Merge pull request #1397 from arc53/tts

Add endpoint for Text-To-Speech conversion

application/api/user/routes.py

@@ -17,6 +17,7 @@
 from application.extensions import api
 from application.utils import check_required_fields
 from application.vectorstore.vector_creator import VectorCreator
+from application.tts.google_tts import GoogleTTS
 
 mongo = MongoClient(settings.MONGO_URI)
 db = mongo["docsgpt"]
@@ -1663,3 +1664,27 @@ def post(self):
             return make_response(jsonify({"success": False, "error": str(err)}), 400)
 
         return make_response(jsonify({"success": True}), 200)
+
+
+@user_ns.route("/api/tts")
+class TextToSpeech(Resource):
+    tts_model = api.model(
+        "TextToSpeechModel",
+        {
+            "text": fields.String(required=True, description="Text to be synthesized as audio"),
+        },
+    )
+
+    @api.expect(tts_model)
+    @api.doc(description="Synthesize audio speech from text")
+    def post(self):
+        data = request.get_json()
+        text = data["text"]
+        try:
+            tts_instance = GoogleTTS(text)
+            audio_base64, detected_language = tts_instance.text_to_speech()
+            return make_response(jsonify({"success": True,'audio_base64': audio_base64,'lang':detected_language}), 200)
+        except Exception as err:
+            return make_response(jsonify({"success": False, "error": str(err)}), 400)
+
+

application/requirements.txt

@@ -85,3 +85,4 @@ vine==5.1.0
 wcwidth==0.2.13
 werkzeug==3.0.4
 yarl==1.11.1
+gTTS==2.3.2

application/tts/base.py

@@ -0,0 +1,10 @@
+from abc import ABC, abstractmethod
+
+
+class BaseTTS(ABC):
+    def __init__(self):
+        pass
+
+    @abstractmethod
+    def text_to_speech(self, *args, **kwargs):
+        pass

application/tts/google_tts.py

@@ -0,0 +1,19 @@
+import io
+import base64
+from gtts import gTTS
+from application.tts.base import BaseTTS
+
+
+class GoogleTTS(BaseTTS):
+    def __init__(self, text):
+        self.text = text
+
+
+    def text_to_speech(self):
+        lang = "en"
+        audio_fp = io.BytesIO()
+        tts = gTTS(text=self.text, lang=lang, slow=False)
+        tts.write_to_fp(audio_fp)
+        audio_fp.seek(0)
+        audio_base64 = base64.b64encode(audio_fp.read()).decode("utf-8")
+        return audio_base64, lang