  $group: {
  _id: "$transaction_reference", 
  "party_uuid" : {$first: "$party_uuid"},
  "transaction_date" : {$first: "$transaction_date"},
  "transaction_net_value" : {$first: "$transaction_net_value"},
  "object_category" : {$first: "$object_category"},
  "transaction_status" : {$first: "$transaction_status"},
  "object_origin_category" : {$first: "$object_origin_category"},
  "object_origin" : {$first: "$object_origin"},
  "customer" : {$first: "$party.customer.customer_name"},
  "goods_and_services" : {$first: "$goods_and_services"}
  }
},

// Stage 5 - Widen set from transaction level to line item level
{
  $unwind: "$goods_and_services"
},

// Stage 6 - Make the complete data set required per record
{
  $project: {
  {"_id" : 1,
    "party_uuid" : "$party_uuid",
    "transaction_date" : "$transaction_date",
    "transaction_net_value" : "$transaction_net_value",
    "object_category" : "$object_category",
    "object_origin_category" : "$object_origin_category",
    "object_origin" : "$object_origin",
    "item_quantity" : "$goods_and_services.item_quantity",
    "item_name" : "$goods_and_services.item_name",
    "item_category" : "$goods_and_services.item_category",
    "item_ net_total_sale_value" : "$goods_and_services.item_net_total_value",
  }
},

// Stage 7 - Restrict records by item category
{
  $match: {
  "item_category" : "Sales Revenue"
  }
}
// Stage 8 - Group by transaction_date and SUM item_net_total_value
]

);

Answer 1

关于 $group 运算符的 docs：

... The argument can be any expression as long as it resolves to a date ...

当您向运算符传递一个字符串时，它显然无法解析为日期，因此不起作用。

解决此问题的方法是获取日期的正确子字符串

$group: {
    _id: { $substr : ["$transaction_date", 5, 2 ] },
    "net_revenue" : { $sum: "item_net_total_value" },
}

顺便说一句：我没有使用 mLab，但如果他们不允许日期，我会感到惊讶。将此答案视为中间 'solution' 但我建议要么针对 mLab 和日期的问题创建一个新问题，要么只是询问他们中的某个人以确保您猜测的限制是正确的，因为从长远来看它是最好使用正确的数据类型。

$group 使用 $month 不在字符串化日期上工作

$group using $month not working on stringified date

mongodb

mongodb-aggregation