Generate fake data for your sequelize model directly from your sequelize schema.
Install from npm npm i faquel
This library exports two function generateEntryFromModel
, generateEntryFromSchema
. Both of them take 2 arguments.
First argument for generateEntryFromModel
is a sequelize model instance which will be used as reference to generate fake data.
First argument for generateEntryFromSchema
on the other hand, is a raw schema definition in data object format.
Second argument is an optional object config that you can use to overwrite the default generated fake data with custom faker methods. Follow the examples below to get a better understanding.
Let's start with a model that looks like:
const PersonSchema = {
name: {
type: DataTypes.CHAR(20),
},
annualIncome: {
type: DataTypes.DOUBLE(11, 2),
},
dob: {
type: DataTypes.DATEONLY
},
option: {
type: DataTypes.ENUM("1", "2", "3", "4"),
allowNull: true,
},
};
class Person extends Model {
static init(sequelize: any) {
const inst = super.init(PersonSchema, {
modelName: 'Person',
timestamps: true,
sequelize,
});
return inst;
};
};
Assuming that you have followed sequelize guide and setup your initialized models in a container named models
.
import { generateEntryFromModel } from 'faquel';
// other code.... i.e: setup for sequelize
...
const fakePersonFromModel = generateEntryFromModel(models.Person);
/*
fakePerson = {
id: 26651,
name: 'Cupiditate nulla del',
annualIncome: 80886196623.68,
monthlyIncome: 20656,
dob: 2018-08-09T13:37:39.846Z,
option: '3',
createdAt: 2019-01-03T00:12:40.279Z,
updatedAt: 2018-08-29T13:52:40.877Z
};
*/
const fakePersonFromSchema = generateEntryFromSchema(PersonSchema);
/*
fakePersonFromSchema = {
name: 'Cupiditate nulla del',
annualIncome: 80886196623.68,
monthlyIncome: 20656,
dob: 2018-08-09T13:37:39.846Z,
option: '3',
};
*/
Notice how in fakePersonFromSchema
the id, createdAt, updatedAt
etc fields are missing? that's because those columns are autogenerated by sequelize model when they are initiated and since the generateEntryFromSchema
function generates the data from data object schema definition, it only outputs field that are present in the object.
That's great and all but the name doesnt look like a real name here, does it? This library uses faker.js under the hood and faker has a lot of nifty methods that can generate "realistic" data. Fortunately, you can use those too.
const fakePerson = generateEntryFromModel(models.Person, {
name: 'name.findName'
});
/*
fakePerson = {
... other columns
name: 'John Doe',
... other columns
};
*/
Notice how the name looks more like a real name in the example above? that's because of the 2nd object parameter we are passing to the generateEntryFromModel
function specifies that the name
column should use the faker.name.findName()
method to generate it's value.
You can pass any faker method name as a string like the example above.
Also, faker may not be good enough for generating the desired data so as a supreme overrider, you can pass a function mapped to column name through fakerMap and the function will be invoked to generate your data.
const fakePerson = generateEntryFromModel(models.Person, {
name: () => {
// do a lot of computation here
return 'the answer to life is 42';
}
});
/*
fakePerson = {
... other columns
name: 'the answer to life is 42',
... other columns
};
*/
- mysql SET columns are currently handled but not tested since we are using sqlite for testing. may be a test suite across dbs would help?
PRs are more than welcome since right now, the library only supports a few of the column types from sequelize.
To run the lib locally, you can clone the repo then run the following commands:
npm i
to install the dependenciesnpm run build:watch
to compile typescript code and run the compiled code through nodemon so that you can develop without restarting the builder over and over again.npm test
to run the only couple of tests we have right now.