mongodb聚合获取匹配文档的总数

Question

I have a following sample docs saved in mongogb, like:

{
    name: 'andy',
    age: 19,
    description: 'aaa aaa aaa'
}
{
    name: 'andy',
    age: 17,
    description: 'bbb bbb bbb'
}
{
    name: 'leo',
    age: 10,
    description: 'aaa aaa aaa'
}
{
    name: 'andy',
    age: 17,
    description: 'ccc ccc ccc'
}

what the pipeline should look like to get the total number of name in each of matched sets? so I can use this sum number for next pipe. the pipeline I currently have is this:

var pip = [
    {
        $match: { name: 'andy' }
    }
]

and I want to get this result like

{
    name: 'andy',
    age: 19,
    description: 'aaa aaa aaa',
    total_andy: 3
}
{
    name: 'andy',
    age: 17,
    description: 'bbb bbb bbb',
    total_andy: 3
}
{
    name: 'andy',
    age: 17,
    description: 'ccc ccc ccc',
    total_andy: 3
}

Answer 1

I am not exactly clear as to what you want. And i don't have enough reputation to ask for that in a comment. So let me have a shot at answering. If the answer isn't what you want, clarify the question further and we'll get to it...

var term1group = {$group : 
                            {'_id' : '$name'},
                            'total_names' : {$sum : 1},
                            'ageAndDescription' : {$addToSet : {'$age', '$description'}}
                    }

var term2unwind = {$unwind : '$ageAndDescription'}

var term3project = {$project : {
                            _id : 0,
                            'name' : '_id',
                            'age' : '$ageAndDescription.age',
                            'description' : '$ageAndDescription.description',
                            'total_name' : 1
                            }

db.collection.aggregate(term1group, term2unwind, term3project);

Haven't tested but i am hopeful this will work.

Answer 2

You just need to use a $group and $sum to do a simple count. The output won't match exactly, but you could reformat it with NodeJS easily.

You apparently want to group on the three fields shown ( name , age , and description ). To do that, just add the fields and a field reference (using $ ):

  { $match: { name: 'andy' } },
  { $group: { 
         _id: { name: "$name", age: "$age", description: "$description"}, 
         count: { $sum: 1} 
       }
  }

To add the count of each group, include a $sum of 1 (for each document that matches the group).

Your output will look something like:

{ "_id" : { "name" : "andy", "age" : 17, "description" : "ccc ccc ccc" }, "count" : 1 }
{ "_id" : { "name" : "andy", "age" : 17, "description" : "bbb bbb bbb" }, "count" : 1 }
{ "_id" : { "name" : "andy", "age" : 19, "description" : "aaa aaa aaa" }, "count" : 3 }

If you used a projection with $project , you could also format the output to more closely match your original request:

{ $match: {name: 'andy' }}, 
{ $group: { _id: { name: "$name", age: "$age", description: "$description"} , 
         count: {$sum: 1}}
}, 
{ $project : { name: "$_id.name", _id: 0, age: "$_id.age", 
              description: "$_id.description", total_andy: "$count" 
             }
}

Results:

{ "name" : "andy", "age" : 17, "description" : "ccc ccc ccc", "total_andy" : 1 }
{ "name" : "andy", "age" : 17, "description" : "bbb bbb bbb", "total_andy" : 1 }
{ "name" : "andy", "age" : 19, "description" : "aaa aaa aaa", "total_andy" : 3 }