I've looked around quite a bit concerning this error, it seems that Mongo won't accept a . or a $ in an update, yet I still get this error
{ [MongoError: not okForStorage]
name: 'MongoError',
err: 'not okForStorage',
code: 12527,
n: 0,
connectionId: 18,
ok: 1 }
This is the object I'm updating:
{
status: "open",
type: "item",
parentId: "4fa13ba2d327ca052d000003",
_id: "4fa13bd6d327ca052d000012",
properties: {
titleInfo: [
{ title: "some item" }
]
}
}
And I'm updating it to:
{
fedoraId: 'aFedoraLib:438',
status: "closed",
type: "item",
parentId: "4fa13ba2d327ca052d000003",
_id: "4fa13bd6d327ca052d000012",
properties: {
titleInfo: [
{ title: "some item" }
]
}
}
Another possible cause I just ran into: storing an object which has periods in the string keys.
So for people getting the same error:
It's due to the fact that I was including the _id, which Mongo doesn't like apparently
I ran into this error when trying to save a JSON structure with this key-value pair (coming straight out of an AngularJS app):
"$$hashKey":"021"
Removing just that key fixed the problem. For others using Angular, it looks like calling Angular's built-in angular.toJson client-side eliminates the $$hashkey keys. From their forums:
$scope.ngObjFixHack = function(ngObj) {
var output;
output = angular.toJson(ngObj);
output = angular.fromJson(output);
return output;
}
Related
I'm working on building a database from a set of CSV files and am running into issues pushing elements to a deeply nested array. I've seen examples for 2D arrays and the use of positional operators but I can't quite figure out how to use them for my situation.
What I am trying to do read the CSV file which has columns for the answer_id and the url string for the photo. I want to push the url string to the photos array for the corresponding answer_id. When I try to use the code below, I get a long error message which starts with:
MongoBulkWriteError: Cannot create field 'answers' in element
{
results: [
{
question_id: "1",
_id: ObjectId('6332e0b015c1d1f4eccebf4e'),
answers: [
{
answer_id: "5",
//....
}
],
}
]
}
I formatted the above message to make things easier to read. It may be worth noting the first row of my CSV file has '5' in the answer_id column which makes me think things are failing at the first try to update my document.
Here is an example of my code:
const ExampleModel = new Schema({
example_id: String,
results: [
{
question_id: String,
answers: [
{
answer_id: String,
photos: [
{ url: String },
]
}
]
}
]
});
// Example Operation
// row.answer_id comes from CSV file
updateOne: {
'filter': {
'results.answers.answer_id': row.answer_id,
},
'update': {
'$push': {
'results.answers.$.photos' : { 'url': 'test'}
}
}
}
I guess my question is can I update an array this deeply nested using Mongoose?
I have a MongoDB collection for storing chat objects with messages embedded as a nested array. The entire collection looks like this:
chats = [
{
_id: 0,
messages: [
{
_id: 0,
text: 'First message'
},
{
_id: 1,
text: 'Second message'
}
]
},
{
_id: 1,
messages: []
}
]
I would like to update a single message and return it to the user. I can update the message like this (in Node):
const chat = chats.findOneAndUpdate({
_id: ObjectId(chatID),
"messages._id": ObjectId(messageID)
}, {
$set: {
"messages.$.text": newText
}
});
The issue is that this query updates a message and returns a chat object, meaning that I have to look for an updated message in my code again. (i.e. chat.messages.find(message => message._id === messageID)).
Is there a way to get a message object from MongoDB directly? It would also be nice to do the update in the same query.
EDIT: I am using Node with mongodb.
Thank you!
Since MongoDB methods findOneAndUpdate, findAndModify do not allow to get the updated document from array field,
projection will return updated sub document of array using positional $ after array field name
returnNewDocument: true will return updated(new) document but this will return whole document object
The problem is, MongoDB cannot allowing to use a positional projection and return the new document together
For temporary solution
try using projection, this will return original document from array field using $ positional,
const chat = chats.findOneAndUpdate(
{
_id: ObjectId(chatID),
"messages._id": ObjectId(messageID)
},
{ $set: { "messages.$.text": newText } },
{
projection: { "messages.$": 1 }
}
);
Result:
{
"_id" : 0.0,
"messages" : [
{
"_id" : 0.0,
"text" : "Original Message"
}
]
}
To get an updated document, a new query could be made like this:
const message = chats.findOne({
_id: ObjectId(chatID),
"messages._id": ObjectId(messageID)
}, {
projection: {'messages.$': 1},
}).messages[0];
Result:
{
"_id": 0.0,
"text": "New message",
}
I am using Redis in my backend to scale subscriptions. I am using this library to implement redis on top of my javascript code. And using mongoose for the models.
During a redis publish, I have to stringify the objects that I get from mongoose. I parse them on the subscribing end and it all works well until there's a nested object in the object that needs to be stringify-ed.
