Related
I have an aggregate like this :
const files = await File.aggregate([
{
$match: { facilityId: { $in: facilities } }
},
{
$sort: { createdAt: 1 }
},
{
$project: {
file: 0,
}
}
])
And i would like to have each "facility" return only 4 files, i used to do something like facilities.map(id => query(id)) but id like to speed things up in production env.
Using $limit will limit the whole query, that's not what i want, i tried using $slice in the projection stage but got en error :
MongoError: Bad projection specification, cannot include fields or add computed fields during an exclusion projection
how can i achieve that in a single query ?
Schema of the collections is :
const FileStorageSchema = new Schema({
name: { type: String, required: true },
userId: { type: String },
facilityId: { type: String },
patientId: { type: String },
type: { type: String },
accessed: { type: Boolean, default: false, required: true },
file: {
type: String, //
required: true,
set: AES.encrypt,
get: AES.decrypt
},
sent: { type: Boolean, default: false, required: true },
},
{
timestamps: true,
toObject: { getters: true },
toJSON: { getters: true }
})
And i would like to returns all fields except for the file fields that contains the encrypted blob encoded as base64.
Also: i have the feeling that my approach is not correct, what i really would like to get is being able to query all facilityId at once but limited to the 4 latest file created for each facility, i though using an aggregate would help me achieve this but im starting to think it's not how its done.
From the question the schema is not clear. So I have two answers based on two Schemas. Please use what works for you
#1
db.collection.aggregate([
{
$match: {
facilityId: {
$in: [
1,
2
]
}
}
},
{
$group: {
_id: "$facilityId",
files: {
$push: "$file"
}
}
},
{
$project: {
files: {
$slice: [
"$files",
0,
4
],
}
}
}
])
Test Here
#2
db.collection.aggregate([
{
$match: {
facilityId: {
$in: [
1,
2
]
}
}
},
{
$project: {
facilityId: 1,
file: {
$slice: [
"$file",
4
]
}
}
}
])
Test Here
Schema
{
chapter: {
required: true,
type: Schema.Types.ObjectId,
ref: "Chapter",
},
questions: {
type: [Number]
},
};
Here is an example document
{
"_id":{
"$oid":"5ff4b728b6af610f0851d2a6"
},
"chapters":[
{
"chapter":{
"$oid":"611478ab34dde61f28dbe4d3"
},
"questions":[
35,
29,
167,
180,
101,
16,
71,
23
]
},
{
"chapter":{
"$oid":"611478ac34dde61f28dbe4d8"
},
"questions":[
162
]
}
]
}
I want to "$addToSet" on "questions", such as
const someId = SOME_ID;
const chapterId = "611478ac34dde61f28dbe4d8";
const update = {
$addToSet: {
"chapters.$.questions": {
$each: [5, 10, 32, 6],
},
},
};
await model.findOneAndUpdate(
{
_id: someId,
"chapters.chapter": chapterId,
},
update,
{ upsert: true }
)
.lean()
.exec();
This works. However, if there is no document, the "upsert" doesn't create the document.
How can I rewrite the operation so that it can update (addToSet) as well as ensure the document is created if it didn't exist?
I checked MongoDB native query use these
db.con.collection('example').updateOne(
{"chapters": {$elemMatch:{"chapter.id":ObjectId("611478ac34dde61f28dbe4d8")}}},
{$addToSet: {
"chapters.$.questions": {
$each: [5, 10, 32, 6],
},
}},
{upsert: true})
you should find the element of array using elemMatch
{"chapters": {$elemMatch:{"chapter.id":"611478ac34dde61f28dbe4d8"}}}
I figured out, for some reason, I can't $addToSet if the parent object is not present. So I had to make one more operation.
Inspired from this Stackoverflow answer.
I fetch the "chapters" which I need to add.
From this list of fetched chapters, I check which ones exist and which ones don't.
Using the knowledge from point 2, I am using $push to add the chapters which didn't exist entirely, and "adding to set ($addToSet)" questions on the chapters which do exist.
I am posting the code which works for me.
