mirror of
https://github.com/SillyTavern/SillyTavern.git
synced 2025-06-05 21:59:27 +02:00
lint: Require semicolons
This commit is contained in:
Cohee
@ -4,11 +4,11 @@ TODO:
|
||||
- Delete useless call
|
||||
*/
|
||||
|
||||
import { doExtrasFetch, extension_settings, getApiUrl, modules } from '../../extensions.js'
|
||||
import { callPopup } from '../../../script.js'
|
||||
import { initVoiceMap } from './index.js'
|
||||
import { doExtrasFetch, extension_settings, getApiUrl, modules } from '../../extensions.js';
|
||||
import { callPopup } from '../../../script.js';
|
||||
import { initVoiceMap } from './index.js';
|
||||
|
||||
export { CoquiTtsProvider }
|
||||
export { CoquiTtsProvider };
|
||||
|
||||
const DEBUG_PREFIX = '<Coqui TTS module> ';
|
||||
|
||||
@ -38,11 +38,11 @@ const languageLabels = {
|
||||
'fr': 'French',
|
||||
'es': 'Spanish',
|
||||
'ja': 'Japanese'
|
||||
}
|
||||
};
|
||||
|
||||
function throwIfModuleMissing() {
|
||||
if (!modules.includes('coqui-tts')) {
|
||||
const message = 'Coqui TTS module not loaded. Add coqui-tts to enable-modules and restart the Extras API.'
|
||||
const message = 'Coqui TTS module not loaded. Add coqui-tts to enable-modules and restart the Extras API.';
|
||||
// toastr.error(message, { timeOut: 10000, extendedTimeOut: 20000, preventDuplicates: true });
|
||||
throw new Error(DEBUG_PREFIX, message);
|
||||
}
|
||||
@ -58,14 +58,14 @@ class CoquiTtsProvider {
|
||||
// Extension UI and Settings //
|
||||
//#############################//
|
||||
|
||||
settings
|
||||
settings;
|
||||
|
||||
defaultSettings = {
|
||||
voiceMap: {},
|
||||
customVoices: {},
|
||||
voiceIds: [],
|
||||
voiceMapDict: {}
|
||||
}
|
||||
};
|
||||
|
||||
get settingsHtml() {
|
||||
let html = `
|
||||
@ -119,19 +119,19 @@ class CoquiTtsProvider {
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
`
|
||||
return html
|
||||
`;
|
||||
return html;
|
||||
}
|
||||
|
||||
async loadSettings(settings) {
|
||||
// Only accept keys defined in defaultSettings
|
||||
this.settings = this.defaultSettings
|
||||
this.settings = this.defaultSettings;
|
||||
|
||||
for (const key in settings) {
|
||||
if (key in this.settings) {
|
||||
this.settings[key] = settings[key]
|
||||
this.settings[key] = settings[key];
|
||||
} else {
|
||||
throw DEBUG_PREFIX + `Invalid setting passed to extension: ${key}`
|
||||
throw DEBUG_PREFIX + `Invalid setting passed to extension: ${key}`;
|
||||
}
|
||||
}
|
||||
|
||||
@ -147,13 +147,13 @@ class CoquiTtsProvider {
|
||||
$('#coqui_api_model_install_status').hide();
|
||||
$('#coqui_api_model_install_button').hide();
|
||||
|
||||
let that = this
|
||||
$('#coqui_model_origin').on('change', function () { that.onModelOriginChange() });
|
||||
$('#coqui_api_language').on('change', function () { that.onModelLanguageChange() });
|
||||
$('#coqui_api_model_name').on('change', function () { that.onModelNameChange() });
|
||||
let that = this;
|
||||
$('#coqui_model_origin').on('change', function () { that.onModelOriginChange(); });
|
||||
$('#coqui_api_language').on('change', function () { that.onModelLanguageChange(); });
|
||||
$('#coqui_api_model_name').on('change', function () { that.onModelNameChange(); });
|
||||
|
||||
$('#coqui_remove_voiceId_mapping').on('click', function () { that.onRemoveClick() });
|
||||
$('#coqui_add_voiceId_mapping').on('click', function () { that.onAddClick() });
|
||||
$('#coqui_remove_voiceId_mapping').on('click', function () { that.onRemoveClick(); });
|
||||
$('#coqui_add_voiceId_mapping').on('click', function () { that.onAddClick(); });
|
||||
|
||||
// Load coqui-api settings from json file
|
||||
await fetch('/scripts/extensions/tts/coqui_api_models_settings.json')
|
||||
@ -198,8 +198,8 @@ class CoquiTtsProvider {
|
||||
|
||||
// Perform a simple readiness check by trying to fetch voiceIds
|
||||
async checkReady(){
|
||||
throwIfModuleMissing()
|
||||
await this.fetchTtsVoiceObjects()
|
||||
throwIfModuleMissing();
|
||||
await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
|
||||
updateCustomVoices() {
|
||||
@ -217,18 +217,18 @@ class CoquiTtsProvider {
|
||||
}
|
||||
|
||||
// Update UI select list with voices
|
||||
$('#coqui_voicename_select').empty()
|
||||
$('#coqui_voicename_select').empty();
|
||||
$('#coqui_voicename_select')
|
||||
.find('option')
|
||||
.remove()
|
||||
.end()
|
||||
.append('<option value="none">Select Voice</option>')
|
||||
.val('none')
|
||||
.val('none');
|
||||
for (const voiceName in this.settings.voiceMapDict) {
|
||||
$('#coqui_voicename_select').append(new Option(voiceName, voiceName));
|
||||
}
|
||||
|
||||
this.onSettingsChange()
|
||||
this.onSettingsChange();
|
||||
}
|
||||
|
||||
onSettingsChange() {
|
||||
@ -237,7 +237,7 @@ class CoquiTtsProvider {
|
||||
}
|
||||
|
||||
async onRefreshClick() {
|
||||
this.checkReady()
|
||||
this.checkReady();
|
||||
}
|
||||
|
||||
async onAddClick() {
|
||||
@ -246,7 +246,7 @@ class CoquiTtsProvider {
|
||||
}
|
||||
|
||||
// Ask user for voiceId name to save voice
|
||||
const voiceName = await callPopup('<h3>Name of Coqui voice to add to voice select dropdown:</h3>', 'input')
|
||||
const voiceName = await callPopup('<h3>Name of Coqui voice to add to voice select dropdown:</h3>', 'input');
|
||||
|
||||
const model_origin = $('#coqui_model_origin').val();
|
||||
const model_language = $('#coqui_api_language').val();
|
||||
@ -303,11 +303,11 @@ class CoquiTtsProvider {
|
||||
const tokens = $('#coqui_api_model_name').val().split('/');
|
||||
const model_dataset = tokens[0];
|
||||
const model_label = tokens[1];
|
||||
const model_id = 'tts_models/' + model_language + '/' + model_dataset + '/' + model_label
|
||||
const model_id = 'tts_models/' + model_language + '/' + model_dataset + '/' + model_label;
|
||||
|
||||
let modelDict = coquiApiModels
|
||||
let modelDict = coquiApiModels;
|
||||
if (model_origin == 'coqui-api-full')
|
||||
modelDict = coquiApiModelsFull
|
||||
modelDict = coquiApiModelsFull;
|
||||
|
||||
if (model_setting_language == null & 'languages' in modelDict[model_language][model_dataset][model_label]) {
|
||||
toastr.error('Model language not selected, please select one.', DEBUG_PREFIX+' voice mapping model language', { timeOut: 10000, extendedTimeOut: 20000, preventDuplicates: true });
|
||||
@ -326,7 +326,7 @@ class CoquiTtsProvider {
|
||||
console.debug(DEBUG_PREFIX, 'Registered new voice map: ', voiceName, ':', this.settings.voiceMapDict[voiceName]);
|
||||
|
||||
this.updateCustomVoices();
|
||||
initVoiceMap() // Update TTS extension voiceMap
|
||||
initVoiceMap(); // Update TTS extension voiceMap
|
||||
|
||||
let successMsg = voiceName + ':' + model_id;
|
||||
if (model_setting_language != null)
|
||||
@ -335,16 +335,16 @@ class CoquiTtsProvider {
|
||||
successMsg += '[' + model_setting_speaker + ']';
|
||||
toastr.info(successMsg, DEBUG_PREFIX + ' voice map updated', { timeOut: 10000, extendedTimeOut: 20000, preventDuplicates: true });
|
||||
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
async getVoice(voiceName) {
|
||||
let match = await this.fetchTtsVoiceObjects()
|
||||
let match = await this.fetchTtsVoiceObjects();
|
||||
match = match.filter(
|
||||
voice => voice.name == voiceName
|
||||
)[0]
|
||||
)[0];
|
||||
if (!match) {
|
||||
throw `TTS Voice name ${voiceName} not found in CoquiTTS Provider voice list`
|
||||
throw `TTS Voice name ${voiceName} not found in CoquiTTS Provider voice list`;
|
||||
}
|
||||
return match;
|
||||
}
|
||||
@ -360,11 +360,11 @@ class CoquiTtsProvider {
|
||||
// Todo erase from voicemap
|
||||
delete (this.settings.voiceMapDict[voiceName]);
|
||||
this.updateCustomVoices();
|
||||
initVoiceMap() // Update TTS extension voiceMap
|
||||
initVoiceMap(); // Update TTS extension voiceMap
|
||||
}
|
||||
|
||||
async onModelOriginChange() {
|
||||
throwIfModuleMissing()
|
||||
throwIfModuleMissing();
|
||||
resetModelSettings();
|
||||
const model_origin = $('#coqui_model_origin').val();
|
||||
|
||||
@ -385,9 +385,9 @@ class CoquiTtsProvider {
|
||||
.val('none');
|
||||
|
||||
for(let language in coquiApiModels) {
|
||||
let languageLabel = language
|
||||
let languageLabel = language;
|
||||
if (language in languageLabels)
|
||||
languageLabel = languageLabels[language]
|
||||
languageLabel = languageLabels[language];
|
||||
$('#coqui_api_language').append(new Option(languageLabel,language));
|
||||
console.log(DEBUG_PREFIX,'added language',languageLabel,'(',language,')');
|
||||
}
|
||||
@ -407,9 +407,9 @@ class CoquiTtsProvider {
|
||||
.val('none');
|
||||
|
||||
for(let language in coquiApiModelsFull) {
|
||||
let languageLabel = language
|
||||
let languageLabel = language;
|
||||
if (language in languageLabels)
|
||||
languageLabel = languageLabels[language]
|
||||
languageLabel = languageLabels[language];
|
||||
$('#coqui_api_language').append(new Option(languageLabel,language));
|
||||
console.log(DEBUG_PREFIX,'added language',languageLabel,'(',language,')');
|
||||
}
|
||||
@ -446,14 +446,14 @@ class CoquiTtsProvider {
|
||||
.append('<option value="none">Select model</option>')
|
||||
.val('none');
|
||||
|
||||
let modelDict = coquiApiModels
|
||||
let modelDict = coquiApiModels;
|
||||
if (model_origin == 'coqui-api-full')
|
||||
modelDict = coquiApiModelsFull
|
||||
modelDict = coquiApiModelsFull;
|
||||
|
||||
for(let model_dataset in modelDict[model_language])
|
||||
for(let model_name in modelDict[model_language][model_dataset]) {
|
||||
const model_id = model_dataset + '/' + model_name
|
||||
const model_label = model_name + ' (' + model_dataset + ' dataset)'
|
||||
const model_id = model_dataset + '/' + model_name;
|
||||
const model_label = model_name + ' (' + model_dataset + ' dataset)';
|
||||
$('#coqui_api_model_name').append(new Option(model_label, model_id));
|
||||
}
|
||||
}
|
||||
@ -477,11 +477,11 @@ class CoquiTtsProvider {
|
||||
const model_dataset = tokens[0];
|
||||
const model_name = tokens[1];
|
||||
|
||||
let modelDict = coquiApiModels
|
||||
let modelDict = coquiApiModels;
|
||||
if (model_origin == 'coqui-api-full')
|
||||
modelDict = coquiApiModelsFull
|
||||
modelDict = coquiApiModelsFull;
|
||||
|
||||
const model_settings = modelDict[model_language][model_dataset][model_name]
|
||||
const model_settings = modelDict[model_language][model_dataset][model_name];
|
||||
|
||||
if ('languages' in model_settings) {
|
||||
$('#coqui_api_model_settings').show();
|
||||
@ -525,22 +525,22 @@ class CoquiTtsProvider {
|
||||
$('#coqui_api_model_install_status').show();
|
||||
|
||||
// Check if already installed and propose to do it otherwise
|
||||
const model_id = modelDict[model_language][model_dataset][model_name]['id']
|
||||
const model_id = modelDict[model_language][model_dataset][model_name]['id'];
|
||||
console.debug(DEBUG_PREFIX,'Check if model is already installed',model_id);
|
||||
let result = await CoquiTtsProvider.checkmodel_state(model_id);
|
||||
result = await result.json();
|
||||
const model_state = result['model_state'];
|
||||
|
||||
console.debug(DEBUG_PREFIX, ' Model state:', model_state)
|
||||
console.debug(DEBUG_PREFIX, ' Model state:', model_state);
|
||||
|
||||
if (model_state == 'installed') {
|
||||
$('#coqui_api_model_install_status').text('Model already installed on extras server');
|
||||
$('#coqui_api_model_install_button').hide();
