Update app.py

app.py

@@ -1,256 +1,64 @@
-import os
-import base64
-import tempfile
-import logging
-
 import gradio as gr
-import requests
-
-logging.basicConfig(
-    level=logging.INFO,
-    format="%(asctime)s | %(levelname)s | %(name)s | %(message)s",
-)
-
-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",
-}
-
-
-def synthesize(text: str, voice: str) -> str:
-    """Generate audio through the Go Edge TTS server."""
-
-    output_path = None
-
-    try:
-        logger.info(
-            "TTS request started | voice=%s | characters=%d",
-            voice,
-            len(text),
-        )
-
-        response = requests.post(
-            GO_TTS_URL,
-            json={
-                "text": text,
-                "voice": voice,
-            },
-            timeout=120,
-        )
-
-        logger.info(
-            "Go TTS response | status=%s | content_type=%s | bytes=%d",
-            response.status_code,
-            response.headers.get("content-type", ""),
-            len(response.content),
-        )
-
-        if response.status_code != 200:
-            raise RuntimeError(
-                f"Go TTS HTTP {response.status_code}: "
-                f"{response.text[:500]}"
-            )
-
-        content_type = response.headers.get(
-            "content-type",
-            "",
-        ).lower()
-
-        # Case 1: Server returns audio directly.
-        if content_type.startswith("audio/"):
-            audio_data = response.content
-
-        # Case 2: Server returns JSON.
-        else:
-            try:
-                result = response.json()
-            except ValueError:
-                raise RuntimeError(
-                    "Unexpected response from Go TTS: "
-                    f"{response.text[:500]}"
-                )
-
-            logger.info(
-                "Go TTS JSON response: %s",
-                str(result)[:1000],
-            )
-
-            if not isinstance(result, dict):
-                raise RuntimeError(
-                    "Go TTS returned unexpected JSON format."
-                )
-
-            # Surface server-side errors.
-            if result.get("error"):
-                raise RuntimeError(
-                    str(result["error"])[:500]
-                )
-
-            # Common Base64 audio fields.
-            audio_data = None
-
-            for key in (
-                "audio",
-                "data",
-                "audioContent",
-                "audio_content",
-            ):
-                value = result.get(key)
-
-                if isinstance(value, str) and value:
-                    try:
-                        if value.startswith("data:") and "," in value:
-                            value = value.split(",", 1)[1]
-
-                        audio_data = base64.b64decode(
-                            value,
-                            validate=True,
-                        )
-
-                        if audio_data:
-                            break
-
-                    except Exception:
-                        audio_data = None
-
-            if not audio_data:
-                raise RuntimeError(
-                    "Go TTS did not return recognizable audio. "
-                    "Response: "
-                    f"{str(result)[:500]}"
-                )
-
-        if len(audio_data) < 100:
-            raise RuntimeError(
-                "Audio response is too small: "
-                f"{len(audio_data)} bytes."
-            )
-
-        with tempfile.NamedTemporaryFile(
-            suffix=".mp3",
-            delete=False,
-        ) as temp:
-            temp.write(audio_data)
-            output_path = temp.name
-
-        logger.info(
-            "TTS succeeded | voice=%s | bytes=%d",
-            voice,
-            len(audio_data),
-        )
-
-        return output_path
-
-    except Exception:
-        if output_path and os.path.exists(output_path):
-            try:
-                os.remove(output_path)
-            except OSError:
-                pass
-
-        logger.exception(
-            "TTS synthesis failed | voice=%s",
-            voice,
-        )
-        raise
-
-
-def generate_audio(text, voice_name):
-    """Gradio interface handler."""
+import edge_tts
+import asyncio
+import tempfile
+import os
 
-    text = (text or "").strip()
+# မြန်မာနဲ့ အင်္ဂလိပ် အသံတချို့ကို Default ထည့်ပေးထားပါတယ်
+VOICES = [
+    "my-MM-NilarNeural",
+    "my-MM-ThihaNeural",
+    "en-US-AriaNeural",
+    "en-US-GuyNeural"
+]
 