//Data to add (which chapter?: questionNumber[])
const docId = "SOMEID";
const questionsToAdd = {
"611478ab34dde61f28dbe4d3": [1,5,6,10],
"611478ab34dde61f28dbe4d8": [5,8,20,30],
};
//Find the chapters from questionsToAdd which exist
const existingDoc = await model.findOne({
_id: docId,
chapters: { $elemMatch: { chapter: { $in: Object.keys(questionsToAdd) } } },
})
.select(["chapters.chapter"])
.lean()
.exec();
// Objectify the array of chapters
const existingChapters = (existingDoc?.chapters ?? []).map((x) => "" + x.chapter);
// Prepare what to insert, what to update
const updateObject = {
$addToSet: {},
arrayFilters: [],
$push: [],
};
for (const [index, [chapterId, questionIndices]] of Object.entries(questionsToAdd).entries()) {
if (existingChapters.includes(chapterId)) {
updateObject.$addToSet["chapters.$[filter" + index + "].questions"] = { $each: questionIndices };
updateObject.arrayFilters.push({
["filter" + index + ".chapter"]: Types.ObjectId(chapterId),
});
} else {
updateObject.$push.push({
chapter: chapterId,
questions: questionIndices,
});
}
}
if (updateObject.arrayFilters.length) {
// *Add to set the chapters which exist
await model.findOneAndUpdate(
{ _id: userId },
{
$addToSet: updateObject.$addToSet,
},
{
arrayFilters: updateObject.arrayFilters,
upsert: true,
}
)
.lean()
.exec();
}
if (updateObject.$push.length) {
// *Push the ones that does not exist
await model.findOneAndUpdate(
{ _id: userId },
{
$push: { chapters: updateObject.$push },
},
{
upsert: true,
}
)
.lean()
.exec();
}
I am trying to update one document using findOneAndUpdate and $set but I clearly missing something very crucial here because the new request is overwriting old values.
My Device schema looks like this:
{
deviceId: {
type: String,
immutable: true,
required: true,
},
version: {
type: String,
required: true,
},
deviceStatus: {
sensors: [
{
sensorId: {
type: String,
enum: ['value1', 'value2', 'value3'],
},
status: { type: Number, min: -1, max: 2 },
},
],
},
}
And I am trying to update the document using this piece of code:
const deviceId = req.params.deviceId;
Device.findOneAndUpdate(
{ deviceId },
{ $set: req.body },
{},
(err, docs) => {
if (err) {
res.send(err);
} else {
res.send({ success: true });
}
}
);
And when I try to send a request from the postman with the body that contains one or multiple sensors, only the last request is saved in the database.
{
"deviceStatus": {
"sensors": [
{
"sensorId": "test",
"status": 1
}
]
}
}
I would like to be able to update values that are already in the database based on req.body or add new ones if needed. Any help will be appreciated.
The documentation said:
The $set operator replaces the value of a field with the specified
value.
You need the $push operator, it appends a specified value to an array.
Having this documents:
[
{
_id: 1,
"array": [
2,
4,
6
]
},
{
_id: 2,
"array": [
1,
3,
5
]
}
]
Using $set operator:
db.collection.update({
_id: 1
},
{
$set: {
array: 10
}
})
Result:
{
"_id": 1,
"array": 10
}
Using $push operator:
db.collection.update({
_id: 1
},
{
$push: {
array: 10
}
})
Result:
{
"_id": 1,
"array": [
2,
4,
6,
10
]
}
you want to using $push and $set in one findOneAndUpdate, that's impossible, I prefer use findById() and process and save() ,so just try
let result = await Device.findById(deviceId )
//implementation business logic on result
await result.save()
If you want to push new sensors every time you make request then update your code as shown below:
const deviceId = req.params.deviceId;
Device.findOneAndUpdate(
{ deviceId },
{
$push: {
"deviceStatus.sensors": { $each: req.body.sensors }
}
},
{},
(err, docs) => {
if (err) {
res.send(err);
} else {
res.send({ success: true });
}
}
);
Update to the old answer:
If you want to update sensors every time you make request then update your code as shown below:
const deviceId = req.params.deviceId;
Device.findOneAndUpdate(
{ "deviceId": deviceId },
{ "deviceStatus": req.body.sensors },
{ upsert: true },
(err, docs) => {
if (err) {
res.send(err);
} else {
res.send({ success: true });
}
}
);
I am trying to find and update a sub document under another sub document. I am not getting the result as I expect. This is what I currently have setup:
const SiteSchema = new mongoose.Schema({
domain: { type: String, required: true },
keywords: [],
campaigns: [
{
campaign: {
type: mongoose.Schema.Types.ObjectId,
ref: "Campaign",
},
responses: [
{
message: { type: String },
asking_fee: { type: Number },
date: { type: Date },
},
],
}],
})
I would like to find and edit a particular response. Here is the code I have now. I am new to mongoose and MongoDB.
const site = await Site.findOneAndUpdate({
"campaigns.responses._id": responseId, // will it fetch the response ?
