@wcstack/speech 1.31.0 → 1.32.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.ja.md +4 -0
- package/README.md +4 -0
- package/package.json +1 -1
package/README.ja.md
CHANGED
|
@@ -265,6 +265,10 @@ core.speak("こんにちは。");
|
|
|
265
265
|
|
|
266
266
|
Core の構造サーフェスは wcstack I/O ノード横断の規範です([async-io-node-guidelines §3.9](../../docs/async-io-node-guidelines.ja.md))。要素なしで signals に束縛するには [@wcstack/signals — Core を直接束縛する](../signals/README.ja.md#core-を直接束縛する要素なし) を参照。
|
|
267
267
|
|
|
268
|
+
## アクセシビリティ
|
|
269
|
+
|
|
270
|
+
**WCAG 1.4.2 Audio Control（レベル A）**: 合成音声も音声である — ユーザー操作なしに始まりうる・長く続きうるなら、システム音量と独立の可視の停止手段（`<wcs-speak>` の `pause` / `cancel` コマンド）を必ず用意する。合成音声はスクリーンリーダー自身の声とも正面衝突する: ユーザー自身のリーダーが読み上げるべき適切なマークアップの代わりに `<wcs-speak>` を使ってはならない（リーダーは*ユーザーが選んだ*声・速度・言語で話す）。認識側の `<wcs-listen>` は入力手段であり、モーションセンサーと同じ扱いを: 同じ操作への並行入力（タイピング）と、マイクを止める可視の手段（`stop` / `abort`）を常に用意する。
|
|
271
|
+
|
|
268
272
|
## ライセンス
|
|
269
273
|
|
|
270
274
|
MIT
|
package/README.md
CHANGED
|
@@ -266,6 +266,10 @@ core.speak("Hello, world.");
|
|
|
266
266
|
|
|
267
267
|
The structural Core surface is normative across wcstack IO nodes ([async-io-node-guidelines §3.9](../../docs/async-io-node-guidelines.md)); to bind it into signals with no element at all, see [@wcstack/signals — Binding a Core directly](../signals/README.md#binding-a-core-directly-no-element).
|
|
268
268
|
|
|
269
|
+
## Accessibility
|
|
270
|
+
|
|
271
|
+
**WCAG 1.4.2 Audio Control (Level A)**: synthesized speech is audio — if it can start without a user action or run long, give the user a visible way to silence it (`<wcs-speak>`'s `pause` / `cancel` commands), independent of system volume. Synthesized speech also collides head-on with a screen reader's own voice: never use `<wcs-speak>` as a substitute for proper markup that the user's own reader would announce (their reader speaks in the voice, speed, and language *they* chose). For recognition, `<wcs-listen>` is an input method — treat it like the motion sensors: always a parallel input (typing) for the same action, and a visible way to stop the microphone (`stop` / `abort`).
|
|
272
|
+
|
|
269
273
|
## License
|
|
270
274
|
|
|
271
275
|
MIT
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@wcstack/speech",
|
|
3
|
-
"version": "1.
|
|
3
|
+
"version": "1.32.0",
|
|
4
4
|
"description": "Declarative Web Speech components for Web Components. Framework-agnostic SpeechSynthesis (TTS) and SpeechRecognition (STT) primitives via wc-bindable-protocol.",
|
|
5
5
|
"type": "module",
|
|
6
6
|
"main": "./dist/index.esm.js",
|