+async def generate_tts(text, voice, proxy):
     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.")
-
-    voice = VOICES[voice_name]
-
+        return None, "ကျေးဇူးပြု၍ စာသားထည့်ပါ။"
+    
     try:
-        return synthesize(text, voice)
-
-    except Exception as exc:
-        logger.exception(
-            "TTS handler failed | type=%s",
-            type(exc).__name__,
-        )
-
-        raise gr.Error(
-            f"TTS failed ({type(exc).__name__}): "
-            f"{str(exc)[:500]}"
-        )
-
-
-with gr.Blocks(title=APP_TITLE) as demo:
-    gr.Markdown(
-        """
-        # Edge TTS Bridge — Go Test
-
-        Test Microsoft Edge TTS using the Go server.
-        """
-    )
-
+        # အသံဖိုင်သိမ်းရန် Temporary file ဖန်တီးခြင်း
+        temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
+        temp_file.close()
+        
+        # Proxy ထည့်ထားရင် သုံးမယ်၊ မထည့်ထားရင် None အနေနဲ့ ပုံမှန်အတိုင်း အလုပ်လုပ်မယ်
+        proxy_url = proxy if proxy.strip() != "" else None
+        
+        # edge_tts ကို proxy parameter ထည့်ပြီး ခေါ်ခြင်း
+        communicate = edge_tts.Communicate(text, voice, proxy=proxy_url)
+        await communicate.save(temp_file.name)
+        
+        return temp_file.name, "အောင်မြင်ပါသည်။"
+    except Exception as e:
+        return None, f"Error: {str(e)}"
+
+def tts_interface(text, voice, proxy):
+    return asyncio.run(generate_tts(text, voice, proxy))
+
+# Gradio Web UI ဒီဇိုင်း
+with gr.Blocks(title="Edge TTS with Proxy") as app:
+    gr.Markdown("## Hugging Face Space အတွက် Edge TTS (Proxy Support)")
+    gr.Markdown("Proxy ကို `http://ip:port` သို့မဟုတ် `http://user:pass@ip:port` ပုံစံဖြင့် ထည့်သွင်းအသုံးပြုနိုင်ပါသည်။")
+    
     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",
-        )
-
-    generate_button = gr.Button(
-        "Generate Audio",
-        variant="primary",
-    )
-
-    audio_output = gr.Audio(
-        label="Generated Audio",
-        type="filepath",
-    )
-
-    generate_button.click(
-        fn=generate_audio,
-        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.
-        """
+        with gr.Column():
+            text_input = gr.Textbox(label="စာသားထည့်ပါ (Text)", lines=7, placeholder="မြန်မာ သို့မဟုတ် အင်္ဂလိပ်စာသား ထည့်ပါ...")
+            voice_dropdown = gr.Dropdown(choices=VOICES, value="my-MM-NilarNeural", label="အသံရွေးချယ်ပါ (Voice)")
+            proxy_input = gr.Textbox(
+                label="Proxy URL (Optional)", 
+                placeholder="http://123.45.67.89:8080 (RunPod သို့မဟုတ် အခြား Proxy ကို ထည့်ပါ)"
+            )
+            submit_btn = gr.Button("အသံဖန်တီးမည် (Generate TTS)", variant="primary")
+            
+        with gr.Column():
+            audio_output = gr.Audio(label="ထွက်ပေါ်လာသောအသံ (Output Audio)", type="filepath")
+            status_output = gr.Textbox(label="အခြေအနေ (Status)")
+            
+    submit_btn.click(
+        fn=tts_interface,
+        inputs=[text_input, voice_dropdown, proxy_input],
+        outputs=[audio_output, status_output]
     )
 
-
 if __name__ == "__main__":
-    demo.queue(max_size=5).launch(
-        server_name="0.0.0.0",
-        server_port=7860,
-    )
+    app.launch()