|
||||
}
|
||||
else {
|
||||
let action = 'download'
|
||||
let action = 'download';
|
||||
if (model_state == 'corrupted') {
|
||||
action = 'repare'
|
||||
action = 'repare';
|
||||
//toastr.error("Click install button to reinstall the model "+$("#coqui_api_model_name").find(":selected").text(), DEBUG_PREFIX+" corrupted model install", { timeOut: 10000, extendedTimeOut: 20000, preventDuplicates: true });
|
||||
$('#coqui_api_model_install_status').text('Model found but incomplete try install again (maybe still downloading)'); // (remove and download again)
|
||||
}
|
||||
@ -573,7 +573,7 @@ class CoquiTtsProvider {
|
||||
$('#coqui_api_model_install_button').show();
|
||||
}
|
||||
} catch (error) {
|
||||
console.error(error)
|
||||
console.error(error);
|
||||
toastr.error(error, DEBUG_PREFIX + ' error with model download', { timeOut: 10000, extendedTimeOut: 20000, preventDuplicates: true });
|
||||
onModelNameChange_pointer();
|
||||
}
|
||||
@ -595,7 +595,7 @@ class CoquiTtsProvider {
|
||||
Check model installation state, return one of ["installed", "corrupted", "absent"]
|
||||
*/
|
||||
static async checkmodel_state(model_id) {
|
||||
throwIfModuleMissing()
|
||||
throwIfModuleMissing();
|
||||
const url = new URL(getApiUrl());
|
||||
url.pathname = '/api/text-to-speech/coqui/coqui-api/check-model-state';
|
||||
|
||||
@ -615,11 +615,11 @@ class CoquiTtsProvider {
|
||||
throw new Error(`HTTP ${apiResult.status}: ${await apiResult.text()}`);
|
||||
}
|
||||
|
||||
return apiResult
|
||||
return apiResult;
|
||||
}
|
||||
|
||||
static async installModel(model_id, action) {
|
||||
throwIfModuleMissing()
|
||||
throwIfModuleMissing();
|
||||
const url = new URL(getApiUrl());
|
||||
url.pathname = '/api/text-to-speech/coqui/coqui-api/install-model';
|
||||
|
||||
@ -640,14 +640,14 @@ class CoquiTtsProvider {
|
||||
throw new Error(`HTTP ${apiResult.status}: ${await apiResult.text()}`);
|
||||
}
|
||||
|
||||
return apiResult
|
||||
return apiResult;
|
||||
}
|
||||
|
||||
/*
|
||||
Retrieve user custom models
|
||||
*/
|
||||
static async getLocalModelList() {
|
||||
throwIfModuleMissing()
|
||||
throwIfModuleMissing();
|
||||
const url = new URL(getApiUrl());
|
||||
url.pathname = '/api/text-to-speech/coqui/local/get-models';
|
||||
|
||||
@ -661,14 +661,14 @@ class CoquiTtsProvider {
|
||||
'model_id': 'model_id',
|
||||
'action': 'action'
|
||||
})
|
||||
})
|
||||
});
|
||||
|
||||
if (!apiResult.ok) {
|
||||
toastr.error(apiResult.statusText, DEBUG_PREFIX + ' Get local model list request failed');
|
||||
throw new Error(`HTTP ${apiResult.status}: ${await apiResult.text()}`);
|
||||
}
|
||||
|
||||
return apiResult
|
||||
return apiResult;
|
||||
}
|
||||
|
||||
|
||||
@ -677,27 +677,27 @@ class CoquiTtsProvider {
|
||||
// tts_models/en/ljspeech/glow-tts
|
||||
// ts_models/ja/kokoro/tacotron2-DDC
|
||||
async generateTts(text, voiceId) {
|
||||
throwIfModuleMissing()
|
||||
voiceId = this.settings.customVoices[voiceId]
|
||||
throwIfModuleMissing();
|
||||
voiceId = this.settings.customVoices[voiceId];
|
||||
|
||||
const url = new URL(getApiUrl());
|
||||
url.pathname = '/api/text-to-speech/coqui/generate-tts';
|
||||
|
||||
let language = 'none'
|
||||
let speaker = 'none'
|
||||
let language = 'none';
|
||||
let speaker = 'none';
|
||||
const tokens = voiceId.replaceAll(']', '').replaceAll('"', '').split('[');
|
||||
const model_id = tokens[0]
|
||||
const model_id = tokens[0];
|
||||
|
||||
console.debug(DEBUG_PREFIX, 'Preparing TTS request for', tokens)
|
||||
console.debug(DEBUG_PREFIX, 'Preparing TTS request for', tokens);
|
||||
|
||||
// First option
|
||||
if (tokens.length > 1) {
|
||||
const option1 = tokens[1]
|
||||
const option1 = tokens[1];
|
||||
|
||||
if (model_id.includes('multilingual'))
|
||||
language = option1
|
||||
language = option1;
|
||||
else
|
||||
speaker = option1
|
||||
speaker = option1;
|
||||
}
|
||||
|
||||
// Second option
|
||||
@ -723,7 +723,7 @@ class CoquiTtsProvider {
|
||||
throw new Error(`HTTP ${apiResult.status}: ${await apiResult.text()}`);
|
||||
}
|
||||
|
||||
return apiResult
|
||||
return apiResult;
|
||||
}
|
||||
|
||||
// Dirty hack to say not implemented
|
||||
@ -731,12 +731,12 @@ class CoquiTtsProvider {
|
||||
const voiceIds = Object
|
||||
.keys(this.settings.voiceMapDict)
|
||||
.map(voice => ({ name: voice, voice_id: voice, preview_url: false }));
|
||||
return voiceIds
|
||||
return voiceIds;
|
||||
}
|
||||
|
||||
// Do nothing
|
||||
previewTtsVoice(id) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
async fetchTtsFromHistory(history_item_id) {
|
||||
@ -746,7 +746,7 @@ class CoquiTtsProvider {
|
||||
|
||||
async function initLocalModels() {
|
||||
if (!modules.includes('coqui-tts'))
|
||||
return
|
||||
return;
|
||||
|
||||
// Initialized local model once
|
||||
if (!coquiLocalModelsReceived) {
|
||||
|
||||
@ -1,73 +1,73 @@
|
||||
import { getRequestHeaders } from '../../../script.js'
|
||||
import { getApiUrl } from '../../extensions.js'
|
||||
import { doExtrasFetch, modules } from '../../extensions.js'
|
||||
import { getPreviewString } from './index.js'
|
||||
import { saveTtsProviderSettings } from './index.js'
|
||||
import { getRequestHeaders } from '../../../script.js';
|
||||
import { getApiUrl } from '../../extensions.js';
|
||||
import { doExtrasFetch, modules } from '../../extensions.js';
|
||||
import { getPreviewString } from './index.js';
|
||||
import { saveTtsProviderSettings } from './index.js';
|
||||
|
||||
export { EdgeTtsProvider }
|
||||
export { EdgeTtsProvider };
|
||||
|
||||
class EdgeTtsProvider {
|
||||
//########//
|
||||
// Config //
|
||||
//########//
|
||||
|
||||
settings
|
||||
voices = []
|
||||
separator = ' . '
|
||||
audioElement = document.createElement('audio')
|
||||
settings;
|
||||
voices = [];
|
||||
separator = ' . ';
|
||||
audioElement = document.createElement('audio');
|
||||
|
||||
defaultSettings = {
|
||||
voiceMap: {},
|
||||
rate: 0,
|
||||
}
|
||||
};
|
||||
|
||||
get settingsHtml() {
|
||||
let html = `Microsoft Edge TTS Provider<br>
|
||||
<label for="edge_tts_rate">Rate: <span id="edge_tts_rate_output"></span></label>
|
||||
<input id="edge_tts_rate" type="range" value="${this.defaultSettings.rate}" min="-100" max="100" step="1" />`
|
||||
return html
|
||||
<input id="edge_tts_rate" type="range" value="${this.defaultSettings.rate}" min="-100" max="100" step="1" />`;
|
||||
return html;
|
||||
}
|
||||
|
||||
onSettingsChange() {
|
||||
this.settings.rate = Number($('#edge_tts_rate').val());
|
||||
$('#edge_tts_rate_output').text(this.settings.rate);
|
||||
saveTtsProviderSettings()
|
||||
saveTtsProviderSettings();
|
||||
}
|
||||
|
||||
async loadSettings(settings) {
|
||||
// Pupulate Provider UI given input settings
|
||||
if (Object.keys(settings).length == 0) {
|
||||
console.info('Using default TTS Provider settings')
|
||||
console.info('Using default TTS Provider settings');
|
||||
}
|
||||
|
||||
// Only accept keys defined in defaultSettings
|
||||
this.settings = this.defaultSettings
|
||||
this.settings = this.defaultSettings;
|
||||
|
||||
for (const key in settings) {
|
||||
if (key in this.settings) {
|
||||
this.settings[key] = settings[key]
|
||||
this.settings[key] = settings[key];
|
||||
} else {
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`;
|
||||
}
|
||||
}
|
||||
|
||||
$('#edge_tts_rate').val(this.settings.rate || 0);
|
||||
$('#edge_tts_rate_output').text(this.settings.rate || 0);
|
||||
$('#edge_tts_rate').on('input', () => {this.onSettingsChange()})
|
||||
await this.checkReady()
|
||||
$('#edge_tts_rate').on('input', () => {this.onSettingsChange();});
|
||||
await this.checkReady();
|
||||
|
||||
console.debug('EdgeTTS: Settings loaded')
|
||||
console.debug('EdgeTTS: Settings loaded');
|
||||
}
|
||||
|
||||
|
||||
// Perform a simple readiness check by trying to fetch voiceIds
|
||||
async checkReady(){
|
||||
throwIfModuleMissing()
|
||||
await this.fetchTtsVoiceObjects()
|
||||
throwIfModuleMissing();
|
||||
await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
|
||||
async onRefreshClick() {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
//#################//
|
||||
@ -76,39 +76,39 @@ class EdgeTtsProvider {
|
||||
|
||||
async getVoice(voiceName) {
|
||||
if (this.voices.length == 0) {
|
||||
this.voices = await this.fetchTtsVoiceObjects()
|
||||
this.voices = await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
const match = this.voices.filter(
|
||||
voice => voice.name == voiceName
|
||||
)[0]
|
||||
)[0];
|
||||
if (!match) {
|
||||
throw `TTS Voice name ${voiceName} not found`
|
||||
throw `TTS Voice name ${voiceName} not found`;
|
||||
}
|
||||
return match
|
||||
return match;
|
||||
}
|
||||
|
||||
async generateTts(text, voiceId) {
|
||||
const response = await this.fetchTtsGeneration(text, voiceId)
|
||||
return response
|
||||
const response = await this.fetchTtsGeneration(text, voiceId);
|
||||
return response;
|
||||
}
|
||||
|
||||
//###########//
|
||||
// API CALLS //
|
||||
//###########//
|
||||
async fetchTtsVoiceObjects() {
|
||||
throwIfModuleMissing()
|
||||
throwIfModuleMissing();
|
||||
|
||||
const url = new URL(getApiUrl());
|
||||
url.pathname = '/api/edge-tts/list'
|
||||
const response = await doExtrasFetch(url)
|
||||
url.pathname = '/api/edge-tts/list';
|
||||
const response = await doExtrasFetch(url);
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`)
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
let responseJson = await response.json()
|
||||
let responseJson = await response.json();
|
||||
responseJson = responseJson
|
||||
.sort((a, b) => a.Locale.localeCompare(b.Locale) || a.ShortName.localeCompare(b.ShortName))
|
||||
.map(x => ({ name: x.ShortName, voice_id: x.ShortName, preview_url: false, lang: x.Locale }));
|
||||
return responseJson
|
||||
return responseJson;
|
||||
}
|
||||
|
||||
|
||||
@ -117,9 +117,9 @@ class EdgeTtsProvider {
|
||||
this.audioElement.currentTime = 0;
|
||||
const voice = await this.getVoice(id);
|
||||
const text = getPreviewString(voice.lang);
|
||||
const response = await this.fetchTtsGeneration(text, id)
|
||||
const response = await this.fetchTtsGeneration(text, id);
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`)
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
|
||||
const audio = await response.blob();
|
||||
@ -129,9 +129,9 @@ class EdgeTtsProvider {
|
||||
}
|
||||
|
||||
async fetchTtsGeneration(inputText, voiceId) {
|
||||
throwIfModuleMissing()
|
||||
throwIfModuleMissing();
|
||||
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`)
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`);
|
||||
const url = new URL(getApiUrl());
|
||||
url.pathname = '/api/edge-tts/generate';
|
||||
const response = await doExtrasFetch(url,
|
||||
@ -144,19 +144,19 @@ class EdgeTtsProvider {
|
||||
'rate': Number(this.settings.rate),
|
||||
})
|
||||
}
|
||||
)
|
||||
);
|
||||
if (!response.ok) {
|
||||
toastr.error(response.statusText, 'TTS Generation Failed');
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
return response
|
||||
return response;
|
||||
}
|
||||
}
|
||||
function throwIfModuleMissing() {
|
||||
if (!modules.includes('edge-tts')) {
|
||||
const message = 'Edge TTS module not loaded. Add edge-tts to enable-modules and restart the Extras API.'
