when I attempt to insert new document manually in system.indexes collection in mongodb,new collection created.here goes the code
{
"v" : 1,
"key" : {
"code" : 1
},
"name" : "code_1",
"ns" : "mydb.collection"
}
where collection is my collection name which is not already present in database and mydb is my database name. Why new collection is getting created?
Is it possible to create collection by adding index manually in system.indexes.
Why are you asking us this? You already tried to add a new index to system.indexes. Has a new collection been created? If yes, then yes it is possible, if no, then not possible.
Is this a correct way?
How do you think? Have you read somewhere in documentation that in order to create a new collection you need to dance around and to create manually indexes in some system defined collection? Or may be it was written in documentation that db.createCollection(name, options) is what you should do or if you so desire you can just insert a document in a non existed collection and it will create it.
So why after all this one might think that the correct way is to do some manipulation with system.indexes?
As a complement to #Salvador Dali's answer strongly discouraging you to do modify system.index directly: if for some reason you really don't want/can't use createCollection, just remember this is a wrapper around the create command.
You can issue yourself such command to create a new collection:
db.runCommand( { create: "collection" } )
As about inserting an entry in system.indexes: from the doc:
Deprecated since version 3.0: Access this data using listIndexes.
The <database>.system.indexes collection lists all the indexes in the database.
By reading that it appears that system.indexes should be considered as read-only (its direct use is even deprecated since 3.0). The behavior you observed should be considered as unspecified. And so unreliable and subject to change without further notices.
If you really need to understand why it behave that way, maybe you should take a look at the source code or ask the question on the MongoDB developer mailing list. There you could have all the insights.
In meteor, when I create an item in a collection, the generated id for the item usually looks like:
"_id" : "vxqbpic8yLdc6Ehor"
However, when I insert a row directly in Mongo, it's generated id looks like:
"_id" : ObjectId("549af35926cee46520611838")
Is there a way for me to insert data directly into mongo generating an id similar to the way meteor does, or is that something special with meteor? I'd be happy with just dropping the "ObjectId()" wrap around the value, if that's possible.
What meteor actually does is create a random 17 character string that consists of characters from within 23456789ABCDEFGHJKLMNPQRSTWXYZabcdefghijkmnopqrstuvwxyz
https://github.com/meteor/meteor/blob/devel/packages/random/random.js is what does that. Namely the RandomGenerator.prototype.id(17) function.
So you can include that in your custom code or any other piece of code that generates 17 character random strings from those characters I've given above, and use its outcome as your ID.
In fact, any other random string would suffice, as long as it is universally random, which Meteor's implementation tries to attain.
Here is a function that uses Math.random() to generate Meteor Id's:
const UNMISTAKABLE_CHARS = '23456789ABCDEFGHJKLMNPQRSTWXYZabcdefghijkmnopqrstuvwxyz';
function newMeteorId () {
return [...Array(17).keys()]
.map(() =>
UNMISTAKABLE_CHARS[
Math.floor(Math.random() * UNMISTAKABLE_CHARS.length)
]
)
.join('')
}
My understanding is that the ids generated using Meteor's random.js are more truly random where as this is not but I believe for most use cases this will be sufficiently random.
I use mongo_insert() three times to insert my data in three different collections. The problem is that the "_id" field must be exactly the same in each of the collections, but I do not know how to (ideally) recover and reuse the "_id" field generated in my first mongo_insert...
Please, advice me how to do it.
Normally, you could have different field, like CustomId for your private needs, and leave _id for mongo generation.
But if you still need it to be exactly the same - there could be 2 variants:
1) setting custom generated _id do each doc.
2) Save first doc, then read it again, check it's _id and set it to the other docs.
in SQL world I could do something to the effect of:
SELECT name FROM table WHERE UPPER(name) = UPPER('Smith');
and this would match a search for "Smith", "SMITH", "SmiTH", etc... because it forces the query and the value to be the same case.
However, MongoDB doesn't seem to have this capability without using a RegEx, which won't use indexes and would be slow for a large amount of data.
Is there a way to convert a stored value to a particular case before doing a search against it in MongoDB?
I've come across the $toUpper aggregate, but I can't figure out how that would be used in this particular case.
If there's not way to convert stored values before searching, is it possible to have MongoDB convert a value when it's created in Mongo? So when I add a document to the collection it would force the "name" attribute to a particular case? Something like a callback in the Rails world.
It looks like there's the ability to create stored JS for MongoDB as well, similar to a Stored Procedure. Would that be a feasible solution as well?
Mostly looking for a push in the right direction; I can figure out the particular code once I know what I'm looking for, but so far I'm not even sure if my desired functionality is doable.
You have to normalize your data before storing them. There is no support for performing normalization as part of a query at runtime.
The simplest thing to do is probably to save both a case-normalized (i.e. all-uppercase) and display version of the field you want to search by. Suppose you are storing users and want to do a case-insensitive search on last name. You might store:
{
_id: ObjectId(...),
first_name: "Dan",
last_name: "Crosta",
last_name_upper: "CROSTA"
}
You can then create an index on last_name_upper, and query like:
> db.users.find({last_name_upper: "CROSTA"})
So this is Day 3 of learning Mongo Db. I'm coming from the MySql universe...
A lot of times when I need to write a query for a MySql table I'm unfamiliar with, I would use the "desc" command - basically telling me what fields I should include in my query.
How would I do that for a Mongo db? I know, I know...I'm searching for a schema in a schema-less database. =) But how else would users know what fields to use in their queries?
