support playht3.0 languages (#459)

* support playht3.0 languages

* wip

src/api/index.ts

@@ -875,6 +875,7 @@ export const getSpeechSupportedLanguagesAndVoices = (
   sid: string | undefined,
   vendor: string,
   label: string,
+  create_new: boolean = false,
 ) => {
   const userData = parseJwt(getToken());
   const apiUrl =
@@ -883,7 +884,7 @@ export const getSpeechSupportedLanguagesAndVoices = (
       : `${API_SERVICE_PROVIDERS}/${sid}`) +
     `/SpeechCredentials/speech/supportedLanguagesAndVoices?vendor=${vendor}${
       label ? `&label=${label}` : ""
-    }`;
+    }${create_new ? "&create_new=true" : ""}`;
   return getFetch<SpeechSupportedLanguagesAndVoices>(apiUrl);
 };
 

src/containers/internal/views/applications/speech-selection.tsx

@@ -29,6 +29,7 @@ import {
   VENDOR_WELLSAID,
   VENDOR_WHISPER,
   VENDOR_SPEECHMATICS,
+  VENDOR_PLAYHT,
 } from "src/vendor";
 import {
   LabelOptions,
@@ -199,6 +200,19 @@ export const SpeechProviderSelection = ({
         if (synthesisGoogleCustomVoiceOptions.length > 0) {
           updateTtsVoice(synthesisGoogleCustomVoiceOptions[0].value);
         }
+      }
+      // PlayHT3.0 all voices are listed under english language, all voices can be used for multiple languages
+      else if (
+        synthVendor === VENDOR_PLAYHT &&
+        synthesisSupportedLanguagesAndVoices.tts.some(
+          (l) => l.value === "english",
+        )
+      ) {
+        setSynthesisVoiceOptions(
+          synthesisSupportedLanguagesAndVoices.tts.find(
+            (tts) => tts.value === "english",
+          )!.voices,
+        );
       } else {
         setSynthesisVoiceOptions(voicesOpts);
       }
@@ -262,6 +276,14 @@ export const SpeechProviderSelection = ({
             updateTtsVoice(newLang!.voices[0].value);
             return;
           }
+          if (synthVendor === VENDOR_PLAYHT) {
+            const newLang = json.tts.find(
+              (lang) => lang.value === LANG_EN_US || lang.value === "english",
+            );
+            setSynthLang(newLang!.value);
+            updateTtsVoice(newLang!.voices[0].value);
+            return;
+          }
           /** Google and AWS have different language lists */
           /** If the new language doesn't map then default to "en-US" */
           let newLang = json.tts.find((lang) => lang.value === synthLang);
@@ -385,6 +407,7 @@ export const SpeechProviderSelection = ({
               value={synthLabel}
               options={ttsLabelOptions}
               onChange={(e) => {
+                shouldUpdateTtsVoice.current = true;
                 setSynthLabel(e.target.value);
               }}
             />
@@ -412,7 +435,9 @@ export const SpeechProviderSelection = ({
                 id="synthesis_lang"
                 name="synthesis_lang"
                 value={synthLang}
-                options={synthesisLanguageOptions}
+                options={synthesisLanguageOptions.sort((a, b) =>
+                  a.name.localeCompare(b.name),
+                )}
                 onChange={(e) => {
                   shouldUpdateTtsVoice.current = true;
                   const language = e.target.value;
@@ -467,7 +492,9 @@ export const SpeechProviderSelection = ({
                   id="synthesis_voice"
                   name="synthesis_voice"
                   value={synthVoice}
-                  options={synthesisVoiceOptions}
+                  options={synthesisVoiceOptions.sort((a, b) =>
+                    a.name.localeCompare(b.name),
+                  )}
                   onChange={(e) => setSynthVoice(e.target.value)}
                 />
               )}

src/containers/internal/views/speech-services/form.tsx

@@ -473,6 +473,7 @@ export const SpeechServiceForm = ({ credential }: SpeechServiceFormProps) => {
         currentServiceProvider?.service_provider_sid,
         vendor,
         "",
+        credential ? false : true,
       ).then(({ json }) => {
         if (json.models) {
           setTtsModels(json.models);

src/vendor/types.ts

@@ -106,6 +106,7 @@ export interface SynthesisVendors {
   elevenlabs: VoiceLanguage[];
   whisper: VoiceLanguage[];
   deepgram: VoiceLanguage[];
+  playht: VoiceLanguage[];
 }
 
 export interface MSRawSpeech {