}, {
$set: { // I am struggling with the following
"campaigns.$.responses.message": message,
"campaigns.$.responses.asking_price": asking_price,
"campaigns.$.responses.date": date,
},
}
);
If you don't have campaigns.campaign id then you have to use update with aggregation pipeline starting from MongoDB 4.2,
$set to update campaigns field, $map to iterate loop of campaigns array, $map to iterate loop of campaigns.responses array and check condition if responseId match then return updateFields otherwise return old fields and merge with current object using $mergeObjects
let responseId = 1;
let updateFields = {
message: "new message",
asking_fee: 10,
date: new Date()
};
const site = await Site.findOneAndUpdate(
{ "campaigns.responses._id": responseId },
[{
$set: {
campaigns: {
$map: {
input: "$campaigns",
in: {
$mergeObjects: [
"$$this",
{
responses: {
$map: {
input: "$$this.responses",
in: {
$mergeObjects: [
"$$this",
{
$cond: [
{ $eq: ["$$this._id", responseId] },
updateFields,
"$$this"
]
}
]
}
}
}
}
]
}
}
}
}
}]
)
Playground
Second option if you have campaigns.campaign id then you can use $[<identifier>] arrayFilters,
let campaign = 1;
let responseId = 1;
let updateFields = {
message: "new message",
asking_fee: 10,
date: new Date()
};
const site = await Site.findOneAndUpdate({
"campaigns.campaign": campaign,
"campaigns.responses._id": responseId
},
{
$set: {
"campaigns.$[parent].responses.$[child]": updateFields
}
},
{
arrayFilters: [
{ "child._id": responseId },
{ "parent.campaign": campaign }
]
})
Playground
I have schema like this:
this.schema = new Schema({
userEmail: String
environments: [
{
envId: String,
appPreference: String,
language: String,
timeZone: String,
summaryNotificationSchedule: {
timeOfTheDay: String
}
}
]
});
Update request:
{
"envId": "u2",
"appPreference": "put2",
"timeZone": "gmt",
"summaryNotificationSchedule.timeOfTheDay": "32400",
}
As you can see, I am not sending "language": "abc", in the update request and in the result I see the language field is removed. I want to update the fields but not remove the other fields
Mongoose find and update call:
await this.model.findOneAndUpdate({ userEmail, 'environments.envId': envId }, { $set: { 'environments.$': setPreferenceFields } }, { new: true });
You can create update object from your request first:
let request = {
"envId": "u2",
"appPreference": "put2",
"timeZone": "gmt",
"summaryNotificationSchedule.timeOfTheDay": "32400",
};
let update = Object.keys(request).reduce((acc, cur) => {
acc[`environments.$.${cur}`] = request[cur];
return acc;
}, {})
console.log(update);
Then pass it to the update:
await this.model.findOneAndUpdate({ userEmail, 'environments.envId': envId }, { $set: update }, { new: true });
You have to specify property with parent key name of an array, it should be like this way,
await this.model.findOneAndUpdate(
{
userEmail,
'environments.envId': envId
},
{
$set: {
'environments.$.envId': "u2",
'environments.$.appPreference': "put2",
'environments.$.timeZone': "gmt",
'environments.$.summaryNotificationSchedule.timeOfTheDay': "32400"
}
},
{ new: true }
)
Another option, update with aggregation pipeline start from MongoDB v4.2, this little lengthy process then above method,
$map to iterate loop of environments array
$cond check condition if envId is equal to matching envId then merge objects update objects and current objects using $mergeObjects otherwise return current object
await this.model.findOneAndUpdate(
{ userEmail },
[
{
$set: {
environments: {
$map: {
input: "$environments",
in: {
$cond: [
{$eq: ["$$this.envId", envId]}, // add update id
{
$mergeObjects: [
"$$this",
setPreferenceFields // your update fields
]
},
"$$this"
]
}
}
}
}
}
],
{new: true}
)