|
||||
const message = 'Edge TTS module not loaded. Add edge-tts to enable-modules and restart the Extras API.';
|
||||
// toastr.error(message)
|
||||
throw new Error(message)
|
||||
throw new Error(message);
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
@ -1,14 +1,14 @@
|
||||
import { saveTtsProviderSettings } from './index.js'
|
||||
export { ElevenLabsTtsProvider }
|
||||
import { saveTtsProviderSettings } from './index.js';
|
||||
export { ElevenLabsTtsProvider };
|
||||
|
||||
class ElevenLabsTtsProvider {
|
||||
//########//
|
||||
// Config //
|
||||
//########//
|
||||
|
||||
settings
|
||||
voices = []
|
||||
separator = ' ... ... ... '
|
||||
settings;
|
||||
voices = [];
|
||||
separator = ' ... ... ... ';
|
||||
|
||||
|
||||
defaultSettings = {
|
||||
@ -17,7 +17,7 @@ class ElevenLabsTtsProvider {
|
||||
apiKey: '',
|
||||
model: 'eleven_monolingual_v1',
|
||||
voiceMap: {}
|
||||
}
|
||||
};
|
||||
|
||||
get settingsHtml() {
|
||||
let html = `
|
||||
@ -36,28 +36,28 @@ class ElevenLabsTtsProvider {
|
||||
<label for="elevenlabs_tts_similarity_boost">Similarity Boost: <span id="elevenlabs_tts_similarity_boost_output"></span></label>
|
||||
<input id="elevenlabs_tts_similarity_boost" type="range" value="${this.defaultSettings.similarity_boost}" min="0" max="1" step="0.05" />
|
||||
</div>
|
||||
`
|
||||
return html
|
||||
`;
|
||||
return html;
|
||||
}
|
||||
|
||||
onSettingsChange() {
|
||||
// Update dynamically
|
||||
this.settings.stability = $('#elevenlabs_tts_stability').val()
|
||||
this.settings.similarity_boost = $('#elevenlabs_tts_similarity_boost').val()
|
||||
this.settings.model = $('#elevenlabs_tts_model').find(':selected').val()
|
||||
this.settings.stability = $('#elevenlabs_tts_stability').val();
|
||||
this.settings.similarity_boost = $('#elevenlabs_tts_similarity_boost').val();
|
||||
this.settings.model = $('#elevenlabs_tts_model').find(':selected').val();
|
||||
$('#elevenlabs_tts_stability_output').text(this.settings.stability);
|
||||
$('#elevenlabs_tts_similarity_boost_output').text(this.settings.similarity_boost);
|
||||
saveTtsProviderSettings()
|
||||
saveTtsProviderSettings();
|
||||
}
|
||||
|
||||
async loadSettings(settings) {
|
||||
// Pupulate Provider UI given input settings
|
||||
if (Object.keys(settings).length == 0) {
|
||||
console.info('Using default TTS Provider settings')
|
||||
console.info('Using default TTS Provider settings');
|
||||
}
|
||||
|
||||
// Only accept keys defined in defaultSettings
|
||||
this.settings = this.defaultSettings
|
||||
this.settings = this.defaultSettings;
|
||||
|
||||
// Migrate old settings
|
||||
if (settings['multilingual'] !== undefined) {
|
||||
@ -67,34 +67,34 @@ class ElevenLabsTtsProvider {
|
||||
|
||||
for (const key in settings) {
|
||||
if (key in this.settings) {
|
||||
this.settings[key] = settings[key]
|
||||
this.settings[key] = settings[key];
|
||||
} else {
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`;
|
||||
}
|
||||
}
|
||||
|
||||
$('#elevenlabs_tts_stability').val(this.settings.stability)
|
||||
$('#elevenlabs_tts_similarity_boost').val(this.settings.similarity_boost)
|
||||
$('#elevenlabs_tts_api_key').val(this.settings.apiKey)
|
||||
$('#elevenlabs_tts_stability').val(this.settings.stability);
|
||||
$('#elevenlabs_tts_similarity_boost').val(this.settings.similarity_boost);
|
||||
$('#elevenlabs_tts_api_key').val(this.settings.apiKey);
|
||||
$('#elevenlabs_tts_model').val(this.settings.model);
|
||||
$('#eleven_labs_connect').on('click', () => { this.onConnectClick() })
|
||||
$('#elevenlabs_tts_similarity_boost').on('input', this.onSettingsChange.bind(this))
|
||||
$('#elevenlabs_tts_stability').on('input', this.onSettingsChange.bind(this))
|
||||
$('#elevenlabs_tts_model').on('change', this.onSettingsChange.bind(this))
|
||||
$('#eleven_labs_connect').on('click', () => { this.onConnectClick(); });
|
||||
$('#elevenlabs_tts_similarity_boost').on('input', this.onSettingsChange.bind(this));
|
||||
$('#elevenlabs_tts_stability').on('input', this.onSettingsChange.bind(this));
|
||||
$('#elevenlabs_tts_model').on('change', this.onSettingsChange.bind(this));
|
||||
$('#elevenlabs_tts_stability_output').text(this.settings.stability);
|
||||
$('#elevenlabs_tts_similarity_boost_output').text(this.settings.similarity_boost);
|
||||
|
||||
try {
|
||||
await this.checkReady()
|
||||
console.debug('ElevenLabs: Settings loaded')
|
||||
await this.checkReady();
|
||||
console.debug('ElevenLabs: Settings loaded');
|
||||
} catch {
|
||||
console.debug('ElevenLabs: Settings loaded, but not ready')
|
||||
console.debug('ElevenLabs: Settings loaded, but not ready');
|
||||
}
|
||||
}
|
||||
|
||||
// Perform a simple readiness check by trying to fetch voiceIds
|
||||
async checkReady() {
|
||||
await this.fetchTtsVoiceObjects()
|
||||
await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
|
||||
async onRefreshClick() {
|
||||
@ -103,21 +103,21 @@ class ElevenLabsTtsProvider {
|
||||
async onConnectClick() {
|
||||
// Update on Apply click
|
||||
return await this.updateApiKey().catch((error) => {
|
||||
toastr.error(`ElevenLabs: ${error}`)
|
||||
})
|
||||
toastr.error(`ElevenLabs: ${error}`);
|
||||
});
|
||||
}
|
||||
|
||||
|
||||
async updateApiKey() {
|
||||
// Using this call to validate API key
|
||||
this.settings.apiKey = $('#elevenlabs_tts_api_key').val()
|
||||
this.settings.apiKey = $('#elevenlabs_tts_api_key').val();
|
||||
|
||||
await this.fetchTtsVoiceObjects().catch(error => {
|
||||
throw 'TTS API key validation failed'
|
||||
})
|
||||
console.debug(`Saved new API_KEY: ${this.settings.apiKey}`)
|
||||
$('#tts_status').text('')
|
||||
this.onSettingsChange()
|
||||
throw 'TTS API key validation failed';
|
||||
});
|
||||
console.debug(`Saved new API_KEY: ${this.settings.apiKey}`);
|
||||
$('#tts_status').text('');
|
||||
this.onSettingsChange();
|
||||
}
|
||||
|
||||
//#################//
|
||||
@ -126,30 +126,30 @@ class ElevenLabsTtsProvider {
|
||||
|
||||
async getVoice(voiceName) {
|
||||
if (this.voices.length == 0) {
|
||||
this.voices = await this.fetchTtsVoiceObjects()
|
||||
this.voices = await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
const match = this.voices.filter(
|
||||
elevenVoice => elevenVoice.name == voiceName
|
||||
)[0]
|
||||
)[0];
|
||||
if (!match) {
|
||||
throw `TTS Voice name ${voiceName} not found in ElevenLabs account`
|
||||
throw `TTS Voice name ${voiceName} not found in ElevenLabs account`;
|
||||
}
|
||||
return match
|
||||
return match;
|
||||
}
|
||||
|
||||
|
||||
async generateTts(text, voiceId) {
|
||||
const historyId = await this.findTtsGenerationInHistory(text, voiceId)
|
||||
const historyId = await this.findTtsGenerationInHistory(text, voiceId);
|
||||
|
||||
let response
|
||||
let response;
|
||||
if (historyId) {
|
||||
console.debug(`Found existing TTS generation with id ${historyId}`)
|
||||
response = await this.fetchTtsFromHistory(historyId)
|
||||
console.debug(`Found existing TTS generation with id ${historyId}`);
|
||||
response = await this.fetchTtsFromHistory(historyId);
|
||||
} else {
|
||||
console.debug('No existing TTS generation found, requesting new generation')
|
||||
response = await this.fetchTtsGeneration(text, voiceId)
|
||||
console.debug('No existing TTS generation found, requesting new generation');
|
||||
response = await this.fetchTtsGeneration(text, voiceId);
|
||||
}
|
||||
return response
|
||||
return response;
|
||||
}
|
||||
|
||||
//###################//
|
||||
@ -157,16 +157,16 @@ class ElevenLabsTtsProvider {
|
||||
//###################//
|
||||
|
||||
async findTtsGenerationInHistory(message, voiceId) {
|
||||
const ttsHistory = await this.fetchTtsHistory()
|
||||
const ttsHistory = await this.fetchTtsHistory();
|
||||
for (const history of ttsHistory) {
|
||||
const text = history.text
|
||||
const itemId = history.history_item_id
|
||||
const text = history.text;
|
||||
const itemId = history.history_item_id;
|
||||
if (message === text && history.voice_id == voiceId) {
|
||||
console.info(`Existing TTS history item ${itemId} found: ${text} `)
|
||||
return itemId
|
||||
console.info(`Existing TTS history item ${itemId} found: ${text} `);
|
||||
return itemId;
|
||||
}
|
||||
}
|
||||
return ''
|
||||
return '';
|
||||
}
|
||||
|
||||
|
||||
@ -176,36 +176,36 @@ class ElevenLabsTtsProvider {
|
||||
async fetchTtsVoiceObjects() {
|
||||
const headers = {
|
||||
'xi-api-key': this.settings.apiKey
|
||||
}
|
||||
};
|
||||
const response = await fetch('https://api.elevenlabs.io/v1/voices', {
|
||||
headers: headers
|
||||
})
|
||||
});
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`)
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
const responseJson = await response.json()
|
||||
return responseJson.voices
|
||||
const responseJson = await response.json();
|
||||
return responseJson.voices;
|
||||
}
|
||||
|
||||
async fetchTtsVoiceSettings() {
|
||||
const headers = {
|
||||
'xi-api-key': this.settings.apiKey
|
||||
}
|
||||
};
|
||||
const response = await fetch(
|
||||
'https://api.elevenlabs.io/v1/voices/settings/default',
|
||||
{
|
||||
headers: headers
|
||||
}
|
||||
)
|
||||
);
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`)
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
return response.json()
|
||||
return response.json();
|
||||
}
|
||||
|
||||
async fetchTtsGeneration(text, voiceId) {
|
||||
let model = this.settings.model ?? 'eleven_monolingual_v1';
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}, model ${model}`)
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}, model ${model}`);
|
||||
const response = await fetch(
|
||||
`https://api.elevenlabs.io/v1/text-to-speech/${voiceId}`,
|
||||
{
|
||||
@ -223,16 +223,16 @@ class ElevenLabsTtsProvider {
|
||||
},
|
||||
})
|
||||
}
|
||||
)
|
||||
);
|
||||
if (!response.ok) {
|
||||
toastr.error(response.statusText, 'TTS Generation Failed');
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
return response
|
||||
return response;
|
||||
}
|
||||
|
||||
async fetchTtsFromHistory(history_item_id) {
|
||||
console.info(`Fetched existing TTS with history_item_id ${history_item_id}`)
|
||||
console.info(`Fetched existing TTS with history_item_id ${history_item_id}`);
|
||||
const response = await fetch(
|
||||
`https://api.elevenlabs.io/v1/history/${history_item_id}/audio`,
|
||||
{
|
||||
@ -240,24 +240,24 @@ class ElevenLabsTtsProvider {
|
||||
'xi-api-key': this.settings.apiKey
|
||||
}
|
||||
}
|
||||
)
|
||||
);
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`)
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
return response
|
||||
return response;
|
||||
}
|
||||
|
||||
async fetchTtsHistory() {
|
||||
const headers = {
|
||||
'xi-api-key': this.settings.apiKey
|
||||
}
|
||||
};
|
||||
const response = await fetch('https://api.elevenlabs.io/v1/history', {
|
||||
headers: headers
|
||||
})
|
||||
});
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`)
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
const responseJson = await response.json()
|
||||
return responseJson.history
|
||||
const responseJson = await response.json();
|
||||
return responseJson.history;
|
||||
}
|
||||
}
|
||||
|
||||
@ -1,25 +1,25 @@
|
||||
import { callPopup, cancelTtsPlay, eventSource, event_types, name2, saveSettingsDebounced } from '../../../script.js'
|
||||
import { ModuleWorkerWrapper, doExtrasFetch, extension_settings, getApiUrl, getContext, modules } from '../../extensions.js'
|
||||
import { escapeRegex, getStringHash } from '../../utils.js'
|
||||
import { EdgeTtsProvider } from './edge.js'
|
||||
import { ElevenLabsTtsProvider } from './elevenlabs.js'
|
||||
import { SileroTtsProvider } from './silerotts.js'
|
||||
import { CoquiTtsProvider } from './coqui.js'
|
||||
import { SystemTtsProvider } from './system.js'
|
||||
import { NovelTtsProvider } from './novel.js'
|
||||
import { power_user } from '../../power-user.js'
|
||||
import { registerSlashCommand } from '../../slash-commands.js'
|
||||
import { OpenAITtsProvider } from './openai.js'
|
||||
import {XTTSTtsProvider} from './xtts.js'
|
||||
import { callPopup, cancelTtsPlay, eventSource, event_types, name2, saveSettingsDebounced } from '../../../script.js';
|
||||
import { ModuleWorkerWrapper, doExtrasFetch, extension_settings, getApiUrl, getContext, modules } from '../../extensions.js';
