MongoDb aggregate the count/sum of 3 fields - mongodb

Question

Im trying to get a sum of content a user has liked, shared or commented.
my content model has the following structure.
Content = {
_id: ObjectId(),
author: ObjectId(),
value: <string of post content>
tags: [<trings>],
likes: [array of user ids],
shares: [array of user ids],
comments: [{
ContentId( of the same schema ),
User Id
}]
}
Now, id like to aggregate over the content documents,
and get a result like this.
{
likes: 20,
shares: 10,
comments: 5
}
So, in short, whenever there is a content,
where user id is in likes array,
likes gets incremented by 1.
Same for shares and comments.
Im not sure if this is possible with aggregation framework.
I think not, but maybe some mongodb gurus know better
Quick Edit. Partially based on first post submitted, i made this.
Now, it seems to work, but im sure there is some sort of gotcha, that im missing, since it seems too simple :)
db.getCollection('contents').aggregate( [
{$facet: {
"likes": [
{$match: {likes: ObjectId("596537d6b63edc2318ee9f0c")} },
{$group: {
_id : "$likes",
count: { $sum: 1 }
}},
{ $project: { count: 1} }
],
"shares": [
{$match: {shares: ObjectId("596537d6b63edc2318ee9f0c")} },
{$group: {
_id : "$shares",
count: { $sum: 1 }
}},
{ $project: { count: 1} }
],
"posts": [
{$match: {$and: [
{user: ObjectId("596537d6b63edc2318ee9f0c")},
{parent: {$exists: false} }
]} },
{$group: {
_id : "$_id",
count: { $sum: 1 }
}},
{ $project: { count: 1} }
]
}
}]);
Can you spot something wrong with the code above ?

Answer 1

You can try the below aggregation in 3.4 version.
The below query will $group all the documents and use $in operator to check if the input user_id is in the arrays.
For likes and shares array, use 1 if the user_id is found else is set 0 and $sum to aggregate over all documents.
For comments it is two step process as it is array of arrays.
$grouping step to go over the user_id's in the content document and check input user_id in each element and output 1 if match and else 0.
The comments will have array of array values [[1, 0], [1], [1]] after group stage. Next step is to sum all the values to get the comments count using $reduce operator.
db.collection_name.aggregate([
{
"$group": {
"_id": null,
"likes": {
"$sum": {
"$cond": [
{
"$in": [
user_id,
"$likes"
]
},
1,
0
]
}
},
"shares": {
"$sum": {
"$cond": [
{
"$in": [
user_id,
"$shares"
]
},
1,
0
]
}
},
"comments": {
"$push": {
"$map": {
"input": "$comments",
"as": "comment",
"in": {
"$cond": [
{
"$in": [
user_id,
"$$comment.user_id"
]
},
1,
0
]
}
}
}
}
}
},
{
"$addFields": {
"comments": {
"$reduce": {
"input": "$comments",
"initialValue": 0,
"in": {
"$add": [
"$$value",
{
"$sum": "$$this"
}
]
}
}
}
}
}
])

Answer 2

db.collection.aggregate( [
{
$facet: {
"LikeCategory": [
{ $unwind: "$likes" },
{ $group : {
_id : "$likes",
count: { $sum: 1 }
}
},{ $project : {
userId : "$_id"
_id : 0,
count : 1
}}
],
"ShareCategory": [
{ $unwind: "$shares" },
{ $group : {
_id : "$shares",
count: { $sum: 1 }
}
},{ $project : {
userId : "$_id"
_id : 0,
count : 1
}}
],
"CommentCategory": [
{ $unwind: "$comments" },
{ $group : {
_id : "$comments.userId",
count: { $sum: 1 }
}
},{ $project : {
userId : "$_id"
_id : 0,
count : 1
}}
]
}
}
];
In this way, you will be able to find out the counts. Above code may have some syntax issues, but i hope you will be able to resolve your problem.