I have this schema
module.exports = function(conn, mongoose) {
// var autoIncrement = require('mongoose-auto-increment');
var UsersSchema = new mongoose.Schema({
first_name: String,
last_name:String,
sex: String,
fk_hobbies: []
}
, {
timestamps: true
}, {collection: 'wt_users'});
return conn.model('wt_users', UsersSchema);
};
And for example I have these users in data base
{
"_id" : ObjectId("5aca2ac25c1d8adeb4a2dab0"),
first_name:"Pierro",
last_name:"pierre",
sex:"H",
fk_hobbies: [
{
"_id" : ObjectId("5ac9f84d5c1f8adeb4a2da97"),
"name" : "Art"
},
{
"_id" : ObjectId("5ac9f84d5c8d8adeb4a2da97"),
"name" : "Sport"
},
{
"_id" : ObjectId("5ac9f84d9c1d8adeb4a2da97"),
"name" : "Fete"
},
{
"_id" : ObjectId("5acaf84d5c1d8adeb4a2da97"),
"name" : "Série"
},
{
"_id" : ObjectId("6ac9f84d5c1d8adeb4a2da97"),
"name" : "Jeux vidéo"
}
]
},
{
"_id" : ObjectId("5ac9fa075c1d8adeb4a2da99"),
first_name:"jean",
last_name:"mark",
sex:"H",
fk_hobbies: [
{
"_id" : ObjectId("5ac7f84d5c1d8adeb4a2da97"),
"name" : "Musique"
},
{
"_id" : ObjectId("5ac9f24d5c1d8adeb4a2da97"),
"name" : "Chiller"
},
{
"_id" : ObjectId("5ac9f84c5c1d8adeb4a2da97"),
"name" : "Papoter"
},
{
"_id" : ObjectId("5ac9f84d2c1d8adeb4a2da97"),
"name" : "Manger"
},
{
"_id" : ObjectId("5ac9f84d5c1d8adeb4a2da97"),
"name" : "Film"
}
]
},
{
"_id" : ObjectId("5aca0a635c1d8adeb4a2da9d"),
first_name:"michael",
last_name:"ferrari",
sex:"H",
fk_hobbies: [
{
"_id" : ObjectId("5ac9f84d5c1d8adeb4a2ea97"),
"name" : "fashion"
},
{
"_id" : ObjectId("5ac9f84d5c1e8adeb4a2da97"),
"name" : "Voyage"
},
{
"_id" : ObjectId("5ac9f84c5c1d8adeb4a2da97"),
"name" : "Papoter"
},
{
"_id" : ObjectId("5ac9f84d2c1d8adeb4a2da97"),
"name" : "Manger"
},
{
"_id" : ObjectId("5ac9f84d5c1d8adeb4a2da97"),
"name" : "Film"
}
]
},
{
"_id" : ObjectId("5ac9fa074c1d8adeb4a2da99"),
first_name:"Philip",
last_name:"roi",
sex:"H",
fk_hobbies:
[
{
"_id" : ObjectId("5ac7f84d5c1d8adeb4a2da97"),
"name" : "Musique"
},
{
"_id" : ObjectId("5ac9f24d5c1d8adeb4a2da97"),
"name" : "Chiller"
},
{
"_id" : ObjectId("5ac9f84c5c1d8adeb4a2da97"),
"name" : "Papoter"
},
{
"_id" : ObjectId("5ac9f84d2c1d8adeb4a2da97"),
"name" : "Manger"
},
{
"_id" : ObjectId("5ac9f84d5c1d8adeb4a2da97"),
"name" : "Film"
}
]
}
I want to create a mongoose query that match user getted by id, with others users in database according this :
the query will return firstly the users that have the max number of the same hobbies, that is 5, then the users that have the same 4 hobbies ...