|
||||
import { escapeRegex, getStringHash } from '../../utils.js';
|
||||
import { EdgeTtsProvider } from './edge.js';
|
||||
import { ElevenLabsTtsProvider } from './elevenlabs.js';
|
||||
import { SileroTtsProvider } from './silerotts.js';
|
||||
import { CoquiTtsProvider } from './coqui.js';
|
||||
import { SystemTtsProvider } from './system.js';
|
||||
import { NovelTtsProvider } from './novel.js';
|
||||
import { power_user } from '../../power-user.js';
|
||||
import { registerSlashCommand } from '../../slash-commands.js';
|
||||
import { OpenAITtsProvider } from './openai.js';
|
||||
import {XTTSTtsProvider} from './xtts.js';
|
||||
export { talkingAnimation };
|
||||
|
||||
const UPDATE_INTERVAL = 1000
|
||||
const UPDATE_INTERVAL = 1000;
|
||||
|
||||
let voiceMapEntries = []
|
||||
let voiceMap = {} // {charName:voiceid, charName2:voiceid2}
|
||||
let voiceMapEntries = [];
|
||||
let voiceMap = {}; // {charName:voiceid, charName2:voiceid2}
|
||||
let storedvalue = false;
|
||||
let lastChatId = null
|
||||
let lastMessageHash = null
|
||||
let lastChatId = null;
|
||||
let lastMessageHash = null;
|
||||
|
||||
const DEFAULT_VOICE_MARKER = '[Default Voice]';
|
||||
const DISABLED_VOICE_MARKER = 'disabled';
|
||||
@ -59,8 +59,8 @@ export function getPreviewString(lang) {
|
||||
'vi-VN': 'Cô bé quàng khăn đỏ đang ngồi trên bãi cỏ xanh',
|
||||
'ar-SA': 'أَبْجَدِيَّة عَرَبِيَّة',
|
||||
'hi-IN': 'श्वेता ने श्वेता के श्वेते हाथों में श्वेता का श्वेता चावल पकड़ा',
|
||||
}
|
||||
const fallbackPreview = 'Neque porro quisquam est qui dolorem ipsum quia dolor sit amet'
|
||||
};
|
||||
const fallbackPreview = 'Neque porro quisquam est qui dolorem ipsum quia dolor sit amet';
|
||||
|
||||
return previewStrings[lang] ?? fallbackPreview;
|
||||
}
|
||||
@ -74,9 +74,9 @@ let ttsProviders = {
|
||||
Edge: EdgeTtsProvider,
|
||||
Novel: NovelTtsProvider,
|
||||
OpenAI: OpenAITtsProvider,
|
||||
}
|
||||
let ttsProvider
|
||||
let ttsProviderName
|
||||
};
|
||||
let ttsProvider;
|
||||
let ttsProviderName;
|
||||
|
||||
let ttsLastMessage = null;
|
||||
|
||||
@ -90,7 +90,7 @@ async function onNarrateOneMessage() {
|
||||
return;
|
||||
}
|
||||
|
||||
resetTtsPlayback()
|
||||
resetTtsPlayback();
|
||||
ttsJobQueue.push(message);
|
||||
moduleWorker();
|
||||
}
|
||||
@ -117,7 +117,7 @@ async function onNarrateText(args, text) {
|
||||
return;
|
||||
}
|
||||
|
||||
resetTtsPlayback()
|
||||
resetTtsPlayback();
|
||||
ttsJobQueue.push({ mes: text, name: name });
|
||||
await moduleWorker();
|
||||
|
||||
@ -127,42 +127,42 @@ async function onNarrateText(args, text) {
|
||||
|
||||
async function moduleWorker() {
|
||||
// Primarily determining when to add new chat to the TTS queue
|
||||
const enabled = $('#tts_enabled').is(':checked')
|
||||
const enabled = $('#tts_enabled').is(':checked');
|
||||
$('body').toggleClass('tts', enabled);
|
||||
if (!enabled) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
const context = getContext()
|
||||
const chat = context.chat
|
||||
const context = getContext();
|
||||
const chat = context.chat;
|
||||
|
||||
processTtsQueue()
|
||||
processAudioJobQueue()
|
||||
updateUiAudioPlayState()
|
||||
processTtsQueue();
|
||||
processAudioJobQueue();
|
||||
updateUiAudioPlayState();
|
||||
|
||||
// Auto generation is disabled
|
||||
if (extension_settings.tts.auto_generation == false) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
// no characters or group selected
|
||||
if (!context.groupId && context.characterId === undefined) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
// Chat changed
|
||||
if (
|
||||
context.chatId !== lastChatId
|
||||
) {
|
||||
currentMessageNumber = context.chat.length ? context.chat.length : 0
|
||||
saveLastValues()
|
||||
currentMessageNumber = context.chat.length ? context.chat.length : 0;
|
||||
saveLastValues();
|
||||
|
||||
// Force to speak on the first message in the new chat
|
||||
if (context.chat.length === 1) {
|
||||
lastMessageHash = -1;
|
||||
}
|
||||
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
// take the count of messages
|
||||
@ -216,13 +216,13 @@ async function moduleWorker() {
|
||||
}
|
||||
|
||||
// New messages, add new chat to history
|
||||
lastMessageHash = hashNew
|
||||
currentMessageNumber = lastMessageNumber
|
||||
lastMessageHash = hashNew;
|
||||
currentMessageNumber = lastMessageNumber;
|
||||
|
||||
console.debug(
|
||||
`Adding message from ${message.name} for TTS processing: "${message.mes}"`
|
||||
)
|
||||
ttsJobQueue.push(message)
|
||||
);
|
||||
ttsJobQueue.push(message);
|
||||
}
|
||||
|
||||
function talkingAnimation(switchValue) {
|
||||
@ -243,7 +243,7 @@ function talkingAnimation(switchValue) {
|
||||
// Handle the error here or simply ignore it to prevent logging
|
||||
}
|
||||
}
|
||||
updateUiAudioPlayState()
|
||||
updateUiAudioPlayState();
|
||||
}
|
||||
|
||||
function resetTtsPlayback() {
|
||||
@ -267,17 +267,17 @@ function resetTtsPlayback() {
|
||||
}
|
||||
|
||||
function isTtsProcessing() {
|
||||
let processing = false
|
||||
let processing = false;
|
||||
|
||||
// Check job queues
|
||||
if (ttsJobQueue.length > 0 || audioJobQueue.length > 0) {
|
||||
processing = true
|
||||
processing = true;
|
||||
}
|
||||
// Check current jobs
|
||||
if (currentTtsJob != null || currentAudioJob != null) {
|
||||
processing = true
|
||||
processing = true;
|
||||
}
|
||||
return processing
|
||||
return processing;
|
||||
}
|
||||
|
||||
function debugTtsPlayback() {
|
||||
@ -294,57 +294,57 @@ function debugTtsPlayback() {
|
||||
'currentTtsJob': currentTtsJob,
|
||||
'ttsConfig': extension_settings.tts
|
||||
}
|
||||
))
|
||||
));
|
||||
}
|
||||
window.debugTtsPlayback = debugTtsPlayback
|
||||
window.debugTtsPlayback = debugTtsPlayback;
|
||||
|
||||
//##################//
|
||||
// Audio Control //
|
||||
//##################//
|
||||
|
||||
let audioElement = new Audio()
|
||||
audioElement.id = 'tts_audio'
|
||||
audioElement.autoplay = true
|
||||
let audioElement = new Audio();
|
||||
audioElement.id = 'tts_audio';
|
||||
audioElement.autoplay = true;
|
||||
|
||||
let audioJobQueue = []
|
||||
let currentAudioJob
|
||||
let audioPaused = false
|
||||
let audioQueueProcessorReady = true
|
||||
let audioJobQueue = [];
|
||||
let currentAudioJob;
|
||||
let audioPaused = false;
|
||||
let audioQueueProcessorReady = true;
|
||||
|
||||
async function playAudioData(audioBlob) {
|
||||
// Since current audio job can be cancelled, don't playback if it is null
|
||||
if (currentAudioJob == null) {
|
||||
console.log('Cancelled TTS playback because currentAudioJob was null')
|
||||
console.log('Cancelled TTS playback because currentAudioJob was null');
|
||||
}
|
||||
const reader = new FileReader()
|
||||
const reader = new FileReader();
|
||||
reader.onload = function (e) {
|
||||
const srcUrl = e.target.result
|
||||
audioElement.src = srcUrl
|
||||
}
|
||||
reader.readAsDataURL(audioBlob)
|
||||
audioElement.addEventListener('ended', completeCurrentAudioJob)
|
||||
const srcUrl = e.target.result;
|
||||
audioElement.src = srcUrl;
|
||||
};
|
||||
reader.readAsDataURL(audioBlob);
|
||||
audioElement.addEventListener('ended', completeCurrentAudioJob);
|
||||
audioElement.addEventListener('canplay', () => {
|
||||
console.debug('Starting TTS playback')
|
||||
audioElement.play()
|
||||
})
|
||||
console.debug('Starting TTS playback');
|
||||
audioElement.play();
|
||||
});
|
||||
}
|
||||
|
||||
window['tts_preview'] = function (id) {
|
||||
const audio = document.getElementById(id)
|
||||
const audio = document.getElementById(id);
|
||||
|
||||
if (audio && !$(audio).data('disabled')) {
|
||||
audio.play()
|
||||
audio.play();
|
||||
}
|
||||
else {
|
||||
ttsProvider.previewTtsVoice(id)
|
||||
ttsProvider.previewTtsVoice(id);
|
||||
}
|
||||
}
|
||||
};
|
||||
|
||||
async function onTtsVoicesClick() {
|
||||
let popupText = ''
|
||||
let popupText = '';
|
||||
|
||||
try {
|
||||
const voiceIds = await ttsProvider.fetchTtsVoiceObjects()
|
||||
const voiceIds = await ttsProvider.fetchTtsVoiceObjects();
|
||||
|
||||
for (const voice of voiceIds) {
|
||||
popupText += `
|
||||
@ -352,27 +352,27 @@ async function onTtsVoicesClick() {
|
||||
<span class="voice_lang">${voice.lang || ''}</span>
|
||||
<b class="voice_name">${voice.name}</b>
|
||||
<i onclick="tts_preview('${voice.voice_id}')" class="fa-solid fa-play"></i>
|
||||
</div>`
|
||||
</div>`;
|
||||
if (voice.preview_url) {
|
||||
popupText += `<audio id="${voice.voice_id}" src="${voice.preview_url}" data-disabled="${voice.preview_url == false}"></audio>`
|
||||
popupText += `<audio id="${voice.voice_id}" src="${voice.preview_url}" data-disabled="${voice.preview_url == false}"></audio>`;
|
||||
}
|
||||
}
|
||||
} catch {
|
||||
popupText = 'Could not load voices list. Check your API key.'
|
||||
popupText = 'Could not load voices list. Check your API key.';
|
||||
}
|
||||
|
||||
callPopup(popupText, 'text')
|
||||
callPopup(popupText, 'text');
|
||||
}
|
||||
|
||||
function updateUiAudioPlayState() {
|
||||
if (extension_settings.tts.enabled == true) {
|
||||
$('#ttsExtensionMenuItem').show();
|
||||
let img
|
||||
let img;
|
||||
// Give user feedback that TTS is active by setting the stop icon if processing or playing
|
||||
if (!audioElement.paused || isTtsProcessing()) {
|
||||
img = 'fa-solid fa-stop-circle extensionsMenuExtensionButton'
|
||||
img = 'fa-solid fa-stop-circle extensionsMenuExtensionButton';
|
||||
} else {
|
||||
img = 'fa-solid fa-circle-play extensionsMenuExtensionButton'
|
||||
img = 'fa-solid fa-circle-play extensionsMenuExtensionButton';
|
||||
}
|
||||
$('#tts_media_control').attr('class', img);
|
||||
} else {
|
||||
@ -382,16 +382,16 @@ function updateUiAudioPlayState() {
|
||||
|
||||
function onAudioControlClicked() {
|
||||
audioElement.src = '/sounds/silence.mp3';
|
||||
let context = getContext()
|
||||
let context = getContext();
|
||||
// Not pausing, doing a full stop to anything TTS is doing. Better UX as pause is not as useful
|
||||
if (!audioElement.paused || isTtsProcessing()) {
|
||||
resetTtsPlayback()
|
||||
resetTtsPlayback();
|
||||
talkingAnimation(false);
|
||||
} else {
|
||||
// Default play behavior if not processing or playing is to play the last message.
|
||||
ttsJobQueue.push(context.chat[context.chat.length - 1])
|
||||
ttsJobQueue.push(context.chat[context.chat.length - 1]);
|
||||
}
|
||||
updateUiAudioPlayState()
|
||||
updateUiAudioPlayState();
|
||||
}
|
||||
|
||||
function addAudioControl() {
|
||||
@ -400,15 +400,15 @@ function addAudioControl() {
|
||||
<div id="ttsExtensionMenuItem" class="list-group-item flex-container flexGap5">
|
||||
<div id="tts_media_control" class="extensionsMenuExtensionButton "/></div>
|
||||
TTS Playback
|
||||
</div>`)
|
||||
$('#ttsExtensionMenuItem').attr('title', 'TTS play/pause').on('click', onAudioControlClicked)
|
||||
updateUiAudioPlayState()
|
||||
</div>`);
|
||||
$('#ttsExtensionMenuItem').attr('title', 'TTS play/pause').on('click', onAudioControlClicked);
|
||||
updateUiAudioPlayState();
|
||||
}
|
||||
|
||||
function completeCurrentAudioJob() {
|
||||
audioQueueProcessorReady = true
|
||||
currentAudioJob = null
|
||||
talkingAnimation(false) //stop lip animation
|
||||
audioQueueProcessorReady = true;
|
||||
currentAudioJob = null;
|
||||
talkingAnimation(false); //stop lip animation
|
||||
// updateUiPlayState();
|
||||
}
|
||||
|
||||
@ -417,27 +417,27 @@ function completeCurrentAudioJob() {
|
||||
* @param {Response} response
|
||||
*/
|
||||
async function addAudioJob(response) {
|
||||
const audioData = await response.blob()
|
||||
const audioData = await response.blob();
|
||||
if (!audioData.type.startsWith('audio/')) {
|
||||
throw `TTS received HTTP response with invalid data format. Expecting audio/*, got ${audioData.type}`
|
||||
throw `TTS received HTTP response with invalid data format. Expecting audio/*, got ${audioData.type}`;
|
||||
}
|
||||
audioJobQueue.push(audioData)
|
||||
console.debug('Pushed audio job to queue.')
|
||||
audioJobQueue.push(audioData);
|
||||
console.debug('Pushed audio job to queue.');
|
||||
}
|
||||
|
||||
async function processAudioJobQueue() {
|
||||
// Nothing to do, audio not completed, or audio paused - stop processing.
