add where condition in aggregate and group function in mongodb - mongodb

Question

I have mongo model lets say MYLIST containing data like:-
{
"_id" : ObjectId("542139f31284ad1461dbc15f"),
"Category" : "CENTER",
"Name" : "STAND",
"Url" : "center/stand",
"Img" : [ {
"url" : "www.google.com/images",
"main" : "1",
"home" : "1",
"id" : "34faf230-43cf-11e4-8743-311ea2261289"
},
{
"url" : "www.google.com/images1",
"main" : "1",
"home" : "0",
"id" : "34faf230-43cf-11e4-8743-311e66441289"
} ]
}
I execute the following query to the MYLIST collection:
db.MYLIST.aggregate([
{ "$group": {
"_id": "$Category",
"Name": { "$addToSet": {
"name": "$Name",
"url": "$Url",
"img": "$Img"
}}
}},
{ "$sort": { "_id" : 1 } }
]);
And I got the following result -
[
{ _id: 'CENTER',
Name:
[ { "name" : "Stand",
"url" : "center/stand",
"img": { "url" : "www.google.com/images" , "main" : "1", "home" : "1", "id" : "350356a0-43cf-11e4-8743-311ea2261289" }
}]
},
{ _id: 'CENTER',
Name:
[ { "name" : "Stand",
"url" : "center/stand",
"img": { "url" : "www.google.com/images1" , "main" : "1", "home" : "0", "id" : "34faf230-43cf-11e4-8743-311ea2261289" }
}]
}
]
As you can see my img key itself is an array of objects, Hence I am getting multiple entries for the same category of each entry in img array.
What I actually need is to get only those images that have some value for home key.
expected result:-
[
{ _id: 'CENTER',
Name:
[ { "name" : "Stand",
"url" : "center/stand",
"img": { "url" : "www.google.com/images" , "main" : "1", "home" : "1", "id" : "350356a0-43cf-11e4-8743-311ea2261289" }
}]
},
]
Hence I would like to add where the condition for img.home > 0 on the above-mentioned query, Could anybody help me to resolve this issue as my relatively new to MongoDB.

Answer 1

Still really not sure if this is what you want or even why you would be using $addToSet on this grouping. But if all you want to do is "filter" the content of the array returned in your result, then what you want to do is $match the array elements to your condition after processing an $unwind pipeline in order to "de-normalize" the content:
db.MYLIST.aggregate([
// If you only want those matching array members it makes sense to match the
// documents that contain them first
{ "$match": { "Img.home": 1 } },
// Unwind to de-normalize or "un-join" the documents
{ "$unwind": "$Img" },
// Match again to "filter" out those elements that do not match
{ "$match": { "Img.home": 1 } },
// Then do your grouping
{ "$group": {
"_id": "$Category",
"Name": {
"$addToSet": {
"name": "$Name",
"url": "$Url",
"img": "$Img"
}
}
}},
// Finally sort
{ "$sort": { "_id" : 1 } }
]);
So the $match pipeline is the equivalent of a general query or "where clause" in SQL terms, and can be used at any stage. It is usually best to have this as a first stage when there is some type of filtering that results from this. It reduces the overall load by reducing documents to be processed even if "all" of the end results are not removed as would be the case of working with an array.
The $unwind stage allows the array elements to be processed just like another document. And of course you can just use another $match pipeline stage after this in order to just match the documents to your query condition.