I create a solution fully Javascipt / node js, Is there any query with mongo ?
this is my solution
//var user : the current user that search other similar users : jean mark : 5ac9fa075c1d8adeb4a2da99
//var users : all other users
var tab = []
async.each(users, function(item, next1){
var j = 0;
var hobbies = item["fk_hobbies"]
for(var i = 0; i < 5; i++)
{
var index = hobbies.findIndex(x => x["_id"] == user[0]["fk_hobbies"][i]["_id"].toString());
if(index != -1)
j++
}
if(j != 0)
tab.push({nbHob:j, user:item})
next1()
}, function ()
{
var tab2 = tab.sort(compare)
res.json({success:true, data:tab2})
})
function compare(a,b) {
if (a.nbHob > b.nbHob)
return -1;
if (a.nbHob < b.nbHob)
return 1;
return 0;
}
the displayed result is like this
nbHob : represents the number of similar hobbies
{"success":true,"data":[{"nbHob":5,"user":{"_id":"5ac9fa074c1d8adeb4a2da99","u_first_name":"Akram","u_last_name":"Cherif","u_email":"","u_login":"","u_password":"","u_user_type":0,"u_date_of_birth":"","u_civility":0,"u_sex":"H","u_phone_number":"","u_facebook_id":"","u_google_id":"","u_twitter_id":"","u_profile_image":"","u_about":"","u_profession":"","u_fk_additional_infos":[null],"u_budget":0,"u_address":{"country":"France","state":"Paris","city":"TM","zip":76001},"u_fk_hobbies":[{"name":"Musique","_id":"5ac7f84d5c1d8adeb4a2da97"},{"name":"Chiller","_id":"5ac9f24d5c1d8adeb4a2da97"},{"name":"Papoter","_id":"5ac9f84c5c1d8adeb4a2da97"},{"name":"Manger","_id":"5ac9f84d2c1d8adeb4a2da97"},{"name":"Film","_id":"5ac9f84d5c1d8adeb4a2da97"}]}},{"nbHob":3,"user":{"_id":"5aca0a635c1d8adeb4a2da9d","u_first_name":"Chawki","u_last_name":"Gasmi","u_email":"","u_login":"","u_password":"","u_user_type":0,"u_date_of_birth":"","u_civility":0,"u_sex":"H","u_phone_number":"","u_facebook_id":"","u_google_id":"","u_twitter_id":"","u_profile_image":"","u_about":"","u_profession":"","u_fk_additional_infos":[null],"u_budget":{"min":500,"max":850},"u_address":{"country":"","state":"","city":"","zip":0},"u_fk_hobbies":[{"name":"fashion","_id":"5ac9f84d5c1d8adeb4a2ea97"},{"name":"Voyage","_id":"5ac9f84d5c1e8adeb4a2da97"},{"name":"Papoter","_id":"5ac9f84c5c1d8adeb4a2da97"},{"name":"Manger","_id":"5ac9f84d2c1d8adeb4a2da97"},{"name":"Film","_id":"5ac9f84d5c1d8adeb4a2da97"}]}}]}
Your question data seems a bit messed up due to probably far to liberal copy/paste since every hobby has the same ObjectId value. But I can correct that with a full self contained example:
const { Schema } = mongoose = require('mongoose');
const uri = 'mongodb://localhost/people';
mongoose.Promise = global.Promise;
mongoose.set('debug', true);
const hobbySchema = new Schema({
name: String
});
const userSchema = new Schema({
first_name: String,
last_name: String,
sex: String,
fk_hobbies: [hobbySchema]
});
const Hobby = mongoose.model('Hobby', hobbySchema)
const User = mongoose.model('User', userSchema);
const userData = [
{
"first_name" : "Pierro",
"last_name" : "pierre",
"sex" : "H",
"fk_hobbies" : [
"Art", "Sport", "Fete", "Série", "Jeux vidéo"
]
},
{
"first_name": "jean",
"last_name" : "mark",
"sex" : "H",
"fk_hobbies" : [
"Musique", "Chiller", "Papoter", "Manger", "Film"
]
},
{
"first_name" : "michael",
"last_name" : "ferrari",
"sex" : "H",
"fk_hobbies" : [
"fashion", "Voyage", "Papoter", "Manger", "Film"
]
},
{
"first_name" : "Philip",
"last_name" : "roi",
"sex" : "H",
"fk_hobbies" : [
"Musique", "Chiller", "Papoter", "Manger", "Film"
]
}
];
const log = data => console.log(JSON.stringify(data, undefined, 2));
(async function() {
try {
const conn = await mongoose.connect(uri);