|
||||
if (audioJobQueue.length == 0 || !audioQueueProcessorReady || audioPaused) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
try {
|
||||
audioQueueProcessorReady = false
|
||||
currentAudioJob = audioJobQueue.pop()
|
||||
playAudioData(currentAudioJob)
|
||||
talkingAnimation(true)
|
||||
audioQueueProcessorReady = false;
|
||||
currentAudioJob = audioJobQueue.pop();
|
||||
playAudioData(currentAudioJob);
|
||||
talkingAnimation(true);
|
||||
} catch (error) {
|
||||
console.error(error)
|
||||
audioQueueProcessorReady = true
|
||||
console.error(error);
|
||||
audioQueueProcessorReady = true;
|
||||
}
|
||||
}
|
||||
|
||||
@ -445,46 +445,46 @@ async function processAudioJobQueue() {
|
||||
// TTS Control //
|
||||
//################//
|
||||
|
||||
let ttsJobQueue = []
|
||||
let currentTtsJob // Null if nothing is currently being processed
|
||||
let currentMessageNumber = 0
|
||||
let ttsJobQueue = [];
|
||||
let currentTtsJob; // Null if nothing is currently being processed
|
||||
let currentMessageNumber = 0;
|
||||
|
||||
function completeTtsJob() {
|
||||
console.info(`Current TTS job for ${currentTtsJob?.name} completed.`)
|
||||
currentTtsJob = null
|
||||
console.info(`Current TTS job for ${currentTtsJob?.name} completed.`);
|
||||
currentTtsJob = null;
|
||||
}
|
||||
|
||||
function saveLastValues() {
|
||||
const context = getContext()
|
||||
lastChatId = context.chatId
|
||||
const context = getContext();
|
||||
lastChatId = context.chatId;
|
||||
lastMessageHash = getStringHash(
|
||||
(context.chat.length && context.chat[context.chat.length - 1].mes) ?? ''
|
||||
)
|
||||
);
|
||||
}
|
||||
|
||||
async function tts(text, voiceId, char) {
|
||||
let response = await ttsProvider.generateTts(text, voiceId)
|
||||
let response = await ttsProvider.generateTts(text, voiceId);
|
||||
|
||||
// RVC injection
|
||||
if (extension_settings.rvc.enabled && typeof window['rvcVoiceConversion'] === 'function')
|
||||
response = await window['rvcVoiceConversion'](response, char, text)
|
||||
response = await window['rvcVoiceConversion'](response, char, text);
|
||||
|
||||
addAudioJob(response)
|
||||
completeTtsJob()
|
||||
addAudioJob(response);
|
||||
completeTtsJob();
|
||||
}
|
||||
|
||||
async function processTtsQueue() {
|
||||
// Called each moduleWorker iteration to pull chat messages from queue
|
||||
if (currentTtsJob || ttsJobQueue.length <= 0 || audioPaused) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
console.debug('New message found, running TTS')
|
||||
currentTtsJob = ttsJobQueue.shift()
|
||||
let text = extension_settings.tts.narrate_translated_only ? (currentTtsJob?.extra?.display_text || currentTtsJob.mes) : currentTtsJob.mes
|
||||
console.debug('New message found, running TTS');
|
||||
currentTtsJob = ttsJobQueue.shift();
|
||||
let text = extension_settings.tts.narrate_translated_only ? (currentTtsJob?.extra?.display_text || currentTtsJob.mes) : currentTtsJob.mes;
|
||||
text = extension_settings.tts.narrate_dialogues_only
|
||||
? text.replace(/\*[^*]*?(\*|$)/g, '').trim() // remove asterisks content
|
||||
: text.replaceAll('*', '').trim() // remove just the asterisks
|
||||
: text.replaceAll('*', '').trim(); // remove just the asterisks
|
||||
|
||||
if (extension_settings.tts.narrate_quoted_only) {
|
||||
const special_quotes = /[“”]/g; // Extend this regex to include other special quotes
|
||||
@ -501,8 +501,8 @@ async function processTtsQueue() {
|
||||
// Collapse newlines and spaces into single space
|
||||
text = text.replace(/\s+/g, ' ').trim();
|
||||
|
||||
console.log(`TTS: ${text}`)
|
||||
const char = currentTtsJob.name
|
||||
console.log(`TTS: ${text}`);
|
||||
const char = currentTtsJob.name;
|
||||
|
||||
// Remove character name from start of the line if power user setting is disabled
|
||||
if (char && !power_user.allow_name2_display) {
|
||||
@ -513,35 +513,35 @@ async function processTtsQueue() {
|
||||
try {
|
||||
if (!text) {
|
||||
console.warn('Got empty text in TTS queue job.');
|
||||
completeTtsJob()
|
||||
completeTtsJob();
|
||||
return;
|
||||
}
|
||||
|
||||
const voiceMapEntry = voiceMap[char] === DEFAULT_VOICE_MARKER ? voiceMap[DEFAULT_VOICE_MARKER] : voiceMap[char]
|
||||
const voiceMapEntry = voiceMap[char] === DEFAULT_VOICE_MARKER ? voiceMap[DEFAULT_VOICE_MARKER] : voiceMap[char];
|
||||
|
||||
if (!voiceMapEntry || voiceMapEntry === DISABLED_VOICE_MARKER) {
|
||||
throw `${char} not in voicemap. Configure character in extension settings voice map`
|
||||
throw `${char} not in voicemap. Configure character in extension settings voice map`;
|
||||
}
|
||||
const voice = await ttsProvider.getVoice(voiceMapEntry)
|
||||
const voiceId = voice.voice_id
|
||||
const voice = await ttsProvider.getVoice(voiceMapEntry);
|
||||
const voiceId = voice.voice_id;
|
||||
if (voiceId == null) {
|
||||
toastr.error(`Specified voice for ${char} was not found. Check the TTS extension settings.`)
|
||||
throw `Unable to attain voiceId for ${char}`
|
||||
toastr.error(`Specified voice for ${char} was not found. Check the TTS extension settings.`);
|
||||
throw `Unable to attain voiceId for ${char}`;
|
||||
}
|
||||
tts(text, voiceId, char)
|
||||
tts(text, voiceId, char);
|
||||
} catch (error) {
|
||||
console.error(error)
|
||||
currentTtsJob = null
|
||||
console.error(error);
|
||||
currentTtsJob = null;
|
||||
}
|
||||
}
|
||||
|
||||
// Secret function for now
|
||||
async function playFullConversation() {
|
||||
const context = getContext()
|
||||
const chat = context.chat
|
||||
ttsJobQueue = chat
|
||||
const context = getContext();
|
||||
const chat = context.chat;
|
||||
ttsJobQueue = chat;
|
||||
}
|
||||
window.playFullConversation = playFullConversation
|
||||
window.playFullConversation = playFullConversation;
|
||||
|
||||
//#############################//
|
||||
// Extension UI and Settings //
|
||||
@ -549,21 +549,21 @@ window.playFullConversation = playFullConversation
|
||||
|
||||
function loadSettings() {
|
||||
if (Object.keys(extension_settings.tts).length === 0) {
|
||||
Object.assign(extension_settings.tts, defaultSettings)
|
||||
Object.assign(extension_settings.tts, defaultSettings);
|
||||
}
|
||||
for (const key in defaultSettings) {
|
||||
if (!(key in extension_settings.tts)) {
|
||||
extension_settings.tts[key] = defaultSettings[key]
|
||||
extension_settings.tts[key] = defaultSettings[key];
|
||||
}
|
||||
}
|
||||
$('#tts_provider').val(extension_settings.tts.currentProvider)
|
||||
$('#tts_provider').val(extension_settings.tts.currentProvider);
|
||||
$('#tts_enabled').prop(
|
||||
'checked',
|
||||
extension_settings.tts.enabled
|
||||
)
|
||||
$('#tts_narrate_dialogues').prop('checked', extension_settings.tts.narrate_dialogues_only)
|
||||
$('#tts_narrate_quoted').prop('checked', extension_settings.tts.narrate_quoted_only)
|
||||
$('#tts_auto_generation').prop('checked', extension_settings.tts.auto_generation)
|
||||
);
|
||||
$('#tts_narrate_dialogues').prop('checked', extension_settings.tts.narrate_dialogues_only);
|
||||
$('#tts_narrate_quoted').prop('checked', extension_settings.tts.narrate_quoted_only);
|
||||
$('#tts_auto_generation').prop('checked', extension_settings.tts.auto_generation);
|
||||
$('#tts_narrate_translated_only').prop('checked', extension_settings.tts.narrate_translated_only);
|
||||
$('#tts_narrate_user').prop('checked', extension_settings.tts.narrate_user);
|
||||
$('body').toggleClass('tts', extension_settings.tts.enabled);
|
||||
@ -575,14 +575,14 @@ const defaultSettings = {
|
||||
currentProvider: 'ElevenLabs',
|
||||
auto_generation: true,
|
||||
narrate_user: false,
|
||||
}
|
||||
};
|
||||
|
||||
function setTtsStatus(status, success) {
|
||||
$('#tts_status').text(status)
|
||||
$('#tts_status').text(status);
|
||||
if (success) {
|
||||
$('#tts_status').removeAttr('style')
|
||||
$('#tts_status').removeAttr('style');
|
||||
} else {
|
||||
$('#tts_status').css('color', 'red')
|
||||
$('#tts_status').css('color', 'red');
|
||||
}
|
||||
}
|
||||
|
||||
@ -591,36 +591,36 @@ function onRefreshClick() {
|
||||
ttsProvider.onRefreshClick(),
|
||||
// updateVoiceMap()
|
||||
]).then(() => {
|
||||
extension_settings.tts[ttsProviderName] = ttsProvider.settings
|
||||
saveSettingsDebounced()
|
||||
setTtsStatus('Successfully applied settings', true)
|
||||
console.info(`Saved settings ${ttsProviderName} ${JSON.stringify(ttsProvider.settings)}`)
|
||||
initVoiceMap()
|
||||
updateVoiceMap()
|
||||
extension_settings.tts[ttsProviderName] = ttsProvider.settings;
|
||||
saveSettingsDebounced();
|
||||
setTtsStatus('Successfully applied settings', true);
|
||||
console.info(`Saved settings ${ttsProviderName} ${JSON.stringify(ttsProvider.settings)}`);
|
||||
initVoiceMap();
|
||||
updateVoiceMap();
|
||||
}).catch(error => {
|
||||
console.error(error)
|
||||
setTtsStatus(error, false)
|
||||
})
|
||||
console.error(error);
|
||||
setTtsStatus(error, false);
|
||||
});
|
||||
}
|
||||
|
||||
function onEnableClick() {
|
||||
extension_settings.tts.enabled = $('#tts_enabled').is(
|
||||
':checked'
|
||||
)
|
||||
updateUiAudioPlayState()
|
||||
saveSettingsDebounced()
|
||||
);
|
||||
updateUiAudioPlayState();
|
||||
saveSettingsDebounced();
|
||||
}
|
||||
|
||||
|
||||
function onAutoGenerationClick() {
|
||||
extension_settings.tts.auto_generation = !!$('#tts_auto_generation').prop('checked');
|
||||
saveSettingsDebounced()
|
||||
saveSettingsDebounced();
|
||||
}
|
||||
|
||||
|
||||
function onNarrateDialoguesClick() {
|
||||
extension_settings.tts.narrate_dialogues_only = !!$('#tts_narrate_dialogues').prop('checked');
|
||||
saveSettingsDebounced()
|
||||
saveSettingsDebounced();
|
||||
}
|
||||
|
||||
function onNarrateUserClick() {
|
||||
@ -630,7 +630,7 @@ function onNarrateUserClick() {
|
||||
|
||||
function onNarrateQuotedClick() {
|
||||
extension_settings.tts.narrate_quoted_only = !!$('#tts_narrate_quoted').prop('checked');
|
||||
saveSettingsDebounced()
|
||||
saveSettingsDebounced();
|
||||
}
|
||||
|
||||
|
||||
@ -645,39 +645,39 @@ function onNarrateTranslatedOnlyClick() {
|
||||
|
||||
async function loadTtsProvider(provider) {
|
||||
//Clear the current config and add new config
|
||||
$('#tts_provider_settings').html('')
|
||||
$('#tts_provider_settings').html('');
|
||||
|
||||
if (!provider) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
// Init provider references
|
||||
extension_settings.tts.currentProvider = provider
|
||||
ttsProviderName = provider
|
||||
ttsProvider = new ttsProviders[provider]
|
||||
extension_settings.tts.currentProvider = provider;
|
||||
ttsProviderName = provider;
|
||||
ttsProvider = new ttsProviders[provider];
|
||||
|
||||
// Init provider settings
|
||||
$('#tts_provider_settings').append(ttsProvider.settingsHtml)
|
||||
$('#tts_provider_settings').append(ttsProvider.settingsHtml);
|
||||
if (!(ttsProviderName in extension_settings.tts)) {
|
||||
console.warn(`Provider ${ttsProviderName} not in Extension Settings, initiatilizing provider in settings`)
|
||||
extension_settings.tts[ttsProviderName] = {}
|
||||
console.warn(`Provider ${ttsProviderName} not in Extension Settings, initiatilizing provider in settings`);
|
||||
extension_settings.tts[ttsProviderName] = {};
|
||||
}
|
||||
await ttsProvider.loadSettings(extension_settings.tts[ttsProviderName])
|
||||
await initVoiceMap()
|
||||
await ttsProvider.loadSettings(extension_settings.tts[ttsProviderName]);
|
||||
await initVoiceMap();
|
||||
}
|
||||
|
||||
function onTtsProviderChange() {
|
||||
const ttsProviderSelection = $('#tts_provider').val()
|
||||
extension_settings.tts.currentProvider = ttsProviderSelection
|
||||
loadTtsProvider(ttsProviderSelection)
|
||||
const ttsProviderSelection = $('#tts_provider').val();
|
||||
extension_settings.tts.currentProvider = ttsProviderSelection;
|
||||
loadTtsProvider(ttsProviderSelection);
|
||||
}
|
||||
|
||||
// Ensure that TTS provider settings are saved to extension settings.
