I have the current mongo collection and need to find and specific id inside an array and update its value.
{
"_id" : ObjectId("111fe6813abdeb1505f5111"),
"objectIdUser" : ObjectId("111fe6813abdeb1505f5111"),
"objectIdClasses" : [
{
"objectIdClass" : ObjectId("111fe6813abdeb1505f5111")
}
],
}
I tried to use the following query but It never updates.
db.classes.findOneAndUpdate(
{
objectIdUser: ObjectId("111fe6813abdeb1505f5111"),
"objectIdClasses.objectIdClass": ObjectId('111fe6813abdeb1505f5111')
},
{
$set:
{
"objectIdClasses.$.objectIdClass": ObjectId('111fe6813abdeb1505f5111')
}
}
);
The records exists, but I always get a null value and the value is never updated.
Follow the below steps carefully;
db.<collectionName>.updateOne(filter, updateDoc, options);
the filter go defines the document you want to change, just like the find.
the updateDoc defines the new changes
the options instruct the method to create a document if no documents match the filter
Example:
I have a database of shop collection with the following data in a products collection
{ "_id" : 1, "name" : "Pen", "price" : 1.2 }
If I want to change the "price" to of "Pen" to 5.0 then:
db.products.updateOne({name: "Pen"}, {$set: {price: 5.0}}, {upsert: true});
Note: all your changes should be inside the <$set> object
Related
I'm fairly new to mongodb so please bear with me.
As per title, what I want to achieve is to convert a specific field in all documents within an array of a document from String to Int how do i do that?
Sample Doc :
{
reviews:[
{
snid:"1242"
},
{
snid:"8392"
}
]
}
And my objective is to convert all of the snid's from String to Int32
so far i understand that we can use something like db.collection.update() but this will update a specific field, not an array.
Another attempt is
db.collection.find({},{reviews:1,_id:0},(err,doc)=>{
//How do i push it back to the document
})
But as you can tell, I'm not entirely sure on how we should push the updated document back into the same array of sorts.
Any insights will be greatly appreciated!
1) If you're using MongoDB version >= 4.2, try below query :
db.collection.update({'reviews.snid' : {$exists : true}}, [
{
$set: {
reviews: {
$map: { input: "$reviews", in: { 'snid': { $toInt: "$$this.snid" } } }
}
}
}
],{multi :true})
Above query uses Aggregation-pipeline in .update() which was introduced in version 4.2, You can also use .updateMany() instead of .update().
It works on documents of below type :
/* 1 */
{
"_id" : ObjectId("5e810f5ec16b5679b43a2f0e"),
"reviews" : [
{
"snid" : '1242'
},
{
"snid" : '8392'
}
]
}
/* 2 */
{
"_id" : ObjectId("5e810f6ac16b5679b43a310c"),
"reviews" : [
{
"snid" : '1242232'
},
{
"snid" : '8391232'
}
]
}
/* 3 */
{
"_id" : ObjectId("5e8110b1c16b5679b43a5148"),
"abc" : 1
}
/* 4 */
{
"_id" : ObjectId("5e8110c3c16b5679b43a52f9"),
"reviews" : []
}
/* 5 */
{
"_id" : ObjectId("5e811359c16b5679b43a9229"),
"reviews" : [
{
"abc" : "1"
}
]
}
But above update query will partially work if you've a doc like below :
{
"_id" : ObjectId("5e811359c16b5679b43a9230"),
"reviews" : [
{
"abc" : "1"
},
{
"snid" : "123"
}
]
}
In that case you need to use $cond to do a conditional check in $map to see if current object has key snid then convert value or else pass on the same object as is to 'reviews' array.
2) Just in Case if your MongoDB version is < 4.2/4.0 & > 3.2 - You can use .bulkWrite() :
Cause you can not use Aggregation Pipeline in update & also $toInt. So you need to do .find() to get entire docs & write code to convert these from strings to integers & use .bulkWrite() to update docs in one update DB call (You can take _id as key for each document).
3) You can also write an aggregation query on existing collection & use $out to update entire collection or write aggregation result to new collection by running just one query. I would prefer to temporarily write it to new collection to check data is correct & rename new collection to what ever is existing by naming existing with something ends with _backup used as backup.
