CLI Usage and Configuration

pgtyped CLI can be launched in build or watch mode. Watch mode is most useful for a local development workflow, while build mode can be used for generating types when running CI.

Flags

The CLI supports a number of flags:

--config config_file_path.json to pass the config file path.
--watch to start in watch mode.
--file file_path.ts if you only want to process one file (which can be useful when working on a big project). Incompatible with watch mode. Uses transforms defined in the config file to determine the mode and emit template, so a file path that doesn't fit the include glob patterns will not be processed.
--uri to specify a PG connection URI (overriding the config value).
--help for a quick flag reference.
--version to show the version number.

Example:

npx pgtyped -w -c config.json

Environment variables

PgTyped supports common PostgreSQL environment variables:

PGHOST
PGUSER
PGPASSWORD
PGDATABASE
PGPORT
PGURI or DATABASE_URL

These variables will override values provided in config.json.

Example configuration file

Below is an example configuration file, with comments explaining each field.
For a full list of options, see the Configuration file format section.

config.json
{
  // You can specify as many transforms as you want
  // Only TS and SQL files (modes) are supported at the moment
  "transforms": [
    {
      "mode": "sql", // SQL mode
      "include": "**/*.sql", // SQL files pattern to scan for queries
      "emitTemplate": "{{dir}}/{{name}}.queries.ts" // File name template to save generated files
    },
    {
      "mode": "ts", // TS mode
      "include": "**/action.ts", // TS file pattern to scan for queries
      "emitTemplate": "{{dir}}/{{name}}.types.ts" // File name template to save generated files
    }
  ],
  "srcDir": "./src/", // Directory to scan or watch for query files
  "failOnError": false, // Whether to fail on a file processing error and abort generation (can be omitted - default is false)
  "camelCaseColumnNames": false, // convert to camelCase column names of result interface
  "dbUrl": "postgres://user:password@host/database", // DB URL (optional - will be merged with db if provided)
  "db": {
    "dbName": "testdb", // DB name
    "user": "user", // DB username
    "password": "password", // DB password (optional)
    "host": "127.0.0.1", // DB host (optional)
    "port": 5432, // DB port (optional)
    "ssl": false // Whether or not to connect to DB with SSL (optional)
  },
  "typesOverrides": { // Override default Postgres => TypeScript mapping
    "date": "string", 
    "timestamptz": "string", 
    "numeric": "number", 
    "my_enum": "./path/to/enums#MyEnum" // Will import MyEnum from ./path/to/enums
  }
}

note

Configuration file can be also be written in CommonJS format and default exported as an object. If your project is of ESM type then you will need to give the config file a .cjs extension instead of .js.

Configuration file format

Name	Type	Description
`transforms`	`Transform[]`	An array of transforms to apply to the files.
`srcDir`	`string`	Directory to scan or watch for query files.
`db`	`DatabaseConfig`	A database config.
`failOnError?`	`boolean`	Whether to fail on a file processing error and abort generation. Default: `false`
`dbUrl?`	`string`	A connection string to the database. Example: `postgres://user:password@host/database`. Overrides (merged) with `db` config.
`camelCaseColumnNames?`	`boolean`	Whether to convert column names to camelCase. Note that this only coverts the types. You need to do this at runtime independently using a library like `pg-camelcase`.
`typesOverrides?`	`Record<string, string>`	A map of type overrides. Similarly to `camelCaseColumnNames`, this only affects the types. You need to do this at runtime independently using a library like `pg-types`.
`maxWorkerThreads`	`number`	The maximum number of worker threads to use for type generation. The default is based on the number of available CPUs.

Fields marked with ? are optional.

Transform

Name	Type	Description
`mode`	`string`	The mode to use. Can be `sql` or `ts`.
`include`	`string`	A glob pattern to match files to process. Example: `"*/.sql"`.
`emitTemplate`	`string`	A template to use for the output file name. See Customizing generated file paths for more details.

DatabaseConfig

Name	Type	Description
`host?`	`string`	The database host. Defaults to `127.0.0.1`.
`port?`	`number`	The database port. Defaults to `5432`.
`user?`	`string`	The database user. Defaults to `postgres`.
`dbName?`	`string`	The database name. Defaults to `postgres`.
`password?`	`string`	The database password. Defaults to empty string.
`ssl?`	`boolean` or `TLSConnectionOptions`	Determines whether to use SSL to connect to the database. Also accepts a TLS connection options object as defined in the Node.js socket method. More details on this in the Configuring SSL section.

Customizing generated file paths

By default, PgTyped saves generated files in the same folder as the source files it parses. This behavior can be customized using the emitTemplate config parameter. In that template, four parameters are available for interpolation: root, dir, base, name and ext. For example, when parsing source/query file /home/user/dir/file.sql, these parameters are assigned the following values:

┌─────────────────────┬────────────┐
│          dir        │    base    │
├──────┬              ├──────┬─────┤
│ root │              │ name │ ext │
"  /    home/user/dir / file  .sql "
└──────┴──────────────┴──────┴─────┘
(All spaces in the "" line should be ignored. They are purely for formatting.)

Configuring SSL

By default, if enabled it will attempt to verify the SSL connection with the local certificates on the machine.

Options can also be provided to customize the certificate used or to ignore SSL errors. More information about options can be found here.

Sample configuration files have been provided below.

custom_ca.json
{
  "transforms": [
    {
      "mode": "sql",
      "include": "**/*.sql",
      "emitTemplate": "{{dir}}/{{name}}.queries.ts"
    }
  ],
  "srcDir": "./src/",
  "failOnError": false,
  "camelCaseColumnNames": false,
  "db": {
    "dbName": "testdb",
    "user": "user",
    "host": "someremote.host.com",
    "ssl": {
      "host": "someremote.host.com",
      "port": 5432,
      "ca": ["insert CA here"]
    }
  }
}

ignore_ssl.json
{
  "transforms": [
    {
      "mode": "sql",
      "include": "**/*.sql",
      "emitTemplate": "{{dir}}/{{name}}.queries.ts"
    }
  ],
  "srcDir": "./src/",
  "failOnError": false,
  "camelCaseColumnNames": false,
  "db": {
    "dbName": "testdb",
    "user": "user",
    "host": "someremote.host.com",
    "ssl": {
      "rejectUnauthorized": false
    }
  }
}

Flags​

Environment variables​

Example configuration file​

Configuration file format​

Transform​

DatabaseConfig​

Customizing generated file paths​

Configuring SSL​