In a nutshell, the purpose of this tool is to generate a GraphQL schema based on a MongoDB's collection.
Note that this tool has been named MongoQL-core since it doesn't have much purpose alone. However, it is the foundation of the concept. In fact, a GraphQL schema can be useful to describe a contract but alone it's not executable. This tool would/should be used conjointly with another tool in order to generate the GraphQL data fetchers required.
A couple tools should be installed beforehand.
npm install variety-cli -g
Please verify that variety is accessible from a command line/terminal.
For Windows, open a command line call and type:
variety.cmd
For Mac OS X/Linux, open a terminal and type:
variety
If it is not, please add an environment variable to make it available.
Here is described exactly how to get a running example of MongoQL-core:
Open a terminal or a command line and connect to Mongo locally with:
mongo
Now you can populate data into a database/collection:
use myDatabase
db.myCollection.insertOne({"name": "David", "age":32})
db.myCollection.insertOne({"name": "Mathieu", "age":32})
db.myCollection.insertOne({"name": "Nuno", "age":32})
db.myCollection.insertOne({"name": "Kevin"})
db.myCollection.insertOne({"name": "Michael", "child": {"name" : "L", "age":1}})
Result:
type Child {
age : Int!
name : String!
}
type Person {
_id : ID!
name : String!
age : Int
child : Child
}
- Connect to MongoDB atlas (or create an account);
- Create a new project;
- Create a new cluster;
- Pick a shared cluster with a free subscription
Please be patient, this step could take a bit of time.
Go in Network Access and add a new ip address. Then select ALLOW ACCESS FROM ANYWHERE.
Please be patient, this step could take a bit of time.
Go in Database Access and add a new database user. In my example, the user is called user and it's password is also user. Note that you will need to give him access to your database in read and write (unfortunately Variety needs write rights as it performs calculations on the server).
Please be patient, this step could take a bit of time.
Retrieve your connection URL from MongoDB atlas. It should look something like this:
mongo "mongodb+srv://****.mongodb.net/test" --username ****
And then enters your password.
Now you can populate data into a database/collection:
use myDatabase
db.myCollection.insertOne({"name": "David", "age":32})
db.myCollection.insertOne({"name": "Mathieu", "age":32})
db.myCollection.insertOne({"name": "Nuno", "age":32})
db.myCollection.insertOne({"name": "Kevin"})
db.myCollection.insertOne({"name": "Michael", "child": {"name" : "L", "age":1}})
Result:
type Child {
age : Int!
name : String!
}
type Person {
_id : ID!
name : String!
age : Int
child : Child
}
Since it's in the early stages, there is, sadly, a couple limitations:
Since MongoDB is schema-less, it enables fields to have multiple types. This tool should be able to handle this in a near future but currently it's not supported.
The tool used to introspect the MongoDB schema of a collection isn't precise enough to give information of content of arrays. For now, it is out of scope of this tool.
There might be schema with sub documents that are the same as other documents. One could determine an fields equivalence between types found to regroup into one type. In the full example above, the type Child is most probably a type Person and it should be only one type rather than two.
If you enjoyed this projet, don't hesitate to pay me a coffee ;)
- Kotlin
- Mockito
- JUnit
- Jackson
- Variety (https://github.com/variety/variety)
