Generate typescript definitions from your pocketbase.io schema.

npx pocketbase-typegen --db ./pb_data/data.db --out pocketbase-types.ts

This will produce types for all your PocketBase collections to use in your frontend typescript codebase.

Options:
  -V, --version          output the version number
  -d, --db <char>        path to the pocketbase SQLite database
  -j, --json <char>      path to JSON schema exported from pocketbase admin UI
  -u, --url <char>       URL to your hosted pocketbase instance. When using this options you must also provide email and password options.
  --email <char>     email for an admin pocketbase user. Use this with the --url option
  -p, --password <char>  password for an admin pocketbase user. Use this with the --url option
  -o, --out <char>       path to save the typescript output file (default: "pocketbase-types.ts")
  --no-sdk               remove the pocketbase package dependency. A typed version of the SDK will not be generated.
  --env [path]       flag to use environment variables for configuration. Add PB_TYPEGEN_URL, PB_TYPEGEN_EMAIL, PB_TYPEGEN_PASSWORD to your .env file. Optionally provide a path to your .env file
  -h, --help             display help for command

DB example:

npx pocketbase-typegen --db ./pb_data/data.db

JSON example (export JSON schema from the pocketbase admin dashboard):

npx pocketbase-typegen --json ./pb_schema.json

URL example:

npx pocketbase-typegen --url https://myproject.pockethost.io --email admin@myproject.com --password 'secr3tp@ssword!'

ENV example (add PB_TYPEGEN_URL, PB_TYPEGEN_EMAIL and PB_TYPEGEN_PASSWORD to your .env file):

npx pocketbase-typegen --env

.env:

PB_TYPEGEN_URL=https://myproject.pockethost.io
PB_TYPEGEN_EMAIL=admin@myproject.com
PB_TYPEGEN_PASSWORD=secr3tp@ssword!

Add it to your projects package.json:

"scripts": {
  "typegen": "pocketbase-typegen --db ./pb_data/data.db",
},

The output is a typescript file pocketbase-types.ts (example) which will contain:

• Collections An enum of all collections.
• [CollectionName]Record One type for each collection (eg ProfilesRecord).
• [CollectionName]Response One response type for each collection (eg ProfilesResponse) which includes system fields. This is what is returned from the PocketBase API.
  • [CollectionName][FieldName]Options If the collection contains a select field with set values, an enum of the options will be generated.
• CollectionRecords A type mapping each collection name to the record type.
• CollectionResponses A type mapping each collection name to the response type.
• TypedPocketBase A type for usage with type asserted PocketBase instance.

Using PocketBase SDK v0.18.3+, collections can be automatically typed using the generated TypedPocketBase type:

import { TypedPocketBase } from "./pocketbase-types"

const pb = new PocketBase("http://127.0.0.1:8090") as TypedPocketBase

await pb.collection("tasks").getOne("RECORD_ID") // -> results in TaskResponse
await pb.collection("posts").getOne("RECORD_ID") // -> results in PostResponse

Alternatively, you can use generic types for each request, eg:

import { Collections, TasksResponse } from "./pocketbase-types"

await pb.collection(Collections.Tasks).getOne<TasksResponse>("RECORD_ID") // -> results in TaskResponse

You can provide types for JSON fields and expanded relations by passing generic arguments to the Response types:

import { Collections, CommentsResponse, UserResponse } from "./pocketbase-types"

/**
  type CommentsRecord<Tmetadata = unknown> = {
    text: string
    metadata: null | Tmetadata // This is a json field
    user: RecordIdString // This is a relation field
  }
*/
type Metadata = {
  likes: number
}
type Expand = {
  user: UsersResponse
}
const result = await pb
  .collection(Collections.Comments)
  .getOne<CommentsResponse<Metadata, Expand>>("RECORD_ID", { expand: "user" })

// Now you can access the expanded relation with type safety and hints in your IDE
result.expand?.user.username