My node.js application will insert a sub-document into a nested sub-document array field of the following MongoDB document, and I need to determine the ID of the newly inserted sub-document:
{
"_id" : ObjectId("578d5a52cc13117022e09def"),
"name" : "Grade 5 - Section A",
"scores" : [{
"studentId" : ObjectId("5776bd36ffc8227405d364d2"),
"performance" : [{
"_id" : ObjectId("57969b8fc164a21c20698261"),
"subjectId" : ObjectId("577694ecbf6f3a781759c54a"),
"score" : 86,
"maximum" : 100,
"grade" : "B+"
}]
}]
}
The sub-document looks like this:
{
"subjectId" : ObjectId("5776ffe1804540e29c602a62"),
"score" : 74,
"maximum" : 100,
"grade" : "A-"
}
I am adding the sub-document using the following Mongoose code:
Class.update({
_id: '578d5a52cc13117022e09def',
'scores.studentId': '5776bd36ffc8227405d364d2'
}, {
$addToSet: {
'scores.$.performance': {
'subjectId' : '5776ffe1804540e29c602a62',
'score' : 74,
'maximum' : 100,
'grade' : 'A-'
}
}
}, function(err, result) {
if (err) {
throw err;
}
console.log(result);
});
The subject sub-document gets added in the performance sub-document array which is itself nested in the scores sub-document array. Notice that the newly inserted sub-document is assigned with its own ID, as instituted by the defined schema. Even if I get back the entire document, that's not very helpful. I specifically need the ID of that newly inserted sub-document. What is the recommended approach to this problem?
In this case I prefer pre-assign the ID to the sub-document (i.e. sub._id = ObjectId() or use uuid package if you prefer uuid): is clear and predictable.
Also remember that if you frequent query by a subdoc id is good to add (using ensureIndex()) an index for this use case in the collection.
There is a good solution for that, try to use the methods create and push of the MongooseArrays.
In your code you could return the Student and do something like this:
const newPerformance = student.performance.create(newData);
student.performance.push(newPerformance);
const updatedStudent = await student.save();
if (updatedStudent) return newPerformance._id;
This is just a simple example.
Using the create method of MongooseArrays, link for doc, the mongoose will create an _id and do all the validations and casts it needs, so if the save process is fine the created subdocument you could just use the _id of the subdocument you got with the create method.
I use addToSet command (to add UpdatedData):
MyTable.update({ "UrlLink": "http://www.someurl.com"}, {
$addToSet: {
UpdatedData: {
TheDate: ThisDate,
NewInfo: ThisInfo
}
}
},
function (err, result) {
next(err,result)
}
);
But then, when I do the query (after UpdatedData is added to my document), I see that it returns two documents. Instead of only one updated document:
db.MyTable.find( {"UrlLink": "http://www.someurl.com"})
{ "_id" : ObjectId("56485922143a886f66c2f8bb"), "UrlLink" : "http://www.someurl.com", "Stuff" : "One", "UpdatedData" : [ { "TheDate" : "11/15/2015", "NewInfo" : "Info1", "_id" : ObjectId("5648599a71efc79c660f76d3") } ] }
{ "_id" : ObjectId("5648599f71efc79c660f76d4"), "UrlLink" : "http://www.someurl.com", "Stuff": "One", "UpdatedData" : [ ] }
So it seems that addToSet creates a new document with new _id, instead of updating the old record (ObjectId("5648599f71efc79c660f76d4")). But I only see the updated document in robomongo (ObjectId("56485922143a886f66c2f8bb")). Any ideas why this happens and how I could prevent that behaviour?
update cannot create a new document, it can only update existing.
This looks like you have created two documents with the same url.. Then when you update it just updates the first one..
To prevent the creation of a document with an already existent url you can create an index and set it to unique
db.collection.createIndex({ UrlLink: 1 }, { unique: true })
This will prevent creation of new documents with the same url, and it will also make queries by UrlLink as fast as possible.
I have a document with a field containing a very long string. I need to concatenate another string to the end of the string already contained in the field.
The way I do it now is that, from Java, I fetch the document, extract the string in the field, append the string to the end and finally update the document with the new string.
The problem: The string contained in the field is very long, which means that it takes time and resources to retrieve and work with this string in Java. Furthermore, this is an operation that is done several times per second.
My question: Is there a way to concatenate a string to an existing field, without having to fetch (db.<doc>.find()) the contents of the field first? In reality all I want is (field.contents += new_string).
I already made this work using Javascript and eval, but as I found out, MongoDB locks the database when it executes javascript, which makes the overall application even slower.