await Promise.all(
Object.entries(conn.models).map(([k,m]) => m.remove())
);
const hobbies = await Hobby.insertMany(
[
...userData
.reduce((o, u) => [ ...o, ...u.fk_hobbies ], [])
.reduce((o, u) => o.set(u,1) , new Map())
]
.map(([name,v]) => ({ name }))
);
const users = await User.insertMany(userData.map(u =>
({
...u,
fk_hobbies: u.fk_hobbies.map(f => hobbies.find(h => f === h.name))
})
));
let user = await User.findOne({
"first_name" : "Philip",
"last_name" : "roi"
});
let user_hobbies = user.fk_hobbies.map(h => h._id );
let result = await User.aggregate([
{ "$match": {
"_id": { "$ne": user._id },
"fk_hobbies._id": { "$in": user_hobbies }
}},
{ "$addFields": {
"numHobbies": {
"$size": {
"$setIntersection": [
"$fk_hobbies._id",
user_hobbies
]
}
},
"fk_hobbies": {
"$map": {
"input": "$fk_hobbies",
"in": {
"$mergeObjects": [
"$$this",
{
"shared": {
"$cond": {
"if": { "$in": [ "$$this._id", user_hobbies ] },
"then": true,
"else": "$$REMOVE"
}
}
}
]
}
}
}
}},
{ "$sort": { "numHobbies": -1 } }
]);
log(result);
mongoose.disconnect();
} catch(e) {
} finally {
process.exit();
}
})()
Most of that is just "setup" to re-create the data set, but simply put we're just adding the users and their hobbies and keeping a "unique" identifier for each "unique hobby" by name. This is probably what you actually meant in the question, and it's the sort of model you should be following.
The interesting part is all in the .aggregate() statement, which is how we "query" then "count" the matching hobbies and enable the "server" to sort the results before returning to the client.
Given a current user ( and the last one in the list you included has the most interesting matches ), we then focus on this section of the code:
// Simulates getting the current user to compare against
let user = await User.findOne({
"first_name" : "Philip",
"last_name" : "roi"
});
// Just get the list of _id values from the current user for reference
let user_hobbies = user.fk_hobbies.map(h => h._id );
let result = await User.aggregate([
// Find all users not the current user with at least one of the hobbies
{ "$match": {
"_id": { "$ne": user._id },
"fk_hobbies._id": { "$in": user_hobbies }
}},
// Add the count of matches, "optionally" we are marking the matched
// hobbies in the array as well.
{ "$addFields": {
"numHobbies": {
"$size": {
"$setIntersection": [
"$fk_hobbies._id",
user_hobbies
]
}
},
"fk_hobbies": {
"$map": {
"input": "$fk_hobbies",
"in": {
"$mergeObjects": [
"$$this",
{
"shared": {
"$cond": {
"if": { "$in": [ "$$this._id", user_hobbies ] },
"then": true,
"else": "$$REMOVE"
}
}
}
]
}
}
}
}},
// Sort the results by the "most" hobbies, which is "descending" order
{ "$sort": { "numHobbies": -1 } }
]);
I've commented those steps for you but let's expand on that.
Firstly we presume you have the current user already returned from the database by whatever means you have already done. For the purposes of the rest of the operations, all your really need from that user is the _id of the "User" itself and of course the _id values from each of that user's chosen hobbies. We can do a quick .map() operation as it shown here, but we keep a copy for ease of reference and not repeating that through the remaining code.
Then we get to the actual aggregate statement. The first condition there is the $match, this works like a standard query expression with all the same operators. We want two things from these query conditions:
Get all users except the current user for consideration;
AND where those users contain at least one match on the same hobbies, by _id value.
