prisma-mock-vitest

Prisma Mock - Vitest

This is a mock of the Prisma API, intended for unit testing with Vitest. All the data is stored in memory.

This library uses vitest-mock-extended, so if functionality you need is not implemented yet, you can mock it yourself.

Pull requests are welcome for missing features.

Installation

yarn add -D prisma-mock-vitest

npm install prisma-mock-vitest --save-dev

You should also have run npx prisma generate for your Prisma schema.

Usage

Simple example how to create a prisma mock instance:

import { createPrismaMock } from 'prisma-mock-vitest';
import { beforeEach, expect, test } from 'vitest';

let client: PrismaClient;

beforeEach(async () => {
  client = await createPrismaMock();
});

test('Initial user count is 0', async () => {
  const totalUsers = await client.user.count();
  expect(totalUsers).toEqual(0);
});

An example how to mock a global prisma instance inside and schema a 'db' directory (like blitzjs):

import type { PrismaClient } from '@prisma/client';
import { createPrismaMock } from 'prisma-mock-vitest';
import { beforeEach } from 'vitest';
import { mockDeep, mockReset } from 'vitest-mock-extended';

vi.mock('db', () => ({
  __esModule: true,
  ...jest.requireActual('db'),
  default: mockDeep<PrismaClient>(),
}));

import db, { Prisma } from 'db';

beforeEach(() => {
  mockReset(db);
  return createPrismaMock({}, Prisma.dmmf.datamodel, db);
});

API

createPrismaMock<P extends PrismaClient = PrismaClient>(
  data: PrismaMockData<P> = {},
  datamodel?: Prisma.DMMF.Datamodel,
  client = mockDeep<P>(),
): Promise<P>;

data

Object with an array default data for each table/model.

createPrismaMock({
  users: [
    {
      id: 1,
      name: 'John Doe',
      accountId: 1,
    },
  ],
  account: [
    {
      id: 1,
      name: 'Company',
    },
  ],
});

datamodel

The datamodel of the prisma client, i.e. Prisma.dmmf.datamodel.

client

vitest-mock-extended instance. If it's not provided, a new instance is created.

Supported Features

A lot of the functionality is implemented, but parts are missing. Here is a list of the (missing) features.

Model Queries

• findUnique,
• findMany,
• findFirst,
• create,
• createMany
• delete,
• update,
• deleteMany,
• updateMany
• upsert
• count
• TODO: aggregate
• TODO: groupBy

Model Query Options

• distinct
• include
• where
• select
• orderBy
• TODO: select: _count

Nested Queries

• create
• createMany
• update
• updateMany
• delete
• deleteMany
• connect
• connectOrCreate
• disconnect
• TODO: set
• TODO: upsert

Filter Conditions and Operators

• equals
• gt
• gte
• lt
• lte
• not
• in
• notIn
• contains
• startWith
• endsWith
• AND
• OR
• NOT
• TODO: search
• TODO: mode

Relation Filters

• some
• every
• none
• TODO: is

Scalar List Methods

TODO (set, push)

Scalar List Filters

TODO (has, hasEvery, hasSome, isEmpty, equals)

Atomic Number Operations

• increment
• decrement
• multiply
• divide
• set

JSON Filters

TODO (path, string_contains, string_starts_with, string_ends_with, array_contains, array_starts_with, array_ends_with)

Attributes

• @@id
• @default
• @unique (TODO: no error if duplicate)
• @@unique (TODO: no error if duplicate)
• @relation
• TODO: @updatedAt

Attribute functions

• autoincrement()
• cuid()
• uuid()
• now()
• TODO: auto()
• TODO: dbgenerated()

Referential actions

• onDelete (SetNull, Cascade)
• TODO: onDelete: Restrict, NoAction, SetDefault
• TODO: onUpdate