Am I going at this the wrong way? Obviously I'm trying to use a MySql way of doing things in a Mongo db. What's the Mongo way?
Type the below query in editor / mongoshell
var col_list= db.emp.findOne();
for (var col in col_list) { print (col) ; }
output will give you name of columns in collection :
_id
name
salary
There is no good answer here. Because there is no schema, you can't 'describe' the collection. In many (most?) MongoDb applications, however, the schema is defined by the structure of the object hierarchy used in the writing application (java or c# or whatever), so you may be able to reflect over the object library to get that information. Otherwise there is a bit of trial and error.
This is my day 30 or something like that of playing around with MongoDB. Unfortunately, we have switched back to MySQL after working with MongoDB because of my company's current infrastructure issues. But having implemented the same model on both MongoDB and MySQL, I can clearly see the difference now.
Of course, there is a schema involved when dealing with schema-less databases like MongoDB, but the schema is dictated by the application, not the database. The database will shove in whatever it is given. As long as you know that admins are not secretly logging into Mongo and making changes, and all access to the database is controller through some wrapper, the only place you should look at for the schema is your model classes. For instance, in our Rails application, these are two of the models we have in Mongo,
class Consumer
include MongoMapper::Document
key :name, String
key :phone_number, String
one :address
end
class Address
include MongoMapper::EmbeddedDocument
key :street, String
key :city, String
key :state, String
key :zip, String
key :state, String
key :country, String
end
Now after switching to MySQL, our classes look like this,
class Consumer < ActiveRecord::Base
has_one :address
end
class Address < ActiveRecord::Base
belongs_to :consumer
end
Don't get fooled by the brevity of the classes. In the latter version with MySQL, the fields are being pulled from the database directly. In the former example, the fields are right there in front of our eyes.
With MongoDB, if we had to change a particular model, we simply add, remove, or modify the fields in the class itself and it works right off the bat. We don't have to worry about keeping the database tables/columns in-sync with the class structure. So if you're looking for the schema in MongoDB, look towards your application for answers and not the database.
Essentially I am saying the exactly same thing as #Chris Shain :)
While factually correct, you're all making this too complex. I think the OP just wants to know what his/her data looks like. If that's the case, you can just
db.collectionName.findOne()
This will show one document (aka. record) in the database in a pretty format.
I had this need too, Cavachon. So I created an open source tool called Variety which does exactly this: link
Hopefully you'll find it to be useful. Let me know if you have questions, or any issues using it.
Good luck!
AFAIK, there isn't a way and it is logical for it to be so.
MongoDB being schema-less allows a single collection to have a documents with different fields. So there can't really be a description of a collection, like the description of a table in the relational databases.
Though this is the case, most applications do maintain a schema for their collections and as said by Chris this is enforced by your application.
As such you wouldn't have to worry about first fetching the available keys to make a query. You can just ask MongoDB for any set of keys (i.e the projection part of the query) or query on any set of keys. In both cases if the keys specified exist on a document they are used, otherwise they aren't. You will not get any error.
For instance (On the mongo shell) :
If this is a sample document in your people collection and all documents follow the same schema:
{
name : "My Name"
place : "My Place"
city : "My City"
}
The following are perfectly valid queries :
These two will return the above document :
db.people.find({name : "My Name"})
db.people.find({name : "My Name"}, {name : 1, place :1})
This will not return anything, but will not raise an error either :
db.people.find({first_name : "My Name"})
This will match the above document, but you will have only the default "_id" property on the returned document.
db.people.find({name : "My Name"}, {first_name : 1, location :1})
print('\n--->', Object.getOwnPropertyNames(db.users.findOne())
.toString()
.replace(/,/g, '\n---> ') + '\n');
---> _id
---> firstName
---> lastName
---> email
---> password
---> terms
---> confirmed
---> userAgent
---> createdAt
This is an incomplete solution because it doesn't give you the exact types, but useful for a quick view.
const doc = db.collectionName.findOne();
for (x in doc) {
print(`${x}: ${typeof doc[x]}`)
};
If you're OK with running a Map / Reduce, you can gather all of the possible document fields.
Start with this post.
The only problem here is that you're running a Map / Reduce on which can be resource intensive. Instead, as others have suggested, you'll want to look at the code that writes the actual data.
Just because the database doesn't have a schema doesn't mean that there is no schema. Generally speaking the schema information will be in the code.
I wrote a small mongo shell script that may help you.
https://gist.github.com/hkasera/9386709
Let me know if it helps.
You can use a UI tool mongo compass for mongoDb. This shows all the fields in that collection and also shows the variation of data in it.
If you are using NodeJS and want to get the all the field names using the API request, this code works for me-
let arrayResult = [];
db.findOne().exec(function (err, docs)){
if(err)
//show error
const JSONobj = JSON.parse(JSON.stringify(docs));
for(let key in JSONobj) {
arrayResult.push(key);
}
return callback(null, arrayResult);
}
The arrayResult will give you entire field/ column names
Output-
[
"_id",
"emp_id",
"emp_type",
"emp_status",
"emp_payment"
]
Hope this works for you!
Consider you have collection called people and you want to find the fields and it's data-types. you can use below query
function printSchema(obj) {
for (var key in obj) {
print( key, typeof obj[key]) ;
}
};
var obj = db.people.findOne();
printSchema(obj)
The result of this query will be like below,
you can use Object.keys like in JavaScript
Object.keys(db.movies.findOne())