So the condition for "everyone else" is essentially to supply the $ne "not equal to" operator in argument to the _id value, comparing of course to the current user _id. The second condition to get only those with the same hobbies uses the $in operator against the _id field of the fk_hobbies array. In MongoDB query parlance we denote this as "$fk_hobbies._id" in order to match against the "inner" _id property values.
The $in operator itself takes a "list" as it's argument and compares each value in the list supplied to the property the condition is assigned to. MongoDB itself does not care that fk_hobbies is an array or a single value, and will simply look for an match for anything in the provided list. Think of $in as a short way of writing $or, except you don't need to explicitly include the same property name on every condition.
Now you have the correct documents selected and have discarded any users who do not share any of the same hobbies we can move on to the next stage. Note also that the whole $match considers it logical that you only want those "matching" users. If you actually wanted to see "all users" including those with "no matches", then you can simply omit the whole $match pipeline stage. Your code is discarding anything that was not counted, so this code simply doesn't bother to count anything which "must" have a 0 count.
The $addFields stage pipeline stage is a quick way to "add new fields" to the document returned in results. The main output you want here is the "numHobbies" in addition to the other user details, so this pipeline stage operator is the optimal way to do this, but if you're MongoDB server is a bit older then you can simply specify "all" fields you want to include in addition to any new ones using $project instead.
In order to "count" the number of hobbies in common we essentially use two aggregation operators, which are $setIntersection and $size. Both of these should be available in an MongoDB version you really should be using in production.
In respective order the $setIntersection operator "compares sets" which is in this case the list of _id values within fk_hobbies, both from the current selected user we stored earlier and from the present document being considered in the expression. The result from this operator is the list of values which are the "same" between both lists.
Naturally the $size operator looks at the returned list ( or set ) from $setIntersection and returns the number of entries in that list. This of course is the "matched count".
The next part involves projecting a "re-written" form of the fk_hobbies array. This is totally optional and by my own design for demonstration purposes. "If" you wanted to do what I am doing here as well, then what this bit of code does is adds an additional property to the objects of the fk_hobbies array to indicate where that particular hobby was one of those which matched the list.
I'm saying this is "optional" because I'm actually demonstrating two features available for MongoDB 3.6 only. These involve the usage of $mergeObjects on the inner array elements and the usage of Conditionally Exlcuding Fields.
Stepping through that, since fk_hobbies is an array we need to use the $map operator in order to "reshape" the objects inside it. This operator allows us to process each array member and return a new value based on the transformations we include as it's argument. It's usage is much the same as .map() for JavaScript or any other language which implements a similar operation.
Therefore for each object in the array ( $$this ) we apply the $mergeObjects operator which will "merge" the result of it's arguments. These are provided as the $$this for the current object as it already is, and the second argument in the expression which is doing something new and interesting.
Here we use the $cond operator, which is a "ternary" operator ( or if..then..else expression ) which considers a condition if and then returns either the then argument where that expression was true, or the else expression where it was false. The expression here is another form of $in used as an aggregation expression. In this form the first argument is a singular value $$this._id which will be compared to a list expression in the second argument. That second argument is of course the list of the current user hobby id's we kept earlier, and are using again for comparison.
That usage of $in alone would return either true or false where it was a match. But the extra demonstrated action here is that within the $cond expresion, our else condition for false returns the new and special $$REMOVE value. What this means is that with our "shared" property we are adding to each object in the array, rather than assigning it a value of false where there was no match, we actually don't include that property in the output document at all.
That "optional" part is really just there as a "nice touch" to indicate which "hobbies" were matched in the conditions, rather than simply returning the count. If you like it then use it, and if you don't have MongoDB 3.6 with those features you can simply do that same alteration in the returned documents from the aggregation output anyway:
let result = await User.aggregate([
{ "$match": {
"_id": { "$ne": user._id },
"fk_hobbies._id": { "$in": user_hobbies }
}},
{ "$addFields": {
"numHobbies": {
"$size": {
"$setIntersection": [
"$fk_hobbies._id",
user_hobbies
]
}
}
}},
{ "$sort": { "numHobbies": -1 } }
]);
// map each result after return
result = result.map(r =>
({
...r,
fk_hobbies: r.fk_hobbies.map(h =>
({
...h,
...(( user_hobbies.map(i => i.toString() ).indexOf( h._id.toString() ) != -1 )
? { "shared": true } : {} )
})
)
})
)
Either way, the main thing you wanted out of any $addFields or $project statement was the actual "numHobbies" value indicating the count. And the main reason we did that on the server was so that we can also $sort on the server, which would in turn allow you to add things like $limit and $skip to larger result sets for purposes of paging where it simply would not be practical to get all the results from the collection, even if they were filtered in the initial match or regular query.
