How sort according to given order and calculation 1st document with second one in MongoDB? - mongodb

Question

I want to get which members not present when state change and which member is present.
My given array for state order like:
[{order:1,text:'CS'}, {order:2,text:'IP'}, {order:3,text:'AC'}]
so I want to sort according to this array want to perform some operation with each pare of documents
My documents like:
{
"count" : 2,
"state" : "CS",
"members" : [
{
"email" : "builuu1998#gmail.com",
"date" : ISODate("2016-12-24T03:39:05.720Z")
},
{
"email" : "bactv.hn#gmail.com",
"date" : ISODate("2016-12-25T02:32:48.698Z")
}
]
},
{
"count" : 1,
"state" : "AC",
"members" : [
{
"email" : "builuu1998#gmail.com",
"date" : ISODate("2016-12-24T03:39:05.720Z")
}
]
},
{
"count" : 3,
"state" : "IP",
"members" : [
{
"email" : "builuu1998#gmail.com",
"date" : ISODate("2016-12-24T03:39:05.720Z")
},
{
"email" : "bactv.hn#gmail.com",
"date" : ISODate("2016-12-25T02:32:48.698Z")
},
{
"email" : "abc.hn#gmail.com",
"date" : ISODate("2016-12-25T02:32:48.698Z")
}
]
}
So I want to know which members are not present from one state to another state and which members are present .
In my exam 1st state to 2nd state means (CS - IP): present 2 members and not present 1 member
{
"state": "CS_IP",
"present": [
{
"email" : "builuu1998#gmail.com",
"date" : ISODate("2016-12-24T03:39:05.720Z")
},
{
"email" : "bactv.hn#gmail.com",
"date" : ISODate("2016-12-25T02:32:48.698Z")
}
],
"notPresent": [
{
"email" : "abc.hn#gmail.com",
"date" : ISODate("2016-12-25T02:32:48.698Z")
}
]
}
and 2nd state to 3rd state means (IP - AC): present 1 members and not present 2 member
{"state": "IP_AC",
"present": [
{
"email" : "builuu1998#gmail.com",
"date" : ISODate("2016-12-24T03:39:05.720Z")
}
],
"notPresent": [
{
"email" : "bactv.hn#gmail.com",
"date" : ISODate("2016-12-25T02:32:48.698Z")
},
{
"email" : "abc.hn#gmail.com",
"date" : ISODate("2016-12-25T02:32:48.698Z")
}
]}
How can I achieve this using aggregate query because I need some aggregate operation after this stage complete

Answer 1

You can perform one aggregation request per "couple" of state like this for ["IP", "AC"] :
db.exams.aggregate([{
$match: {
"state": {
$in: ["IP", "AC"]
}
}
}, {
"$group": {
"_id": 1,
"group1": { "$first": "$members" },
"group2": { "$last": "$members" }
}
}, {
"$project": {
"present": { "$setIntersection": ["$group1", "$group2"] },
"notPresent": {
$setUnion: [
{ "$setDifference": ["$group1", "$group2"] },
{ "$setDifference": ["$group2", "$group1"] }
]
}
}
}])
Note that this will only work for 2 elements to match (here ["IP", "AC"]) since we have to create two new fields group1 & group2 to $setIntersection and $setDifference (as there is no $group with $setIntersection and $setDifference)
The query above will give :
{
"_id": 1,
"present": [
{ "email": "builuu1998#gmail.com", "date": ISODate("2016-12-24T03:39:05.720Z") }
],
"notPresent": [
{ "email": "abc.hn#gmail.com", "date": ISODate("2016-12-25T02:32:48.698Z") },
{ "email": "bactv.hn#gmail.com", "date": ISODate("2016-12-25T02:32:48.698Z") }
]
}