Inner array total count in mongodb - mongodb

Question

I have a journal doc. Which contains details of journal_volumes, journal issues and journal articles. I have to list the journals along with the count of volumes, issues and articles in each journal.
Here is my doc:
{
"_id" : ObjectId("5c470fc3135edb4413b0ea24"),
"jnl_code" : "KEG",
"jnl_volumes" : [
{
"name" : "1",
"created_date" : "2019-03-01",
"status" : "0",
"issue_flag" : "0",
"jnl_issues" : [
{
"issue_name" : "1",
"created_date" : "2019-03-04",
"jnl_articles" : [
"test",
"test2"
]
},
{
"issue_name" : "2",
"created_date" : "2019-03-04",
"jnl_articles" : [
"a"
]
},
{
"issue_name" : "3",
"created_date" : "2019-03-04",
"jnl_articles" : [
"b"
]
},
{
"issue_name" : "3",
"created_date" : "2019-03-05",
"jnl_articles" : [
"Q"
]
}
]
},
{
"name" : "2",
"created_date" : "2019-03-01",
"status" : "0",
"issue_flag" : "0",
"jnl_issues" : [
{
"issue_name" : "1",
"created_date" : "2019-03-05",
"jnl_articles" : [
"W"
]
},
{
"issue_name" : "1",
"created_date" : "2019-03-05",
"jnl_articles" : [
"S"
]
},
{
"issue_name" : "1",
"created_date" : "2019-03-05",
"jnl_articles" : [
"R"
]
},
{
"issue_name" : "1",
"created_date" : "2019-03-05",
"jnl_articles" : [
"R"
]
}
]
},
{
"name" : "3",
"created_date" : "2019-03-05",
"status" : "0",
"issue_flag" : "0"
}
]
}
My requirement is to get the count of jnl_volumes, total jnl_issues count and total jnl_articles count in single query..
Thanks to Neil Lunn to redirect me to the similar question (Calculate the count of nested objects with C#
). I referred the answer and wrote a query:
db.getCollection('rvh_journals').aggregate([
{
$project: {
"volumes" : { "$size" : { "$ifNull" : [ "$jnl_volumes", [] ] } },
"issues" : {
"$sum" : {
"$map" : {
"input" : "$jnl_volumes",
"in": { "$size" : { "$ifNull" : [ "$$this.jnl_issues", [] ] } }
}
}
},
"articles" : {
"$sum" : {
"$map" : {
"input" : "$jnl_volumes.jnl_issues.jnl_articles",
"in" : { "$size" : { "$ifNull" : [ "$$this", [] ] } }
}
}
}
}
}
])
This returns with an incorrect article count. Actual article count is 9 but the query returns 8
{
"_id" : ObjectId("5c470fc3135edb4413b0ea24"),
"volumes" : 3,
"issues" : 8,
"articles" : 8
}

Answer 1

Yes I got curious after your edit to the previous question, and noticed your statement was incorrect.
This one is correct:
db.getCollection('rvh_journals').aggregate([
{ "$project": {
"volumes": { "$size": "$jnl_volumes" },
"issues": {
"$sum": {
"$map": {
"input": "$jnl_volumes",
"in": { "$size": { "$ifNull": ["$$this.jnl_issues", [] ] } }
}
}
},
"articles": {
"$sum": {
"$map": {
"input": "$jnl_volumes",
"in": {
"$sum": {
"$map": {
"input": { "$ifNull": [ "$$this.jnl_issues", [] ] },
"in": { "$size": { "$ifNull": [ "$$this.jnl_articles", [] ] } }
}
}
}
}
}
}
}}
])
Returns:
{
"_id" : ObjectId("5c470fc3135edb4413b0ea24"),
"volumes" : 3,
"issues" : 8,
"articles" : 9
}
Note the traversal of the arrays.
You might go and read some of the actual words I used on that original linked answer, because I would have explained that nesting arrays like this is not a good idea. More details on why it's not a good idea and practical approaches to take otherwise are on Updating a Nested Array with MongoDB

Answer 2

enter image description here
.itcount() - is showing distinct count.
Need to show total count of count variable which is declare in $group
Currently it is showing based on company.