So if my object is this:
{ subtitle: '',
description: '',
checklists:
[ { _id: 5cee450c0fa29d0b54275da0, items: [] },
{ _id: 5cee455c0c31785b0875e09d, items: [] },
{ _id: 5cee47dc6d32e72c6411ce2d, items: [] } ],
attachments: [],
labels: [],
_id: 5ced1af26547233798f943f6,
title: 'asfasf',
box: 5cece1c3e6c3c13ff098658d,
workflow: 5cece1cbe6c3c13ff0986591,
}
I receive:
{ cardUpdated:
{
subtitle: '',
description: '',
checklists: [ [Object], [Object], [Object] ],
attachments: [],
labels: [],
_id: '5ced1af26547233798f943f6',
title: 'asfasf',
box: '5cece1c3e6c3c13ff098658d',
workflow: '5cece1cbe6c3c13ff0986591',
}
}
When I publish I use the following line:
pub.publish(types.CARD_UPDATED,
JSON.stringify(
{ cardUpdated: await getUpdatedCardStats(checklist.card) },
));
Note: I know that I am wrapping the argument for stringify in {} and without it the nested objects would not be ignored, but I need to do that because I need the key property name on the subscription end i.e. I am using this publish command with different key names at several places.
Is there a way to about this to get the nested objects stringify-ed?
EDIT: Turns out, I was getting the proper full object as a string on the subscribing end of Redis, but it was actually JSON.parse() that was the culprit. After parsing, it completely ignores the nested objects. Is there anyway to avoid this?
Try:
const value = JSON.stringify({
cardUpdated: await getUpdatedCardStats(checklist.card)
});
pub.publish(types.CARD_UPDATED, value);
This is not a valid JS object:
{ _id: 5cee450c0fa29d0b54275da0, items: [] }
I think it's the output of .toString() of an object of type {_id: ObjectId, items: any[], with ObjectId defined here. In any case, the JSONification of this object is not trivial and that is why JSON.stringify simply outputs [Object].
To bypass this limitation, you might implement a custom function to transform your object into one that can be trivially JSONified, possibly with the help of ObjectId.toString().
I have a structure like this:
{
...
_id: <projectId>
en-GB: [{
_id: <entryId>,
key: 'some key',
value: 'some value',
}]
}
And I've tried updating it with Mongoose (and raw mongodb too) like this:
const result = await Project
.update({
_id: projectId,
'en-GB._id': entryId,
}, {
$set: {
'en-GB.$.key': 'g000gle!!',
},
})
.exec();
I've checked that the IDs are correct. But it doesn't update anything:
{ n: 0, nModified: 0, ok: 1 }
What am I doing wrong? Thanks
As discussed in the comments on the question, the issue is directly related to passing in a string representation of an id in the query as opposed to using an ObjectId. In general, it's good practice to treat the use of ObjectIds as the rule and the use of string representations as special exceptions (e.g. in methods like findByIdAndUpdate) in order to avoid this issue.
const { ObjectId } = require('mongodb');
.update({
_id: ObjectId(projectId),
'en-GB._id': ObjectId(entryId),
})
I have this code to get some stock data with deep nested associations by using nested-pop:
Stock.find({limit:100, sort:'name ASC'})
.populate('scans', {limit:1,sort:'createdAt DESC'})
.then((stocks) => {
return nestedPop(stocks, {
scans: ['values']
})
}).then(stocks => {
console.log(stocks[0].scans[0]);
res.json(stocks);
})
In the server console the scan is logged and the values array is printed. But res.json returns the stocks with scans that are all missing the values array.
Expected output:
[{
id: 1,
scans: [
{
id: 1,
values: [
{
id: 1,
value: 0.5,
type: 1
},
...
]
}
]
}, ...]
The returned output:
[{
id: 1,
scans: [
{
id: 1,
}
]
}, ...]
The error doesn't seem to be related to the depth of the object, because I tested it with some deeper test object and everything was returned. I just don't understand how this is not returned by res.json if it's there.
UPDATE:
I made other tests now and found out that both, .toObject and .toJSON called on a single stock cause the values to disappear on the server side. And because res.json is calling .toJSON on the passed objects it also doesn't appear in the browser.
nestedPop is returning values that can not be parsed correctly by .toJSON, because they are already converted.
I solved the problem by converting all stocks before I pass them into nestedPop, so the working code is:
Stock.find({limit:100, sort:'name ASC'})
.populate('scans', {limit:1,sort:'createdAt DESC'})
.then((stocks) => {
stocks = stocks.map((s) => s.toJSON());
return nestedPop(stocks, {
scans: ['values']
})
}).then(stocks => {
console.log(stocks[0].scans[0]);
res.json(stocks);
})
Now everything is logged on the server side AND passed to browser.