mongodb aggregate(group) sub document 问题

Question

我的数据：

    { 
        "_id" : ObjectId("542911f4d7054ab6ee0de476"), 
        "name" : "policy2", 
        "url" : [ "www.url22.com", "www.url21.com" ], 
        "countries" : [ 
            { "country" : "USA" }, 
            { "country" : "GER" } 
        ] 
    }

我想单独group by countries.country字段，单独统计USA出现几次，China出现几次，而不是[USA,GER]在一起出现几次，如下：

{ "_id" : "China", "total" : 1 }
{ "_id" : "USA", "total" : 2 }
{ "_id" : "GER", "total" : 1 }
{ "_id" : "JP", "total" : 1 }

我尝试使用aggregate

db.policy.aggregate({$group:{_id:"$countries.country",total:{$sum:1}}})

但是得到的结果是countries合在一起的统计结果：

{ "_id" : [ "China" ], "total" : 1 }
{ "_id" : [ "USA", "JP" ], "total" : 1 }
{ "_id" : [ "USA", "GER" ], "total" : 1 }

求解决方案！！

Answer 1

使用 $unwind：

db.policy.aggregate([ 
  { $unwind: '$countries' },
  { $group: { _id: '$countries.country', total: { $sum: 1 } } }
]);

Answer 2

我找到解决方案了原来是使用$unwind函数

> db.policy.aggregate([{$unwind:"$country"},{$group:{_id:"$country",total:{$sum:1}}}])