I have a collection named 'captures' and the documents within it have the field 'username'
a document looks something like this
/* 1 */
{
"_id" : ObjectId("622b951a026ca3a73f5a2a1c"),
"username" : "andre",
"data" : {
"metadata" : {
"start" : "2022-02-24T09:32:22.390Z",
...
},
...
}
}
/* 2 */
{
"_id" : ObjectId("9255941b026ca3a73f5a2a1c"),
"username" : "andre",
"data" : {
"metadata" : {
"start" : "2022-05-10T03:12:23.440Z",
...
},
...
}
}
/* 3 */
{
"_id" : ObjectId("7775941b026ca3a73f5a2a1c"),
"username" : "bob",
"data" : {
"metadata" : {
"start" : "2022-05-16T12:24:12.002Z",
...
},
...
}
}
/* 4 */
{
"_id" : ObjectId("3215331b026ca3a73f5a2a1c"),
"username" : "bob",
"data" : {
"metadata" : {
"start" : "2022-05-18T12:24:12.002Z",
...
},
...
}
}
I'd like to return a count of documents for each distinct username, where 'start' is after 2022-02-24T09:32:22.390Z
the above example would return something like:
{ "user" : "andre", "count" : 1 }
{ "user" : "bob", "count" : 2 }
I've tried using count, distinct, aggregate without success...
CodePudding user response:
This is pretty simple to do with the aggregation framework:
[
{
$project: {
_id: 0,
user: '$username',
start: {
$toDate: '$data.metadata.start'
}
}
},
{
$match: {
start: {
$gt: Date('2022-02-24T09:32:22.390Z')
}
}
},
{
$group: {
_id: '$user',
user: {
$first: '$user'
},
count: {
$sum: 1
}
}
}
]
By the way you should store dates as Date objects, not strings, it will make your life easier.