Starting Mongo 4.2, db.collection.updateMany() can accept an aggregation pipeline, finally allowing the update of a field based on its current value:
// { a: "Hello" }
db.collection.updateMany(
{},
[{ $set: { a: { $concat: [ "$a", "World" ] } } }]
)
// { a: "HelloWorld" }
The first part {} is the match query, filtering which documents to update (in this case all documents).
The second part [{ $set: { a: { $concat: [ "$a", "World" ] } } }] is the update aggregation pipeline (note the squared brackets signifying the use of an aggregation pipeline). $set (alias of $addFields) is a new aggregation operator which in this case replaces the field's value (by concatenating a itself with the suffix "World"). Note how a is modified directly based on its own value ($a).
For example (it's append to the start, the same story ):
before
{ "_id" : ObjectId("56993251e843bb7e0447829d"), "name" : "London
City", "city" : "London" }
db.airports
.find( { $text: { $search: "City" } })
.forEach(
function(e, i){
e.name='Big ' + e.name;
db.airports.save(e);
}
)
after:
{ "_id" : ObjectId("56993251e843bb7e0447829d"), "name" : "Big London
City", "city" : "London" }
Old topic but i had the same problem.
Since mongo 2.4, you can use $concat from aggregation framework.
Example
Consider these documents :
{
"_id" : ObjectId("5941003d5e785b5c0b2ac78d"),
"title" : "cov"
}
{
"_id" : ObjectId("594109b45e785b5c0b2ac97d"),
"title" : "fefe"
}
Append fefe to title field :
db.getCollection('test_append_string').aggregate(
[
{ $project: { title: { $concat: [ "$title", "fefe"] } } }
]
)
The result of aggregation will be :
{
"_id" : ObjectId("5941003d5e785b5c0b2ac78d"),
"title" : "covfefe"
}
{
"_id" : ObjectId("594109b45e785b5c0b2ac97d"),
"title" : "fefefefe"
}
You can then save the results with a bulk, see this answer for that.
this is a sample of one document i have :
{
"_id" : 1,
"s" : 1,
"ser" : 2,
"p" : "9919871172",
"d" : ISODate("2018-05-30T05:00:38.057Z"),
"per" : "10"
}
to append a string to any feild you can run a forEach loop throught all documents and then update desired field:
db.getCollection('jafar').find({}).forEach(function(el){
db.getCollection('jafar').update(
{p:el.p},
{$set:{p:'98'+el.p}})
})
This would not be possible.
One optimization you can do is create batches of updates.
i.e. fetch 10K documents, append relevant strings to each of their keys,
and then save them as single batch.
Most mongodb drivers support batch operations.
db.getCollection('<collection>').update(
// query
{},
// update
{
$set: {<field>:this.<field>+"<new string>"}
},
// options
{
"multi" : true, // update only one document
"upsert" : false // insert a new document, if no existing document match the query
});
Want to upsert in object properties in a array of a document
Consider a document in collection m
{ "_id" : ObjectId("524bfc39e6bed5cc5a9f3a33"),
"x" : [
{ "id":0.0, "name":"aaa"},{ "id":1.0, "name":"bbb"}
]
}
Want to add age:100 to { "id":0.0, "name":"aaa"} .
Not just age .. But but provision for upsert in the array element {}. So it can contain {age:100,"city":"amd"} (since i am getting this from the application service)
Was trying this... But did not worked as it replaced the entire array element
db.m.update({_id:ObjectId("524bfc39e6bed5cc5a9f3a33"),
"x" : {
"$elemMatch" : {
"id" : 0.0
}
}},
{
$set : {
"x.$" : {
"age": 100
}
}
},
{upsert: true}
)
Changed the document to (which i did not wanted)
{ "_id" : ObjectId("524bfc39e6bed5cc5a9f3a33"),
"x" : [
{ "age":100},{ "id":1.0, "name":"bbb"}
]
}
Is this possible without changing schema.
$set : {"x.$" : {"age": 100}}
x.$ sets the entire matched array element to {age: 100}
This should work:
db.m.update({_id:ObjectId("524bfc39e6bed5cc5a9f3a33"),
"x.id": 0.0}, {$set: {"x.$.age": 100 }});
Using elemMatch:
db.test.update({x: {$elemMatch: {id: 1}}},{$set: {"x.$.age": 44}})
Note that the upsert option here, is redundant and wouldn't work if the id isn't present in x because the positional operator $ doesn't support upserting.
This is not possible without changing schema. If you can change schema to use an object to store your items (rather than an array), you can follow the approach I outlined in this answer.