Anyhow, from the small sample of documents in the question as also generated in the sample listing, we get a result like this:
[
{
"_id": "5ad6bbe63365bc3428feed8a",
"first_name": "jean",
"last_name": "mark",
"sex": "H",
"fk_hobbies": [
{
"_id": "5ad6bbe63365bc3428feed7d",
"name": "Musique",
"__v": 0,
"shared": true
},
{
"_id": "5ad6bbe63365bc3428feed7e",
"name": "Chiller",
"__v": 0,
"shared": true
},
{
"_id": "5ad6bbe63365bc3428feed7f",
"name": "Papoter",
"__v": 0,
"shared": true
},
{
"_id": "5ad6bbe63365bc3428feed80",
"name": "Manger",
"__v": 0,
"shared": true
},
{
"_id": "5ad6bbe63365bc3428feed81",
"name": "Film",
"__v": 0,
"shared": true
}
],
"__v": 0,
"numHobbies": 5
},
{
"_id": "5ad6bbe63365bc3428feed90",
"first_name": "michael",
"last_name": "ferrari",
"sex": "H",
"fk_hobbies": [
{
"_id": "5ad6bbe63365bc3428feed82",
"name": "fashion",
"__v": 0
},
{
"_id": "5ad6bbe63365bc3428feed83",
"name": "Voyage",
"__v": 0
},
{
"_id": "5ad6bbe63365bc3428feed7f",
"name": "Papoter",
"__v": 0,
"shared": true
},
{
"_id": "5ad6bbe63365bc3428feed80",
"name": "Manger",
"__v": 0,
"shared": true
},
{
"_id": "5ad6bbe63365bc3428feed81",
"name": "Film",
"__v": 0,
"shared": true
}
],
"__v": 0,
"numHobbies": 3
}
]
So there are two users that were returned and we counted the matching hobbies as 5 and 3 respectively and returned the one with the most matched first. You can also see the addition of the "shared" property on each of the matched hobbies to indicate which of the hobbies in each of the returned users lists were also shared with the original user they were compared with.
NOTE: You were probably just "trying things" but your usage of async.each() in your question was not really necessary since none of the inner code is actually "async" itself. Even in the listing here, the only thing you actually need to "await" as an async call after you have the current user to compare is the .aggregate() response itself.
So if at any part of this you were presuming you would be "awaiting requests within a loop", then you were mistaken. Simply ask the database for the results and await their return.
One request to the database is all that is required.
N.B It's also 2018, so you really should start to understand Promises and usage of async/await with them. The code is much cleaner that way and surely any newly developed application should be running in an environment with this support. So "callback helper" libraries like "node async", are a little "old hat" and outmoded in a modern context.
MongoDB selectors become quickly complicated, especially when you come from mySQL using JOIN and other fancy keywords. I did my best to make the title of this question as clear as possible, but failed miserably.
As an example, let a MongoDB collection have the following schema for its documents:
{
_id : int
products : [
{
qte : int
status : string
},
{
qte : int
status : string
},
{
qte : int
status : string
},
...
]
}
I'm trying to run a db.collection.find({ }) query returning documents where all products do not have the string "finished" as status. Please note that the products array has a variable length.
We could also say we want all documents that has at least one product with a status that is not "finished".