|
||||
export function saveTtsProviderSettings() {
|
||||
extension_settings.tts[ttsProviderName] = ttsProvider.settings
|
||||
updateVoiceMap()
|
||||
saveSettingsDebounced()
|
||||
console.info(`Saved settings ${ttsProviderName} ${JSON.stringify(ttsProvider.settings)}`)
|
||||
extension_settings.tts[ttsProviderName] = ttsProvider.settings;
|
||||
updateVoiceMap();
|
||||
saveSettingsDebounced();
|
||||
console.info(`Saved settings ${ttsProviderName} ${JSON.stringify(ttsProvider.settings)}`);
|
||||
}
|
||||
|
||||
|
||||
@ -686,28 +686,28 @@ export function saveTtsProviderSettings() {
|
||||
//###################//
|
||||
|
||||
async function onChatChanged() {
|
||||
await resetTtsPlayback()
|
||||
await initVoiceMap()
|
||||
ttsLastMessage = null
|
||||
await resetTtsPlayback();
|
||||
await initVoiceMap();
|
||||
ttsLastMessage = null;
|
||||
}
|
||||
|
||||
async function onChatDeleted() {
|
||||
const context = getContext()
|
||||
const context = getContext();
|
||||
|
||||
// update internal references to new last message
|
||||
lastChatId = context.chatId
|
||||
currentMessageNumber = context.chat.length ? context.chat.length : 0
|
||||
lastChatId = context.chatId;
|
||||
currentMessageNumber = context.chat.length ? context.chat.length : 0;
|
||||
|
||||
// compare against lastMessageHash. If it's the same, we did not delete the last chat item, so no need to reset tts queue
|
||||
let messageHash = getStringHash((context.chat.length && context.chat[context.chat.length - 1].mes) ?? '')
|
||||
let messageHash = getStringHash((context.chat.length && context.chat[context.chat.length - 1].mes) ?? '');
|
||||
if (messageHash === lastMessageHash) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
lastMessageHash = messageHash
|
||||
lastMessageHash = messageHash;
|
||||
ttsLastMessage = (context.chat.length && context.chat[context.chat.length - 1].mes) ?? '';
|
||||
|
||||
// stop any tts playback since message might not exist anymore
|
||||
await resetTtsPlayback()
|
||||
await resetTtsPlayback();
|
||||
}
|
||||
|
||||
/**
|
||||
@ -716,7 +716,7 @@ async function onChatDeleted() {
|
||||
* @returns {string[]} - Array of character names
|
||||
*/
|
||||
function getCharacters(unrestricted) {
|
||||
const context = getContext()
|
||||
const context = getContext();
|
||||
|
||||
if (unrestricted) {
|
||||
const names = context.characters.map(char => char.name);
|
||||
@ -724,26 +724,26 @@ function getCharacters(unrestricted) {
|
||||
return names;
|
||||
}
|
||||
|
||||
let characters = []
|
||||
let characters = [];
|
||||
if (context.groupId === null) {
|
||||
// Single char chat
|
||||
characters.push(DEFAULT_VOICE_MARKER)
|
||||
characters.push(context.name1)
|
||||
characters.push(context.name2)
|
||||
characters.push(DEFAULT_VOICE_MARKER);
|
||||
characters.push(context.name1);
|
||||
characters.push(context.name2);
|
||||
} else {
|
||||
// Group chat
|
||||
characters.push(DEFAULT_VOICE_MARKER)
|
||||
characters.push(context.name1)
|
||||
const group = context.groups.find(group => context.groupId == group.id)
|
||||
characters.push(DEFAULT_VOICE_MARKER);
|
||||
characters.push(context.name1);
|
||||
const group = context.groups.find(group => context.groupId == group.id);
|
||||
for (let member of group.members) {
|
||||
// Remove suffix
|
||||
if (member.endsWith('.png')) {
|
||||
member = member.slice(0, -4)
|
||||
member = member.slice(0, -4);
|
||||
}
|
||||
characters.push(member)
|
||||
characters.push(member);
|
||||
}
|
||||
}
|
||||
return characters
|
||||
return characters;
|
||||
}
|
||||
|
||||
function sanitizeId(input) {
|
||||
@ -759,15 +759,15 @@ function sanitizeId(input) {
|
||||
}
|
||||
|
||||
function parseVoiceMap(voiceMapString) {
|
||||
let parsedVoiceMap = {}
|
||||
let parsedVoiceMap = {};
|
||||
for (const [charName, voiceId] of voiceMapString
|
||||
.split(',')
|
||||
.map(s => s.split(':'))) {
|
||||
if (charName && voiceId) {
|
||||
parsedVoiceMap[charName.trim()] = voiceId.trim()
|
||||
parsedVoiceMap[charName.trim()] = voiceId.trim();
|
||||
}
|
||||
}
|
||||
return parsedVoiceMap
|
||||
return parsedVoiceMap;
|
||||
}
|
||||
|
||||
|
||||
@ -776,39 +776,39 @@ function parseVoiceMap(voiceMapString) {
|
||||
* Apply voiceMap based on current voiceMapEntries
|
||||
*/
|
||||
function updateVoiceMap() {
|
||||
const tempVoiceMap = {}
|
||||
const tempVoiceMap = {};
|
||||
for (const voice of voiceMapEntries) {
|
||||
if (voice.voiceId === null) {
|
||||
continue
|
||||
continue;
|
||||
}
|
||||
tempVoiceMap[voice.name] = voice.voiceId
|
||||
tempVoiceMap[voice.name] = voice.voiceId;
|
||||
}
|
||||
if (Object.keys(tempVoiceMap).length !== 0) {
|
||||
voiceMap = tempVoiceMap
|
||||
console.log(`Voicemap updated to ${JSON.stringify(voiceMap)}`)
|
||||
voiceMap = tempVoiceMap;
|
||||
console.log(`Voicemap updated to ${JSON.stringify(voiceMap)}`);
|
||||
}
|
||||
if (!extension_settings.tts[ttsProviderName].voiceMap) {
|
||||
extension_settings.tts[ttsProviderName].voiceMap = {}
|
||||
extension_settings.tts[ttsProviderName].voiceMap = {};
|
||||
}
|
||||
Object.assign(extension_settings.tts[ttsProviderName].voiceMap, voiceMap)
|
||||
saveSettingsDebounced()
|
||||
Object.assign(extension_settings.tts[ttsProviderName].voiceMap, voiceMap);
|
||||
saveSettingsDebounced();
|
||||
}
|
||||
|
||||
class VoiceMapEntry {
|
||||
name
|
||||
voiceId
|
||||
selectElement
|
||||
name;
|
||||
voiceId;
|
||||
selectElement;
|
||||
constructor(name, voiceId = DEFAULT_VOICE_MARKER) {
|
||||
this.name = name
|
||||
this.voiceId = voiceId
|
||||
this.selectElement = null
|
||||
this.name = name;
|
||||
this.voiceId = voiceId;
|
||||
this.selectElement = null;
|
||||
}
|
||||
|
||||
addUI(voiceIds) {
|
||||
let sanitizedName = sanitizeId(this.name)
|
||||
let sanitizedName = sanitizeId(this.name);
|
||||
let defaultOption = this.name === DEFAULT_VOICE_MARKER ?
|
||||
`<option>${DISABLED_VOICE_MARKER}</option>` :
|
||||
`<option>${DEFAULT_VOICE_MARKER}</option><option>${DISABLED_VOICE_MARKER}</option>`
|
||||
`<option>${DEFAULT_VOICE_MARKER}</option><option>${DISABLED_VOICE_MARKER}</option>`;
|
||||
let template = `
|
||||
<div class='tts_voicemap_block_char flex-container flexGap5'>
|
||||
<span id='tts_voicemap_char_${sanitizedName}'>${this.name}</span>
|
||||
@ -816,25 +816,25 @@ class VoiceMapEntry {
|
||||
${defaultOption}
|
||||
</select>
|
||||
</div>
|
||||
`
|
||||
$('#tts_voicemap_block').append(template)
|
||||
`;
|
||||
$('#tts_voicemap_block').append(template);
|
||||
|
||||
// Populate voice ID select list
|
||||
for (const voiceId of voiceIds) {
|
||||
const option = document.createElement('option');
|
||||
option.innerText = voiceId.name;
|
||||
option.value = voiceId.name;
|
||||
$(`#tts_voicemap_char_${sanitizedName}_voice`).append(option)
|
||||
$(`#tts_voicemap_char_${sanitizedName}_voice`).append(option);
|
||||
}
|
||||
|
||||
this.selectElement = $(`#tts_voicemap_char_${sanitizedName}_voice`)
|
||||
this.selectElement.on('change', args => this.onSelectChange(args))
|
||||
this.selectElement.val(this.voiceId)
|
||||
this.selectElement = $(`#tts_voicemap_char_${sanitizedName}_voice`);
|
||||
this.selectElement.on('change', args => this.onSelectChange(args));
|
||||
this.selectElement.val(this.voiceId);
|
||||
}
|
||||
|
||||
onSelectChange(args) {
|
||||
this.voiceId = this.selectElement.find(':selected').val()
|
||||
updateVoiceMap()
|
||||
this.voiceId = this.selectElement.find(':selected').val();
|
||||
updateVoiceMap();
|
||||
}
|
||||
}
|
||||
|
||||
@ -844,69 +844,69 @@ class VoiceMapEntry {
|
||||
*/
|
||||
export async function initVoiceMap(unrestricted = false) {
|
||||
// Gate initialization if not enabled or TTS Provider not ready. Prevents error popups.
|
||||
const enabled = $('#tts_enabled').is(':checked')
|
||||
const enabled = $('#tts_enabled').is(':checked');
|
||||
if (!enabled) {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
// Keep errors inside extension UI rather than toastr. Toastr errors for TTS are annoying.
|
||||
try {
|
||||
await ttsProvider.checkReady()
|
||||
await ttsProvider.checkReady();
|
||||
} catch (error) {
|
||||
const message = `TTS Provider not ready. ${error}`
|
||||
setTtsStatus(message, false)
|
||||
return
|
||||
const message = `TTS Provider not ready. ${error}`;
|
||||
setTtsStatus(message, false);
|
||||
return;
|
||||
}
|
||||
|
||||
setTtsStatus('TTS Provider Loaded', true)
|
||||
setTtsStatus('TTS Provider Loaded', true);
|
||||
|
||||
// Clear existing voiceMap state
|
||||
$('#tts_voicemap_block').empty()
|
||||
voiceMapEntries = []
|
||||
$('#tts_voicemap_block').empty();
|
||||
voiceMapEntries = [];
|
||||
|
||||
// Get characters in current chat
|
||||
const characters = getCharacters(unrestricted);
|
||||
|
||||
// Get saved voicemap from provider settings, handling new and old representations
|
||||
let voiceMapFromSettings = {}
|
||||
let voiceMapFromSettings = {};
|
||||
if ('voiceMap' in extension_settings.tts[ttsProviderName]) {
|
||||
// Handle previous representation
|
||||
if (typeof extension_settings.tts[ttsProviderName].voiceMap === 'string') {
|
||||
voiceMapFromSettings = parseVoiceMap(extension_settings.tts[ttsProviderName].voiceMap)
|
||||
voiceMapFromSettings = parseVoiceMap(extension_settings.tts[ttsProviderName].voiceMap);
|
||||
// Handle new representation
|
||||
} else if (typeof extension_settings.tts[ttsProviderName].voiceMap === 'object') {
|
||||
voiceMapFromSettings = extension_settings.tts[ttsProviderName].voiceMap
|
||||
voiceMapFromSettings = extension_settings.tts[ttsProviderName].voiceMap;
|
||||
}
|
||||
}
|
||||
|
||||
// Get voiceIds from provider
|
||||
let voiceIdsFromProvider
|
||||
let voiceIdsFromProvider;
|
||||
try {
|
||||
voiceIdsFromProvider = await ttsProvider.fetchTtsVoiceObjects()
|
||||
voiceIdsFromProvider = await ttsProvider.fetchTtsVoiceObjects();
|
||||
}
|
||||
catch {
|
||||
toastr.error('TTS Provider failed to return voice ids.')
|
||||
toastr.error('TTS Provider failed to return voice ids.');
|
||||
}
|
||||
|
||||
// Build UI using VoiceMapEntry objects
|
||||
for (const character of characters) {
|
||||
if (character === 'SillyTavern System') {
|
||||
continue
|
||||
continue;
|
||||
}
|
||||
// Check provider settings for voiceIds
|
||||
let voiceId
|
||||
let voiceId;
|
||||
if (character in voiceMapFromSettings) {
|
||||
voiceId = voiceMapFromSettings[character]
|
||||
voiceId = voiceMapFromSettings[character];
|
||||
} else if (character === DEFAULT_VOICE_MARKER) {
|
||||
voiceId = DISABLED_VOICE_MARKER
|
||||
voiceId = DISABLED_VOICE_MARKER;
|
||||
} else {
|
||||
voiceId = DEFAULT_VOICE_MARKER
|
||||
voiceId = DEFAULT_VOICE_MARKER;
|
||||
}
|
||||
const voiceMapEntry = new VoiceMapEntry(character, voiceId)
|
||||
voiceMapEntry.addUI(voiceIdsFromProvider)
|
||||
voiceMapEntries.push(voiceMapEntry)
|
||||
const voiceMapEntry = new VoiceMapEntry(character, voiceId);
|
||||
voiceMapEntry.addUI(voiceIdsFromProvider);
|
||||
voiceMapEntries.push(voiceMapEntry);
|
||||
}
|
||||
updateVoiceMap()
|
||||
updateVoiceMap();
|
||||
}
|
||||
|
||||
$(document).ready(function () {
|
||||
@ -965,32 +965,32 @@ $(document).ready(function () {
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
`
|
||||
$('#extensions_settings').append(settingsHtml)
|
||||
$('#tts_refresh').on('click', onRefreshClick)
|
||||
$('#tts_enabled').on('click', onEnableClick)
|
||||
`;
|
||||
$('#extensions_settings').append(settingsHtml);
|
||||
$('#tts_refresh').on('click', onRefreshClick);
|
||||
$('#tts_enabled').on('click', onEnableClick);
|
||||
$('#tts_narrate_dialogues').on('click', onNarrateDialoguesClick);
|
||||
$('#tts_narrate_quoted').on('click', onNarrateQuotedClick);
|
||||
$('#tts_narrate_translated_only').on('click', onNarrateTranslatedOnlyClick);
|
||||
$('#tts_auto_generation').on('click', onAutoGenerationClick);
|
||||
$('#tts_narrate_user').on('click', onNarrateUserClick);
|
||||
$('#tts_voices').on('click', onTtsVoicesClick)
|
||||
$('#tts_voices').on('click', onTtsVoicesClick);
|
||||
for (const provider in ttsProviders) {
|
||||
$('#tts_provider').append($('<option />').val(provider).text(provider))
|
||||
$('#tts_provider').append($('<option />').val(provider).text(provider));
|
||||
}
|
||||
$('#tts_provider').on('change', onTtsProviderChange)
|
||||
$('#tts_provider').on('change', onTtsProviderChange);
|
||||
$(document).on('click', '.mes_narrate', onNarrateOneMessage);
|
||||
}
|
||||
addExtensionControls() // No init dependencies
|
||||
loadSettings() // Depends on Extension Controls and loadTtsProvider
|
||||
loadTtsProvider(extension_settings.tts.currentProvider) // No dependencies
|
||||
addAudioControl() // Depends on Extension Controls
|
||||
addExtensionControls(); // No init dependencies
|
||||
loadSettings(); // Depends on Extension Controls and loadTtsProvider
|
||||
loadTtsProvider(extension_settings.tts.currentProvider); // No dependencies
|
||||
addAudioControl(); // Depends on Extension Controls
|
||||
const wrapper = new ModuleWorkerWrapper(moduleWorker);
|
||||
setInterval(wrapper.update.bind(wrapper), UPDATE_INTERVAL) // Init depends on all the things
|
||||
setInterval(wrapper.update.bind(wrapper), UPDATE_INTERVAL); // Init depends on all the things
|
||||
eventSource.on(event_types.MESSAGE_SWIPED, resetTtsPlayback);
|
||||
eventSource.on(event_types.CHAT_CHANGED, onChatChanged)
|
||||
eventSource.on(event_types.CHAT_CHANGED, onChatChanged);
|
||||
eventSource.on(event_types.MESSAGE_DELETED, onChatDeleted);
|
||||
eventSource.on(event_types.GROUP_UPDATED, onChatChanged)
|
||||
eventSource.on(event_types.GROUP_UPDATED, onChatChanged);
|
||||
registerSlashCommand('speak', onNarrateText, ['narrate', 'tts'], '<span class="monospace">(text)</span> – narrate any text using currently selected character\'s voice. Use voice="Character Name" argument to set other voice from the voice map, example: <tt>/speak voice="Donald Duck" Quack!</tt>', true, true);
|
||||
document.body.appendChild(audioElement);
|
||||
})
|
||||
});
|
||||
|
||||
@ -1,23 +1,23 @@
|
||||
import { getRequestHeaders, callPopup } from '../../../script.js'
|
||||
import { getPreviewString, saveTtsProviderSettings } from './index.js'
|
||||
import { initVoiceMap } from './index.js'
|
||||
import { getRequestHeaders, callPopup } from '../../../script.js';
|
||||
import { getPreviewString, saveTtsProviderSettings } from './index.js';
|
||||
import { initVoiceMap } from './index.js';
|
||||
|
||||
export { NovelTtsProvider }
|
||||
export { NovelTtsProvider };
|
||||
|
||||
class NovelTtsProvider {
|
||||
//########//
|
||||
// Config //
|
||||
//########//
|
||||
|
||||
settings
|
||||
voices = []
|
||||
separator = ' . '
|
||||
audioElement = document.createElement('audio')
|
||||
settings;
|
||||
voices = [];
|
||||
separator = ' . ';
|
||||
audioElement = document.createElement('audio');
|
||||
|
||||
defaultSettings = {
|
||||
voiceMap: {},
|
||||
customVoices: []
|
||||
}
|
||||
};
|
||||
|
||||
/**
|
||||
* Perform any text processing before passing to TTS engine.
