Update app.py

app.py

@@ -1,146 +1,101 @@
 import os
-import asyncio
 import tempfile
 import logging
 
 import gradio as gr
-import edge_tts
+import requests
 
-logging.basicConfig(
-    level=logging.INFO,
-    format="%(asctime)s | %(levelname)s | %(name)s | %(message)s",
-)
+logging.basicConfig(level=logging.INFO)
 
 logger = logging.getLogger("edge-tts-bridge")
 
 APP_TITLE = "Edge TTS Bridge"
 
+GO_TTS_URL = "http://127.0.0.1:8080/v1/synthesize"
+
 VOICES = {
     "Myanmar Female": "my-MM-NilarNeural",
     "Myanmar Male": "my-MM-ThihaNeural",
 }
 
 
-async def synthesize(text: str, voice: str) -> str:
-    """Generate speech using Microsoft Edge TTS."""
+def generate_audio(text, voice_name):
+    text = (text or "").strip()
+
+    if not text:
+        raise gr.Error("Please enter text.")
+
+    if len(text) > 3000:
+        raise gr.Error("Maximum 3000 characters per request.")
+
+    if voice_name not in VOICES:
+        raise gr.Error("Unsupported voice.")
 
-    with tempfile.NamedTemporaryFile(
-        suffix=".mp3",
-        delete=False,
-    ) as temp:
-        output_path = temp.name
+    voice = VOICES[voice_name]
 
     try:
         logger.info(
-            "TTS request started | voice=%s | characters=%d",
+            "Generating audio: voice=%s, characters=%d",
             voice,
             len(text),
         )
 
-        communicate = edge_tts.Communicate(
-            text=text,
-            voice=voice,
+        response = requests.post(
+            GO_TTS_URL,
+            json={
+                "text": text,
+                "voice": voice,
+            },
+            timeout=120,
         )
 
-        await communicate.save(output_path)
-
-        if not os.path.exists(output_path):
-            raise RuntimeError("Audio file was not created.")
-
-        file_size = os.path.getsize(output_path)
-
-        if file_size < 100:
+        if response.status_code != 200:
             raise RuntimeError(
-                f"Audio file is unexpectedly small: {file_size} bytes"
+                f"Go TTS HTTP {response.status_code}: "
+                f"{response.text[:400]}"
             )
 
-        logger.info(
-            "TTS request succeeded | voice=%s | file_size=%d",
-            voice,
-            file_size,
-        )
+        audio_data = response.content
 
-        return output_path
+        if len(audio_data) < 100:
+            raise RuntimeError("Audio response is too small.")
 
-    except Exception as exc:
-        logger.exception(
-            "TTS synthesis failed | voice=%s | "
-            "error_type=%s | error=%r",
-            voice,
-            type(exc).__name__,
-            exc,
-        )
-
-        try:
-            os.remove(output_path)
-        except OSError:
-            pass
-
-        raise
-
-
-def generate_audio(text, voice_name):
-    """Gradio interface handler."""
+        with tempfile.NamedTemporaryFile(
+            suffix=".mp3",
+            delete=False,
+        ) as temp:
+            temp.write(audio_data)
+            output_path = temp.name
 
-    text = (text or "").strip()
-
-    if not text:
-        raise gr.Error("Please enter text.")
-
-    if len(text) > 3000:
-        raise gr.Error(
-            "Maximum 3000 characters per request."
+        logger.info(
+            "Audio generated successfully: %d bytes",
+            len(audio_data),
         )
 
-    if voice_name not in VOICES:
-        raise gr.Error("Unsupported voice.")
-
-    voice = VOICES[voice_name]
-
-    try:
-        return asyncio.run(
-            synthesize(text, voice)
-        )
+        return output_path
 
     except Exception as exc:
-        logger.exception(
-            "Edge TTS handler failed | type=%s",
-            type(exc).__name__,
-        )
-
-        message = str(exc).strip() or repr(exc)
+        logger.exception("TTS failed")
 
         raise gr.Error(
-            f"Edge TTS failed ({type(exc).__name__}): "
-            f"{message[:500]}"
+            f"TTS failed ({type(exc).__name__}): {str(exc)[:400]}"
         )
 
 
 with gr.Blocks(title=APP_TITLE) as demo:
-    gr.Markdown(
-        """
-        # Edge TTS Bridge
-
-        Test Microsoft Edge TTS from Hugging Face Spaces.
+    gr.Markdown("# Edge TTS Bridge — Go Test")
 
-        This service requires an active connection to
-        Microsoft's speech service.
-        """
+    text_input = gr.Textbox(
+        label="Text",
+        placeholder="မင်္ဂလာပါ၊ အသံစမ်းသပ်နေပါတယ်။",
+        lines=4,
     )
 
-    with gr.Row():
-        text_input = gr.Textbox(
-            label="Text",
-            placeholder="မင်္ဂလာပါ၊ အသံစမ်းသပ်နေပါတယ်။",
-            lines=4,
-            max_lines=10,
-        )
-
-        voice_input = gr.Dropdown(
-            choices=list(VOICES.keys()),
-            value="Myanmar Female",
-            label="Voice",
-        )
+    voice_input = gr.Dropdown(
+        choices=list(VOICES.keys()),
+        value="Myanmar Female",
+        label="Voice",
+    )
 
     generate_button = gr.Button(
         "Generate Audio",
@@ -154,25 +109,14 @@ with gr.Blocks(title=APP_TITLE) as demo:
 
     generate_button.click(
         fn=generate_audio,
-        inputs=[
-            text_input,
-            voice_input,
-        ],
+        inputs=[text_input, voice_input],
         outputs=audio_output,
         api_name="generate",
     )
 
-    gr.Markdown(
-        """
-        ## Test instructions
-
-        1. Enter a short Myanmar sentence.
-        2. Select a voice.
-        3. Click Generate Audio.
-        4. Check whether the MP3 is playable.
-        """
-    )
-
 
 if __name__ == "__main__":
-    demo.queue(max_size=5).launch()
+    demo.queue(max_size=5).launch(
+        server_name="0.0.0.0",
+        server_port=7860,
+    )