If I were to run it as a Javascript loop, we would have something like the following :
// Will contain queried documents
var matches = new Array();
// The documents variable contains all documents of the collection
for (var i = 0, len = documents.length; i < len; i++) {
var match = false;
if (documents[i].products && documents[i].products.length !== 0) {
for (var j = 0; j < documents[i].products; j++) {
if (documents[i].products[j].status !== "finished") {
match = true;
break;
}
}
}
if (match) {
matches.push(documents[i]);
}
}
// The previous snippet was coded directly in the Stack Overflow textarea; I might have done nasty typos.
The matches array would contain the documents I'm looking for. Now, I wish there would be a way of doing something similar to collection.find({"products.$.status" : {"$ne":"finished"}}) but MongoDB hates my face when I do so.
Also, documents that do not have any products need to be ignored, but I already figured this one out with a $and clause. Please note that I need the ENTIRE document to be returned, not just the product array. If a document has products that are not "finished", then the entire document should be present. If a document has all of its products set at "finished", the document is not returned at all.
MongoDB Version: 3.2.4
Example
Let's say we have a collection that contains three documents.
This one would match because one of the status is not "finished".
{
_id : 1,
products : [
{
qte : 10,
status : "finished"
},
{
qte : 21,
status : "ongoing"
},
]
}
This would not match because all statuses are set to "finished"
{
_id : 2,
products : [
{
qte : 35,
status : "finished"
},
{
qte : 210,
status : "finished"
},
{
qte : 2,
status : "finished"
},
]
}
This would also not match because there are no products. It would also not match if the products field was undefined.
{
_id : 3,
products : []
}
Again, if we ran the query in a collection that had the three documents in this example, the output would be:
[
{
_id : 1,
products : [
{
qte : 10,
status : "finished"
},
{
qte : 21,
status : "ongoing"
},
]
}
]
Only the first document gets returned because it has at least one product that doesn't have a status of "finished", but the last two did not make the cut since they either have all their products' statuses set as "finished", or don't have any products at all.
Try following query. It's fetching documents where status is not equals to "finished"
Note: This query will work with MongoDB 3.2+ only
db.collection.aggregate([
{
$project:{
"projectid" : 1,
"campname" : 1,
"campstatus" : 1,
"clientid" : 1,
"paymentreq" : 1,
products:{
$filter:{
input:"$products",
as: "product",
cond:{$ne: ["$$product.status", "finished"]}
}
}
}
},
{
$match:{"products":{$gt: [0, {$size:"products"}]}}
}
])
You need .aggregate() rather than .find() here. That is the only way to determine if ALL elements actually don't contain what you want:
// Sample data
db.products.insertMany([
{ "products": [
{ "qte": 1 },
{ "status": "finished" },
{ "status": "working" }
]},
{ "products": [
{ "qte": 2 },
{ "status": "working" },
{ "status": "other" }
]}
])
Then the aggregate operation with $redact:
db.products.aggregate([
{ "$redact": {
"$cond": {
"if": {
"$anyElementTrue": [
{ "$map": {
"input": "$products",
"as": "product",
"in": {
"$eq": [ "$$product.status", "finshed" ]
}
}}
]
},
"then": "$$PRUNE",
"else": "$$KEEP"
}
}}
])
Or alternately you can use the poorer and slower cousin with $where
db.products.find(function(){
return !this.products.some(function(product){
return product.status == "finished"
})
})
Both return just the one sample document:
{
"_id" : ObjectId("56fb4791ae26432047413455"),
"products" : [
{
"qte" : 2
},
{
"status" : "working"
},
{
"status" : "other"
}
]
}
So the $anyElementTrue with the $map input or the .some() are basically doing the same thing here and evaluating if there was any match at all. You use the "negative" assertion to "exclude" documents that actually find a match.
I have a mongodb collection which has documents like this :
{
"_id" : ObjectId("safdsd435tdg54trgds"),
"startDate" : ISODate("2013-07-02T17:35:01.000Z"),
"endDate" : ISODate("2013-08-02T17:35:01.000Z"),
"active" : true,
"channels" : [
1, 2, 3, 4
],
}
I want to convert this to something like this :
{
"_id" : ObjectId("safdsd435tdg54trgds"),
"startDate" : ISODate("2013-07-02T17:35:01.000Z"),
"endDate" : ISODate("2013-08-02T17:35:01.000Z"),
"active" : true,
"channels" : [
1, 2, 3, 4
],
"tags" :[
{
"name": one
"type": channel
},
{
"name": two
"type": channel
},
{
"name": three
"type": channel
},
{
"name": four
"type": channel
}
]
}
I already have a mapping of what 1,2,3,4 mean. Just for the sake of simplicity I put them as their alphabetical format. the values could be different, but they're static mappings.