|
||||
@ -53,11 +53,11 @@ class NovelTtsProvider {
|
||||
|
||||
// Add a new Novel custom voice to provider
|
||||
async addCustomVoice(){
|
||||
const voiceName = await callPopup('<h3>Custom Voice name:</h3>', 'input')
|
||||
this.settings.customVoices.push(voiceName)
|
||||
this.populateCustomVoices()
|
||||
initVoiceMap() // Update TTS extension voiceMap
|
||||
saveTtsProviderSettings()
|
||||
const voiceName = await callPopup('<h3>Custom Voice name:</h3>', 'input');
|
||||
this.settings.customVoices.push(voiceName);
|
||||
this.populateCustomVoices();
|
||||
initVoiceMap(); // Update TTS extension voiceMap
|
||||
saveTtsProviderSettings();
|
||||
}
|
||||
|
||||
// Delete selected custom voice from provider
|
||||
@ -68,52 +68,52 @@ class NovelTtsProvider {
|
||||
if (voiceIndex !== -1) {
|
||||
this.settings.customVoices.splice(voiceIndex, 1);
|
||||
}
|
||||
this.populateCustomVoices()
|
||||
initVoiceMap() // Update TTS extension voiceMap
|
||||
saveTtsProviderSettings()
|
||||
this.populateCustomVoices();
|
||||
initVoiceMap(); // Update TTS extension voiceMap
|
||||
saveTtsProviderSettings();
|
||||
}
|
||||
|
||||
// Create the UI dropdown list of voices in provider
|
||||
populateCustomVoices(){
|
||||
let voiceSelect = $('#tts-novel-custom-voices-select')
|
||||
voiceSelect.empty()
|
||||
let voiceSelect = $('#tts-novel-custom-voices-select');
|
||||
voiceSelect.empty();
|
||||
this.settings.customVoices.forEach(voice => {
|
||||
voiceSelect.append(`<option>${voice}</option>`)
|
||||
})
|
||||
voiceSelect.append(`<option>${voice}</option>`);
|
||||
});
|
||||
}
|
||||
|
||||
async loadSettings(settings) {
|
||||
// Populate Provider UI given input settings
|
||||
if (Object.keys(settings).length == 0) {
|
||||
console.info('Using default TTS Provider settings')
|
||||
console.info('Using default TTS Provider settings');
|
||||
}
|
||||
$('#tts-novel-custom-voices-add').on('click', () => (this.addCustomVoice()))
|
||||
$('#tts-novel-custom-voices-delete').on('click',() => (this.deleteCustomVoice()))
|
||||
$('#tts-novel-custom-voices-add').on('click', () => (this.addCustomVoice()));
|
||||
$('#tts-novel-custom-voices-delete').on('click',() => (this.deleteCustomVoice()));
|
||||
|
||||
// Only accept keys defined in defaultSettings
|
||||
this.settings = this.defaultSettings
|
||||
this.settings = this.defaultSettings;
|
||||
|
||||
for (const key in settings) {
|
||||
if (key in this.settings) {
|
||||
this.settings[key] = settings[key]
|
||||
this.settings[key] = settings[key];
|
||||
} else {
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`;
|
||||
}
|
||||
}
|
||||
|
||||
this.populateCustomVoices()
|
||||
await this.checkReady()
|
||||
console.debug('NovelTTS: Settings loaded')
|
||||
this.populateCustomVoices();
|
||||
await this.checkReady();
|
||||
console.debug('NovelTTS: Settings loaded');
|
||||
}
|
||||
|
||||
// Perform a simple readiness check by trying to fetch voiceIds
|
||||
// Doesnt really do much for Novel, not seeing a good way to test this at the moment.
|
||||
async checkReady(){
|
||||
await this.fetchTtsVoiceObjects()
|
||||
await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
|
||||
async onRefreshClick() {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
//#################//
|
||||
@ -122,15 +122,15 @@ class NovelTtsProvider {
|
||||
|
||||
async getVoice(voiceName) {
|
||||
if (!voiceName) {
|
||||
throw 'TTS Voice name not provided'
|
||||
throw 'TTS Voice name not provided';
|
||||
}
|
||||
|
||||
return { name: voiceName, voice_id: voiceName, lang: 'en-US', preview_url: false}
|
||||
return { name: voiceName, voice_id: voiceName, lang: 'en-US', preview_url: false};
|
||||
}
|
||||
|
||||
async generateTts(text, voiceId) {
|
||||
const response = await this.fetchTtsGeneration(text, voiceId)
|
||||
return response
|
||||
const response = await this.fetchTtsGeneration(text, voiceId);
|
||||
return response;
|
||||
}
|
||||
|
||||
//###########//
|
||||
@ -156,8 +156,8 @@ class NovelTtsProvider {
|
||||
// Add in custom voices to the map
|
||||
let addVoices = this.settings.customVoices.map(voice =>
|
||||
({ name: voice, voice_id: voice, lang: 'en-US', preview_url: false })
|
||||
)
|
||||
voices = voices.concat(addVoices)
|
||||
);
|
||||
voices = voices.concat(addVoices);
|
||||
|
||||
return voices;
|
||||
}
|
||||
@ -167,10 +167,10 @@ class NovelTtsProvider {
|
||||
this.audioElement.pause();
|
||||
this.audioElement.currentTime = 0;
|
||||
|
||||
const text = getPreviewString('en-US')
|
||||
const response = await this.fetchTtsGeneration(text, id)
|
||||
const text = getPreviewString('en-US');
|
||||
const response = await this.fetchTtsGeneration(text, id);
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}`)
|
||||
throw new Error(`HTTP ${response.status}`);
|
||||
}
|
||||
|
||||
const audio = await response.blob();
|
||||
@ -180,7 +180,7 @@ class NovelTtsProvider {
|
||||
}
|
||||
|
||||
async fetchTtsGeneration(inputText, voiceId) {
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`)
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`);
|
||||
const response = await fetch('/api/novelai/generate-voice',
|
||||
{
|
||||
method: 'POST',
|
||||
@ -190,11 +190,11 @@ class NovelTtsProvider {
|
||||
'voice': voiceId,
|
||||
})
|
||||
}
|
||||
)
|
||||
);
|
||||
if (!response.ok) {
|
||||
toastr.error(response.statusText, 'TTS Generation Failed');
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
return response
|
||||
return response;
|
||||
}
|
||||
}
|
||||
|
||||
@ -1,7 +1,7 @@
|
||||
import { getRequestHeaders } from '../../../script.js'
|
||||
import { getRequestHeaders } from '../../../script.js';
|
||||
import { saveTtsProviderSettings } from './index.js';
|
||||
|
||||
export { OpenAITtsProvider }
|
||||
export { OpenAITtsProvider };
|
||||
|
||||
class OpenAITtsProvider {
|
||||
static voices = [
|
||||
@ -13,17 +13,17 @@ class OpenAITtsProvider {
|
||||
{ name: 'Shimmer', voice_id: 'shimmer', lang: 'en-US', preview_url: 'https://cdn.openai.com/API/docs/audio/shimmer.wav' },
|
||||
];
|
||||
|
||||
settings
|
||||
voices = []
|
||||
separator = ' . '
|
||||
audioElement = document.createElement('audio')
|
||||
settings;
|
||||
voices = [];
|
||||
separator = ' . ';
|
||||
audioElement = document.createElement('audio');
|
||||
|
||||
defaultSettings = {
|
||||
voiceMap: {},
|
||||
customVoices: [],
|
||||
model: 'tts-1',
|
||||
speed: 1,
|
||||
}
|
||||
};
|
||||
|
||||
get settingsHtml() {
|
||||
let html = `
|
||||
@ -52,7 +52,7 @@ class OpenAITtsProvider {
|
||||
async loadSettings(settings) {
|
||||
// Populate Provider UI given input settings
|
||||
if (Object.keys(settings).length == 0) {
|
||||
console.info('Using default TTS Provider settings')
|
||||
console.info('Using default TTS Provider settings');
|
||||
}
|
||||
|
||||
// Only accept keys defined in defaultSettings
|
||||
@ -100,21 +100,21 @@ class OpenAITtsProvider {
|
||||
|
||||
async getVoice(voiceName) {
|
||||
if (!voiceName) {
|
||||
throw 'TTS Voice name not provided'
|
||||
throw 'TTS Voice name not provided';
|
||||
}
|
||||
|
||||
const voice = OpenAITtsProvider.voices.find(voice => voice.voice_id === voiceName || voice.name === voiceName);
|
||||
|
||||
if (!voice) {
|
||||
throw `TTS Voice not found: ${voiceName}`
|
||||
throw `TTS Voice not found: ${voiceName}`;
|
||||
}
|
||||
|
||||
return voice;
|
||||
}
|
||||
|
||||
async generateTts(text, voiceId) {
|
||||
const response = await this.fetchTtsGeneration(text, voiceId)
|
||||
return response
|
||||
const response = await this.fetchTtsGeneration(text, voiceId);
|
||||
return response;
|
||||
}
|
||||
|
||||
async fetchTtsVoiceObjects() {
|
||||
@ -126,7 +126,7 @@ class OpenAITtsProvider {
|
||||
}
|
||||
|
||||
async fetchTtsGeneration(inputText, voiceId) {
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`)
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`);
|
||||
const response = await fetch('/api/openai/generate-voice', {
|
||||
method: 'POST',
|
||||
headers: getRequestHeaders(),
|
||||
|
||||
@ -1,22 +1,22 @@
|
||||
import { doExtrasFetch, getApiUrl, modules } from '../../extensions.js'
|
||||
import { saveTtsProviderSettings } from './index.js'
|
||||
import { doExtrasFetch, getApiUrl, modules } from '../../extensions.js';
|
||||
import { saveTtsProviderSettings } from './index.js';
|
||||
|
||||
export { SileroTtsProvider }
|
||||
export { SileroTtsProvider };
|
||||
|
||||
class SileroTtsProvider {
|
||||
//########//
|
||||
// Config //
|
||||
//########//
|
||||
|
||||
settings
|
||||
ready = false
|
||||
voices = []
|
||||
separator = ' .. '
|
||||
settings;
|
||||
ready = false;
|
||||
voices = [];
|
||||
separator = ' .. ';
|
||||
|
||||
defaultSettings = {
|
||||
provider_endpoint: 'http://localhost:8001/tts',
|
||||
voiceMap: {}
|
||||
}
|
||||
};
|
||||
|
||||
get settingsHtml() {
|
||||
let html = `
|
||||
@ -24,31 +24,31 @@ class SileroTtsProvider {
|
||||
<input id="silero_tts_endpoint" type="text" class="text_pole" maxlength="250" value="${this.defaultSettings.provider_endpoint}"/>
|
||||
<span>
|
||||
<span>Use <a target="_blank" href="https://github.com/SillyTavern/SillyTavern-extras">SillyTavern Extras API</a> or <a target="_blank" href="https://github.com/ouoertheo/silero-api-server">Silero TTS Server</a>.</span>
|
||||
`
|
||||
return html
|
||||
`;
|
||||
return html;
|
||||
}
|
||||
|
||||
onSettingsChange() {
|
||||
// Used when provider settings are updated from UI
|
||||
this.settings.provider_endpoint = $('#silero_tts_endpoint').val()
|
||||
saveTtsProviderSettings()
|
||||
this.refreshSession()
|
||||
this.settings.provider_endpoint = $('#silero_tts_endpoint').val();
|
||||
saveTtsProviderSettings();
|
||||
this.refreshSession();
|
||||
}
|
||||
|
||||
async loadSettings(settings) {
|
||||
// Pupulate Provider UI given input settings
|
||||
if (Object.keys(settings).length == 0) {
|
||||
console.info('Using default TTS Provider settings')
|
||||
console.info('Using default TTS Provider settings');
|
||||
}
|
||||
|
||||
// Only accept keys defined in defaultSettings
|
||||
this.settings = this.defaultSettings
|
||||
this.settings = this.defaultSettings;
|
||||
|
||||
for (const key in settings) {
|
||||
if (key in this.settings) {
|
||||
this.settings[key] = settings[key]
|
||||
this.settings[key] = settings[key];
|
||||
} else {
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`;
|
||||
}
|
||||
}
|
||||
|
||||
@ -63,26 +63,26 @@ class SileroTtsProvider {
|
||||
}
|
||||
}, 2000);
|
||||
|
||||
$('#silero_tts_endpoint').val(this.settings.provider_endpoint)
|
||||
$('#silero_tts_endpoint').on('input', () => { this.onSettingsChange() })
|
||||
this.refreshSession()
|
||||
$('#silero_tts_endpoint').val(this.settings.provider_endpoint);
|
||||
$('#silero_tts_endpoint').on('input', () => { this.onSettingsChange(); });
|
||||
this.refreshSession();
|
||||
|
||||
await this.checkReady()
|
||||
await this.checkReady();
|
||||
|
||||
console.debug('SileroTTS: Settings loaded')
|
||||
console.debug('SileroTTS: Settings loaded');
|
||||
}
|
||||
|
||||
// Perform a simple readiness check by trying to fetch voiceIds
|
||||
async checkReady() {
|
||||
await this.fetchTtsVoiceObjects()
|
||||
await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
|
||||
async onRefreshClick() {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
async refreshSession() {
|
||||
await this.initSession()
|
||||
await this.initSession();
