Hooks

Lifecycle hooks let you intercept and transform input, cancel execution, handle errors, and clean up on exit.

Usage

const h = term
    .hook()
    .beforeParse()
    .do((input: string) => (input === 'h' ? 'help' : input));

h.dispose(); // unregister (idempotent)

Events

Selector	Callback	Purpose
.beforeParse()	(input: string) => string	Transform raw input before tokenizing
.afterParse()	(tokens: string[]) => string[]	Rewrite token list after parsing
.onStart()	() => void \| Promise<void>	Run before the first prompt on start
.beforeParse()	(input: string) => string	Transform raw input before tokenizing
.afterParse()	(tokens: string[]) => string[]	Rewrite token list after parsing
.beforeExecute()	(command, ctx, args) => void \| false	Cancel execution — return false to skip
.afterExecute()	(result: unknown) => void	Post-process after command returns
.beforeExit()	() => void \| Promise<void>	Run before cleanup when terminal stops
.onStop()	() => void \| Promise<void>	Run after cleanup when terminal stops
.onError()	(error: Error) => void \| boolean	Return true to suppress default error output

Hooks execute in registration order.

Error handling

Errors from any hook propagate to handleError (see Architecture). onError hooks run first — the first returning true suppresses the error. If an onError callback itself throws, it's caught individually and remaining hooks still run. The terminal loop never crashes.

dispose() {#dispose}

Every .do() returns a Hook. Call .dispose() to unregister. Safe to call multiple times.

---

Terminal — class reference, register, start/stop
Commands — defining commands
Arguments — typed accessors