yomichan/ext/mixed/js/audio.js

297 lines
8.8 KiB
JavaScript
Raw Normal View History

2019-10-10 23:58:06 +00:00
/*
2020-01-01 17:00:00 +00:00
* Copyright (C) 2019-2020 Alex Yatskov <alex@foosoft.net>
2019-10-10 23:58:06 +00:00
* Author: Alex Yatskov <alex@foosoft.net>
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
2020-01-01 17:00:31 +00:00
* along with this program. If not, see <https://www.gnu.org/licenses/>.
2019-10-10 23:58:06 +00:00
*/
2020-02-01 20:00:34 +00:00
/*global apiAudioGetUrl*/
2019-10-10 23:58:06 +00:00
class TextToSpeechAudio {
constructor(text, voice) {
this.text = text;
this.voice = voice;
this._utterance = null;
this._volume = 1;
}
get currentTime() {
return 0;
}
set currentTime(value) {
// NOP
}
get volume() {
return this._volume;
}
set volume(value) {
this._volume = value;
if (this._utterance !== null) {
this._utterance.volume = value;
}
}
play() {
try {
if (this._utterance === null) {
this._utterance = new SpeechSynthesisUtterance(this.text || '');
this._utterance.lang = 'ja-JP';
this._utterance.volume = this._volume;
this._utterance.voice = this.voice;
}
speechSynthesis.cancel();
speechSynthesis.speak(this._utterance);
} catch (e) {
// NOP
}
}
pause() {
try {
speechSynthesis.cancel();
} catch (e) {
// NOP
}
}
static createFromUri(ttsUri) {
const m = /^tts:[^#?]*\?([^#]*)/.exec(ttsUri);
if (m === null) { return null; }
2020-02-15 18:36:32 +00:00
const searchParameters = new URLSearchParams(m[1]);
const text = searchParameters.get('text');
let voice = searchParameters.get('voice');
if (text === null || voice === null) { return null; }
2020-02-15 18:36:32 +00:00
voice = audioGetTextToSpeechVoice(voice);
if (voice === null) { return null; }
2020-02-15 18:36:32 +00:00
return new TextToSpeechAudio(text, voice);
}
}
2020-03-07 17:44:14 +00:00
class AudioSystem {
constructor() {
this._cache = new Map();
this._cacheSizeMaximum = 32;
if (typeof speechSynthesis !== 'undefined') {
// speechSynthesis.getVoices() will not be populated unless some API call is made.
speechSynthesis.addEventListener('voiceschanged', this._onVoicesChanged.bind(this));
}
}
async getExpressionAudio(expression, sources, optionsContext, details) {
const key = `${expression.expression}:${expression.reading}`;
const cacheValue = this._cache.get(expression);
if (typeof cacheValue !== 'undefined') {
const {audio, uri, source} = cacheValue;
return {audio, uri, source};
}
for (const source of sources) {
const uri = await apiAudioGetUrl(expression, source, optionsContext);
if (uri === null) { continue; }
try {
const audio = await this._createAudio(uri, details);
this._cacheCheck();
this._cache.set(key, {audio, uri, source});
return {audio, uri, source};
} catch (e) {
// NOP
}
}
throw new Error('Could not create audio');
}
createTextToSpeechAudio({text, voiceUri}) {
const voice = this._getTextToSpeechVoiceFromVoiceUri(voiceUri);
if (voice === null) {
throw new Error('Invalid text-to-speech voice');
}
return new TextToSpeechAudio(text, voice);
}
_onVoicesChanged() {
// NOP
}
async _createAudio(uri, details) {
const ttsParameters = this._getTextToSpeechParameters(uri);
if (ttsParameters !== null) {
if (typeof details === 'object' && details !== null) {
if (details.tts === false) {
throw new Error('Text-to-speech not permitted');
}
}
return this.createTextToSpeechAudio(ttsParameters);
}
return await this._createAudioFromUrl(uri);
}
_createAudioFromUrl(url) {
return new Promise((resolve, reject) => {
const audio = new Audio(url);
audio.addEventListener('loadeddata', () => {
