Prisma JSON Schema Generator

A generator, which takes a Prisma 2 schema.prisma and generates a JSON Schema in version 7 of the specification (https://json-schema.org/).

Getting Started

1. Install

npm:

``shell
npm install prisma-json-schema-generator --save-dev
`

yarn:

`shell
yarn add -D prisma-json-schema-generator
`

2. Add the generator to the schema

`prisma
generator jsonSchema {
provider = "prisma-json-schema-generator"
}
`

With a custom output path (default=./json-schema)

`prisma
generator jsonSchema {
provider = "prisma-json-schema-generator"
output = "custom-output-path"
}
`

Additional options

`prisma
generator jsonSchema {
provider = "prisma-json-schema-generator"
keepRelationScalarFields = "true"
schemaId = "some-schema-id"
includeRequiredFields = "true"
persistOriginalType = "true"
forceAnyOf = "true"
}
`

The generator currently supports a few options

| Key | Default Value | Description |
| ------------------------ | ------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| keepRelationScalarFields | "false" | By default, the JSON Schema that's generated will output only objects for related model records. If set to "true", this will cause the generator to also output foreign key fields for related records |
| keepRelationFields | "true" | Determines whether to include fields from related models in the generated schema. Setting it to "false" allows excluding related model fields from the schema. |
| schemaId | undefined | Add an id to the generated schema. All references will include the schema id |
| includeRequiredFields | "false" | If this flag is "true" all required scalar prisma fields that do not have a default value, will be added to the required properties field for that schema definition. |
| persistOriginalType | "false" | If this flag is "true" the original type will be outputed under the property key "originalType" |
| forceAnyOf | "false" | If this flag is "true" the union types will be forced to use anyOf. Check contradictory types for details |

3. Run generation

prisma:

`shell
prisma generate
`

nexus with prisma plugin:

`shell
nexus build
`

Supported Node Versions

| Node Version | Support |
| -------------------: | :----------------- |
| (Maintenance LTS) 18 | :heavy_check_mark: |
| (LTS) 20 | :heavy_check_mark: |
| (Current) 21 | :heavy_check_mark: |

Examples

$3

This generator converts a prisma schema like this:

`prisma
datasource db {
provider = "postgresql"
url = env("DATABASE_URL")
}

model User {
id Int @id @default(autoincrement())
// Double Slash Comment: It will NOT show up in JSON schema
createdAt DateTime @default(now())
/// Triple Slash Comment: It will show up in JSON schema [EMAIL]
email String @unique
weight Float?
is18 Boolean?
name String?
number BigInt @default(34534535435353)
favouriteDecimal Decimal
bytes Bytes /// Triple Slash Inline Comment: It will show up in JSON schema [BYTES]
successorId Int? @unique
successor User? @relation("BlogOwnerHistory", fields: [successorId], references: [id])
predecessor User? @relation("BlogOwnerHistory")
role Role @default(USER)
posts Post[]
keywords String[]
biography Json
}

model Post {
id Int @id @default(autoincrement())
user User? @relation(fields: [userId], references: [id])
userId Int?
}

enum Role {
USER
ADMIN
}
`

into:

`javascript
{
$schema: 'http://json-schema.org/draft-07/schema#',
definitions: {
Post: {
properties: {
id: { type: 'integer' },
user: {
anyOf: [
{ $ref: '#/definitions/User' },
{ type: 'null' },
],
},
},
type: 'object',
},
User: {
properties: {
biography: {
type: [
'number',
'string',
'boolean',
'object',
'array',
'null'
],
},
createdAt: { format: 'date-time', type: 'string' },
email: {
description: 'Triple Slash Comment: Will show up in JSON schema [EMAIL]',
type: 'string'
},
id: { type: 'integer' },
is18: { type: ['boolean', 'null'] },
keywords: { items: { type: 'string' }, type: 'array' },
name: { type: ['string', 'null'] },
number: { type: 'integer', default: '34534535435353' },
bytes: {
description: 'Triple Slash Inline Comment: Will show up in JSON schema [BYTES]',
type: 'string'
},
favouriteDecimal: { type: 'number' },
posts: {
items: { $ref: '#/definitions/Post' },
type: 'array',
},
predecessor: {
anyOf: [
{ $ref: '#/definitions/User' },
{ type: 'null' },
],
},
role: { enum: ['USER', 'ADMIN'], type: 'string', default: 'USER' },
successor: {
anyOf: [
{ $ref: '#/definitions/User' },
{ type: 'null' },
],
},
weight: { type: ['integer', 'null'] },
},
type: 'object',
},
},
properties: {
post: { $ref: '#/definitions/Post' },
user: { $ref: '#/definitions/User' },
},
type: 'object',
}
`