You seem to be trying to do this update without a big iteration of your collection, So you "could" do this with mapReduce, albeit in a very "mapReduce way" as it has it's own way of doing things.
So first you want to define a mapper that encapsulates your current document :
var mapFunction = function (){
var key = this._id;
var value = {
startDate: this.startDate,
endDate: this.endDate,
active: this.active,
channels: this.channels
};
emit( key, value );
};
Now here the reducer is actually not going to be called as all the keys from the mapper will be unique, being of course the _id values from the original document. But to make the call happy:
var reduceFunction = function(){};
As this is a one to one thing this will go to finalize. It could be in the mapper, but for cleanliness sake
var finalizeFunction = function (key, reducedValue) {
var tags = [
{ name: "one", type: "channel" },
{ name: "two", type: "channel" },
{ name: "three", type: "channel" },
{ name: "four", type: "channel" }
];
reducedValue.tags = [];
reducedValue.channels.forEach(function(channel) {
reducedValue.tags.push( tags[ channel -1 ] );
});
return reducedValue;
};
Then call the mapReduce:
db.docs.mapReduce(
mapFunction,
reduceFunction,
{
out: { replace: "newdocs" },
finalize: finalizeFunction
}
)
So that will output to a new collection, but in the way that mapReduce does it so you have this:
{
"_id" : ObjectId("53112b2d0ceb66905ae41259"),
"value" : {
"startDate" : ISODate("2013-07-02T17:35:01Z"),
"endDate" : ISODate("2013-08-02T17:35:01Z"),
"active" : true,
"channels" : [ 1, 2, 3, 4 ],
"tags" : [
{
"name" : "one",
"type" : "channel"
},
{
"name" : "two",
"type" : "channel"
},
{
"name" : "three",
"type" : "channel"
},
{
"name" : "four",
"type" : "channel"
}
]
}
}
So all your document fields other than _id are stuck under that value field, so that's not the document that you want. But that is how mapReduce works.
If you really need to get out of jail from this and are willing to wait a bit, the upcoming 2.6 release has added an $out pipeline stage. So you "could" transform the documents in your new collection with $project like this:
db.newdocs.aggregate([
// Transform the document
{"$project": {
"startDate": "$value.startDate",
"endDate": "$value.endDate",
"active": "$value.active",
"channels": "$value.channels",
"tags": "$value.tags"
}},
// Output to new collection
{"$out": "fixeddocs" }
])
So that will be right. But of course this is not your original collection. So to back to that state you are going to have to .drop() collections and use .renameCollection() :
db.newdocs.drop();
db.docs.drop();
db.fixeddocs.renameCollection("docs");
Now please READ the documentation carefully on this, there are several limitations, and of course you would have to re-create indexes as well.
All of this, and in particular the last stage is going to result in a lot of disk thrashing and also keep in mind that you are dropping collections here. It almost certainly is a case for taking access to your database off-line while this is performed.
And even as such the dangers here are real enough that perhaps you can just live with running an iterative loop to update the documents, using arbitrary JavaScript. And if you really must have to do so, you could always do that using db.eval() to have that all execute on the server. But if you do, then please read the documentation for that very carefully as well.
But for completeness even if I'm not advocating this:
db.eval(function(){
db.docs.find().forEach(function(document) {
var tags = [
{ name: "one", type: "channel" },
{ name: "two", type: "channel" },
{ name: "three", type: "channel" },
{ name: "four", type: "channel" }
];
document.tags = [];
document.channels.forEach(function(channel) {
document.tags.push( tags[ channel -1 ] );
});
var id = document._id;
delete document._id;
db.docs.update({ "_id": id },document);
});
})