adding in tts logic with gcp

voice-server/app.py

@@ -1,5 +1,6 @@
-from flask import Flask, request, abort
+from flask import Flask, request, abort, send_file
 from tempfile import NamedTemporaryFile
+import tts
 import whisper
 
 # Load the Whisper model:
@@ -44,6 +45,24 @@ def transcribe():
     return {"results": results}
 
 
+# endpoint to convert text to speech and return the file
+@app.route("/tts", methods=["GET"])
+def text_to_speech():
+    # get the text from the request
+    text = request.form.get("text")
+    if not text:
+        abort(400, "Please provide text to convert to speech!")
+
+    # call the text_to_wav function from tts.py
+    fileName = tts.text_to_wav(
+        voice_name="en-GB-Neural2-B",
+        text=text,
+    )
+
+    # return the file
+    return send_file(f"{fileName}.wav", mimetype="audio/wav")
+
+
 if __name__ == "__main__":
     # 0.0.0.0 opens up to all ipv4 addresses
     app.run(host="0.0.0.0", port=5001, debug=True)

voice-server/tts.py

@@ -0,0 +1,24 @@
+import google.cloud.texttospeech as tts
+
+
+def text_to_wav(voice_name: str, text: str):
+    language_code = "-".join(voice_name.split("-")[:2])
+    text_input = tts.SynthesisInput(text=text)
+    voice_params = tts.VoiceSelectionParams(
+        language_code=language_code, name=voice_name
+    )
+    audio_config = tts.AudioConfig(audio_encoding=tts.AudioEncoding.LINEAR16)
+
+    client = tts.TextToSpeechClient()
+    response = client.synthesize_speech(
+        input=text_input,
+        voice=voice_params,
+        audio_config=audio_config,
+    )
+
+    filename = f"{voice_name}.wav"
+    with open(filename, "wb") as out:
+        out.write(response.audio_content)
+        print(f'Generated speech saved to "{filename}"')
+
+    return filename