So the following input will correctly be validated:

`javascript
{
post: {
id: 0,
user: {
id: 100,
},
},
user: {
id: 10,
createdAt: '1997-07-16T19:20:30.45+01:00',
email: 'jan@scharnow.city',
biography: {
bornIn: 'Scharnow',
},
is18: true,
keywords: ['prisma2', 'json-schema', 'generator'],
name: null,
posts: [
{
id: 4,
},
{
id: 20,
},
],
predecessor: {
id: 10,
email: 'horst@wassermann.de',
},
successor: null,
role: 'USER',
weight: 10.14,
},
}
`

$3

The generator also takes care of composite types in MongoDB:

`prisma
datasource db {
provider = "mongodb"
url = env("DATABASE_URL")
}

model User {
id String @id @default(auto()) @map("_id") @db.ObjectId
photos Photo[]
}

type Photo {
height Int @default(200)
width Int @default(100)
url String
}
`

Output:

`javascript
{
$schema: 'http://json-schema.org/draft-07/schema#',
definitions: {
User: {
properties: {
id: { type: 'string' },
photos: {
items: { $ref: '#/definitions/Photo' },
type: 'array',
},
},
type: 'object',
},
Photo: {
properties: {
height: {
type: 'integer',
default: 200,
},
width: {
type: 'integer',
default: 100,
},
url: {
type: 'string',
},
},
type: 'object',
},
},
properties: {
user: { $ref: '#/definitions/User' },
},
type: 'object',
}
`

$3

For some use cases, it might be useful to not include relation fields in the generated schema. This can be achieved by setting the keepRelationFields option to "false" and the keepRelationScalarFields option to "true". For example if you want to use the generated schema to validate POST request object for instance, you might want to use this option.

`prisma
datasource db {
provider = "postgresql"
url = env("DATABASE_URL")
}

generator jsonSchema {
provider = "prisma-json-schema-generator"
keepRelationScalarFields = "true" // default is "false"
keepRelationFields = "false" // default is "true"
}

model User {
id Int @id @default(autoincrement())
createdAt DateTime @default(now())
email String @unique
weight Float?
is18 Boolean?
name String?
number BigInt @default(34534535435353)
favouriteDecimal Decimal
bytes Bytes
successorId Int? @unique
successor User? @relation("BlogOwnerHistory", fields: [successorId], references: [id])
predecessor User? @relation("BlogOwnerHistory")
role Role @default(USER)
posts Post[]
keywords String[]
biography Json
}

model Post {
id Int @id @default(autoincrement())
user User? @relation(fields: [userId], references: [id])
userId Int?
}

enum Role {
USER
ADMIN
}
`

Output:

`javascript
{
$schema: 'http://json-schema.org/draft-07/schema#',
definitions: {
Post: {
properties: {
id: { type: 'integer' },
userId: { type: ['integer', 'null'] },
},
type: 'object',
},
User: {
properties: {
biography: {
type: [
'number',
'string',
'boolean',
'object',
'array',
'null'
],
},
createdAt: { format: 'date-time', type: 'string' },
email: {
description: 'Triple Slash Comment: Will show up in JSON schema [EMAIL]',
type: 'string'
},
id: { type: 'integer' },
is18: { type: ['boolean', 'null'] },
keywords: { items: { type: 'string' }, type: 'array' },
name: { type: ['string', 'null'] },
number: { type: 'integer', default: '34534535435353' },
bytes: {
description: 'Triple Slash Inline Comment: Will show up in JSON schema [BYTES]',
type: 'string'
},
favouriteDecimal: { type: 'number' },
role: { enum: ['USER', 'ADMIN'], type: 'string', default: 'USER' },
successorId: { type: ['integer', 'null'] },
weight: { type: ['integer', 'null'] },
},
type: 'object',
},
},
properties: {
post: { $ref: '#/definitions/Post' },
user: { $ref: '#/definitions/User' },
},
type: 'object',
}
`

License: MIT

Copyright (c) 2020 Valentin Palkovič

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

`

``