|
||||
}
|
||||
|
||||
//#################//
|
||||
@ -91,36 +91,36 @@ class SileroTtsProvider {
|
||||
|
||||
async getVoice(voiceName) {
|
||||
if (this.voices.length == 0) {
|
||||
this.voices = await this.fetchTtsVoiceObjects()
|
||||
this.voices = await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
const match = this.voices.filter(
|
||||
sileroVoice => sileroVoice.name == voiceName
|
||||
)[0]
|
||||
)[0];
|
||||
if (!match) {
|
||||
throw `TTS Voice name ${voiceName} not found`
|
||||
throw `TTS Voice name ${voiceName} not found`;
|
||||
}
|
||||
return match
|
||||
return match;
|
||||
}
|
||||
|
||||
async generateTts(text, voiceId) {
|
||||
const response = await this.fetchTtsGeneration(text, voiceId)
|
||||
return response
|
||||
const response = await this.fetchTtsGeneration(text, voiceId);
|
||||
return response;
|
||||
}
|
||||
|
||||
//###########//
|
||||
// API CALLS //
|
||||
//###########//
|
||||
async fetchTtsVoiceObjects() {
|
||||
const response = await doExtrasFetch(`${this.settings.provider_endpoint}/speakers`)
|
||||
const response = await doExtrasFetch(`${this.settings.provider_endpoint}/speakers`);
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.json()}`)
|
||||
throw new Error(`HTTP ${response.status}: ${await response.json()}`);
|
||||
}
|
||||
const responseJson = await response.json()
|
||||
return responseJson
|
||||
const responseJson = await response.json();
|
||||
return responseJson;
|
||||
}
|
||||
|
||||
async fetchTtsGeneration(inputText, voiceId) {
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`)
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`);
|
||||
const response = await doExtrasFetch(
|
||||
`${this.settings.provider_endpoint}/generate`,
|
||||
{
|
||||
@ -135,12 +135,12 @@ class SileroTtsProvider {
|
||||
'session': 'sillytavern'
|
||||
})
|
||||
}
|
||||
)
|
||||
);
|
||||
if (!response.ok) {
|
||||
toastr.error(response.statusText, 'TTS Generation Failed');
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
return response
|
||||
return response;
|
||||
}
|
||||
|
||||
async initSession() {
|
||||
@ -158,7 +158,7 @@ class SileroTtsProvider {
|
||||
'path': 'sillytavern',
|
||||
}),
|
||||
}
|
||||
)
|
||||
);
|
||||
|
||||
if (!response.ok && response.status !== 404) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
|
||||
@ -1,8 +1,8 @@
|
||||
import { isMobile } from '../../RossAscends-mods.js';
|
||||
import { getPreviewString } from './index.js';
|
||||
import { talkingAnimation } from './index.js';
|
||||
import { saveTtsProviderSettings } from './index.js'
|
||||
export { SystemTtsProvider }
|
||||
import { saveTtsProviderSettings } from './index.js';
|
||||
export { SystemTtsProvider };
|
||||
|
||||
/**
|
||||
* Chunkify
|
||||
@ -79,16 +79,16 @@ class SystemTtsProvider {
|
||||
// Config //
|
||||
//########//
|
||||
|
||||
settings
|
||||
ready = false
|
||||
voices = []
|
||||
separator = ' ... '
|
||||
settings;
|
||||
ready = false;
|
||||
voices = [];
|
||||
separator = ' ... ';
|
||||
|
||||
defaultSettings = {
|
||||
voiceMap: {},
|
||||
rate: 1,
|
||||
pitch: 1,
|
||||
}
|
||||
};
|
||||
|
||||
get settingsHtml() {
|
||||
if (!('speechSynthesis' in window)) {
|
||||
@ -107,7 +107,7 @@ class SystemTtsProvider {
|
||||
this.settings.pitch = Number($('#system_tts_pitch').val());
|
||||
$('#system_tts_pitch_output').text(this.settings.pitch);
|
||||
$('#system_tts_rate_output').text(this.settings.rate);
|
||||
saveTtsProviderSettings()
|
||||
saveTtsProviderSettings();
|
||||
}
|
||||
|
||||
async loadSettings(settings) {
|
||||
@ -146,8 +146,8 @@ class SystemTtsProvider {
|
||||
$('#system_tts_pitch').val(this.settings.pitch || this.defaultSettings.pitch);
|
||||
|
||||
// Trigger updates
|
||||
$('#system_tts_rate').on('input', () => { this.onSettingsChange() })
|
||||
$('#system_tts_rate').on('input', () => { this.onSettingsChange() })
|
||||
$('#system_tts_rate').on('input', () => { this.onSettingsChange(); });
|
||||
$('#system_tts_rate').on('input', () => { this.onSettingsChange(); });
|
||||
|
||||
$('#system_tts_pitch_output').text(this.settings.pitch);
|
||||
$('#system_tts_rate_output').text(this.settings.rate);
|
||||
@ -156,11 +156,11 @@ class SystemTtsProvider {
|
||||
|
||||
// Perform a simple readiness check by trying to fetch voiceIds
|
||||
async checkReady() {
|
||||
await this.fetchTtsVoiceObjects()
|
||||
await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
|
||||
async onRefreshClick() {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
//#################//
|
||||
@ -191,7 +191,7 @@ class SystemTtsProvider {
|
||||
const voice = speechSynthesis.getVoices().find(x => x.voiceURI === voiceId);
|
||||
|
||||
if (!voice) {
|
||||
throw `TTS Voice id ${voiceId} not found`
|
||||
throw `TTS Voice id ${voiceId} not found`;
|
||||
}
|
||||
|
||||
speechSynthesis.cancel();
|
||||
@ -205,14 +205,14 @@ class SystemTtsProvider {
|
||||
|
||||
async getVoice(voiceName) {
|
||||
if (!('speechSynthesis' in window)) {
|
||||
return { voice_id: null }
|
||||
return { voice_id: null };
|
||||
}
|
||||
|
||||
const voices = speechSynthesis.getVoices();
|
||||
const match = voices.find(x => x.name == voiceName);
|
||||
|
||||
if (!match) {
|
||||
throw `TTS Voice name ${voiceName} not found`
|
||||
throw `TTS Voice name ${voiceName} not found`;
|
||||
}
|
||||
|
||||
return { voice_id: match.voiceURI, name: match.name };
|
||||
|
||||
@ -1,17 +1,17 @@
|
||||
import { doExtrasFetch, getApiUrl, modules } from '../../extensions.js'
|
||||
import { saveTtsProviderSettings } from './index.js'
|
||||
import { doExtrasFetch, getApiUrl, modules } from '../../extensions.js';
|
||||
import { saveTtsProviderSettings } from './index.js';
|
||||
|
||||
export { XTTSTtsProvider }
|
||||
export { XTTSTtsProvider };
|
||||
|
||||
class XTTSTtsProvider {
|
||||
//########//
|
||||
// Config //
|
||||
//########//
|
||||
|
||||
settings
|
||||
ready = false
|
||||
voices = []
|
||||
separator = '. '
|
||||
settings;
|
||||
ready = false;
|
||||
voices = [];
|
||||
separator = '. ';
|
||||
|
||||
/**
|
||||
* Perform any text processing before passing to TTS engine.
|
||||
@ -46,13 +46,13 @@ class XTTSTtsProvider {
|
||||
'Korean': 'ko',
|
||||
'Hungarian': 'hu',
|
||||
'Hindi': 'hi',
|
||||
}
|
||||
};
|
||||
|
||||
defaultSettings = {
|
||||
provider_endpoint: 'http://localhost:8020',
|
||||
language: 'en',
|
||||
voiceMap: {}
|
||||
}
|
||||
};
|
||||
|
||||
get settingsHtml() {
|
||||
let html = `
|
||||
@ -64,7 +64,7 @@ class XTTSTtsProvider {
|
||||
|
||||
if (this.languageLabels[language] == this.settings?.language) {
|
||||
html += `<option value="${this.languageLabels[language]}" selected="selected">${language}</option>`;
|
||||
continue
|
||||
continue;
|
||||
}
|
||||
|
||||
html += `<option value="${this.languageLabels[language]}">${language}</option>`;
|
||||
@ -88,25 +88,25 @@ class XTTSTtsProvider {
|
||||
}
|
||||
onSettingsChange() {
|
||||
// Used when provider settings are updated from UI
|
||||
this.settings.provider_endpoint = $('#xtts_tts_endpoint').val()
|
||||
this.settings.language = $('#xtts_api_language').val()
|
||||
saveTtsProviderSettings()
|
||||
this.settings.provider_endpoint = $('#xtts_tts_endpoint').val();
|
||||
this.settings.language = $('#xtts_api_language').val();
|
||||
saveTtsProviderSettings();
|
||||
}
|
||||
|
||||
async loadSettings(settings) {
|
||||
// Pupulate Provider UI given input settings
|
||||
if (Object.keys(settings).length == 0) {
|
||||
console.info('Using default TTS Provider settings')
|
||||
console.info('Using default TTS Provider settings');
|
||||
}
|
||||
|
||||
// Only accept keys defined in defaultSettings
|
||||
this.settings = this.defaultSettings
|
||||
this.settings = this.defaultSettings;
|
||||
|
||||
for (const key in settings) {
|
||||
if (key in this.settings) {
|
||||
this.settings[key] = settings[key]
|
||||
this.settings[key] = settings[key];
|
||||
} else {
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`
|
||||
throw `Invalid setting passed to TTS Provider: ${key}`;
|
||||
}
|
||||
}
|
||||
|
||||
@ -121,23 +121,23 @@ class XTTSTtsProvider {
|
||||
}
|
||||
}, 2000);
|
||||
|
||||
$('#xtts_tts_endpoint').val(this.settings.provider_endpoint)
|
||||
$('#xtts_tts_endpoint').on('input', () => { this.onSettingsChange() })
|
||||
$('#xtts_api_language').val(this.settings.language)
|
||||
$('#xtts_api_language').on('change', () => { this.onSettingsChange() })
|
||||
$('#xtts_tts_endpoint').val(this.settings.provider_endpoint);
|
||||
$('#xtts_tts_endpoint').on('input', () => { this.onSettingsChange(); });
|
||||
$('#xtts_api_language').val(this.settings.language);
|
||||
$('#xtts_api_language').on('change', () => { this.onSettingsChange(); });
|
||||
|
||||
await this.checkReady()
|
||||
await this.checkReady();
|
||||
|
||||
console.debug('XTTS: Settings loaded')
|
||||
console.debug('XTTS: Settings loaded');
|
||||
}
|
||||
|
||||
// Perform a simple readiness check by trying to fetch voiceIds
|
||||
async checkReady() {
|
||||
await this.fetchTtsVoiceObjects()
|
||||
await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
|
||||
async onRefreshClick() {
|
||||
return
|
||||
return;
|
||||
}
|
||||
|
||||
//#################//
|
||||
@ -146,36 +146,36 @@ class XTTSTtsProvider {
|
||||
|
||||
async getVoice(voiceName) {
|
||||
if (this.voices.length == 0) {
|
||||
this.voices = await this.fetchTtsVoiceObjects()
|
||||
this.voices = await this.fetchTtsVoiceObjects();
|
||||
}
|
||||
const match = this.voices.filter(
|
||||
XTTSVoice => XTTSVoice.name == voiceName
|
||||
)[0]
|
||||
)[0];
|
||||
if (!match) {
|
||||
throw `TTS Voice name ${voiceName} not found`
|
||||
throw `TTS Voice name ${voiceName} not found`;
|
||||
}
|
||||
return match
|
||||
return match;
|
||||
}
|
||||
|
||||
async generateTts(text, voiceId) {
|
||||
const response = await this.fetchTtsGeneration(text, voiceId)
|
||||
return response
|
||||
const response = await this.fetchTtsGeneration(text, voiceId);
|
||||
return response;
|
||||
}
|
||||
|
||||
//###########//
|
||||
// API CALLS //
|
||||
//###########//
|
||||
async fetchTtsVoiceObjects() {
|
||||
const response = await doExtrasFetch(`${this.settings.provider_endpoint}/speakers`)
|
||||
const response = await doExtrasFetch(`${this.settings.provider_endpoint}/speakers`);
|
||||
if (!response.ok) {
|
||||
throw new Error(`HTTP ${response.status}: ${await response.json()}`)
|
||||
throw new Error(`HTTP ${response.status}: ${await response.json()}`);
|
||||
}
|
||||
const responseJson = await response.json()
|
||||
return responseJson
|
||||
const responseJson = await response.json();
|
||||
return responseJson;
|
||||
}
|
||||
|
||||
async fetchTtsGeneration(inputText, voiceId) {
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`)
|
||||
console.info(`Generating new TTS for voice_id ${voiceId}`);
|
||||
const response = await doExtrasFetch(
|
||||
`${this.settings.provider_endpoint}/tts_to_audio/`,
|
||||
{
|
||||
@ -190,12 +190,12 @@ class XTTSTtsProvider {
|
||||
'language': this.settings.language
|
||||
})
|
||||
}
|
||||
)
|
||||
);
|
||||
if (!response.ok) {
|
||||
toastr.error(response.statusText, 'TTS Generation Failed');
|
||||
throw new Error(`HTTP ${response.status}: ${await response.text()}`);
|
||||
}
|
||||
return response
|
||||
return response;
|
||||
}
|
||||
|
||||
// Interface not used by XTTS TTS
|
||||
|
||||
Reference in New Issue
Block a user