GenerateTts
generateTts
PLANNED FOR UPDATE 0.1.7
This function is planned for G1R:MP 0.1.7. It is not included in the public 0.1.6 release.
Queues asynchronous speech rendering with persistent cache reuse; generateTts and renderTts are aliases.
Syntax
number|false generateTts(string text, table options)
Parameters
| Name | Type | Required | Description |
|---|---|---|---|
text |
string |
yes | Nonempty UTF-8 text: at most 12000 bytes and 4000 Unicode characters. |
options |
table |
yes | engine + language, or an existing voice/preset (not both). Optional seed, reference/hash, parameters and saveAs. See Server TTS and TTS voices. |
Returns
A resource-owned job ID, or false, reason. Acceptance is not completion: wait for onTtsReady or onTtsError.
Notes
- Server-side only. Coming in 0.1.7 — not released yet; not included in public 0.1.6.
- See Server TTS, TTS configuration and TTS voices for installation, options, ownership, cache and examples.
- Invalid requests return false, reason. Profiles, jobs and playback handles belong to the calling resource; do not use handles from another resource or a previous resource lifetime.
- The exact text/options/model revision/reference hash form the cache request. Identical requests are reused. saveAs creates a stable resource-private name; a different recording cannot silently overwrite it.
- Overrides replace top-level fields; replacing parameters uses defaults for omitted parameter keys. Release completed job handles with releaseTtsJob; stored recordings remain.