const duration = audio.duration;
if (duration === 5.694694 || duration === 5.720718) {
// Hardcoded values for invalid audio
reject(new Error('Could not retrieve audio'));
} else {
resolve(audio);
}
});
audio.addEventListener('error', () => reject(audio.error));
});
}
_getTextToSpeechVoiceFromVoiceUri(voiceUri) {
try {
for (const voice of speechSynthesis.getVoices()) {
if (voice.voiceURI === voiceUri) {
return voice;
}
}
} catch (e) {
// NOP
}
return null;
}
_getTextToSpeechParameters(uri) {
const m = /^tts:[^#?]*\?([^#]*)/.exec(uri);
if (m === null) { return null; }
const searchParameters = new URLSearchParams(m[1]);
const text = searchParameters.get('text');
const voiceUri = searchParameters.get('voice');
return (text !== null && voiceUri !== null ? {text, voiceUri} : null);
}
_cacheCheck() {
const removeCount = this._cache.size - this._cacheSizeMaximum;
if (removeCount <= 0) { return; }
const removeKeys = [];
for (const key of this._cache.keys()) {
removeKeys.push(key);
if (removeKeys.length >= removeCount) { break; }
}
for (const key of removeKeys) {
this._cache.delete(key);
}
}
}
function audioGetFromUrl(url, willDownload) {
const tts = TextToSpeechAudio.createFromUri(url);
if (tts !== null) {
if (willDownload) {
throw new Error('AnkiConnect does not support downloading text-to-speech audio.');
}
return Promise.resolve(tts);
}
2019-10-10 23:58:06 +00:00
return new Promise((resolve, reject) => {
const audio = new Audio(url);
audio.addEventListener('loadeddata', () => {
if (audio.duration === 5.694694 || audio.duration === 5.720718) {
// Hardcoded values for invalid audio
reject(new Error('Could not retrieve audio'));
} else {
resolve(audio);
}
});
audio.addEventListener('error', () => reject(audio.error));
});
}
async function audioGetFromSources(expression, sources, optionsContext, willDownload, cache=null) {
2019-10-10 23:58:06 +00:00
const key = `${expression.expression}:${expression.reading}`;
if (cache !== null) {
const cacheValue = cache.get(expression);
if (typeof cacheValue !== 'undefined') {
return cacheValue;
}
2019-10-10 23:58:06 +00:00
}
for (let i = 0, ii = sources.length; i < ii; ++i) {
const source = sources[i];
const url = await apiAudioGetUrl(expression, source, optionsContext);
if (url === null) {
continue;
}
try {
2019-11-10 18:56:06 +00:00
let audio = await audioGetFromUrl(url, willDownload);
if (willDownload) {
// AnkiConnect handles downloading URLs into cards
2019-11-10 18:56:06 +00:00
audio = null;
}
2019-10-10 23:58:06 +00:00
const result = {audio, url, source};
if (cache !== null) {
cache.set(key, result);
2019-10-10 23:58:06 +00:00
}
return result;
} catch (e) {
// NOP
}
}
return {audio: null, url: null, source: null};
2019-10-10 23:58:06 +00:00
}
2019-10-13 02:50:22 +00:00
function audioGetTextToSpeechVoice(voiceURI) {
try {
for (const voice of speechSynthesis.getVoices()) {
if (voice.voiceURI === voiceURI) {
return voice;
}
}
} catch (e) {
// NOP
}
return null;
}
function audioPrepareTextToSpeech(options) {
if (
audioPrepareTextToSpeech.state ||
!options.audio.textToSpeechVoice ||
!(
options.audio.sources.includes('text-to-speech') ||
options.audio.sources.includes('text-to-speech-reading')
)
) {
// Text-to-speech not in use.
return;
}
// Chrome needs this value called once before it will become populated.
// The first call will return an empty list.
audioPrepareTextToSpeech.state = true;
try {
speechSynthesis.getVoices();
} catch (e) {
// NOP
}
}
audioPrepareTextToSpeech.state = false;