Update app.py

app.py

@@ -1,4 +1,4 @@
-# app.py — FastAPI REST API + mounted Gradio UI (Hugging Face Spaces compatible)
+# app.py — FastAPI + Gradio (External API + UI)
 import os
 import shutil
 import asyncio
@@ -6,31 +6,28 @@ import inspect
 from typing import Optional
 
 from fastapi import FastAPI, UploadFile, File, Form
-from fastapi.responses import JSONResponse, PlainTextResponse
 from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import JSONResponse
 import gradio as gr
-import uvicorn
 
-# ---- Your module ----
 from multimodal_module import MultiModalChatModule
 
-# Instantiate once at import time
+# Instantiate AI module
 AI = MultiModalChatModule()
 
 TMP_DIR = "/tmp"
 os.makedirs(TMP_DIR, exist_ok=True)
 
-# ---------------- Helpers ----------------
+# --- File wrapper ---
 class FileWrapper:
-    """Tiny adapter so your module can .download_to_drive(path)."""
     def __init__(self, path: str):
         self._path = path
-
-    async def download_to_drive(self, dst_path: str) -> None:
+    async def download_to_drive(self, dst_path: str):
         loop = asyncio.get_event_loop()
         await loop.run_in_executor(None, shutil.copyfile, self._path, dst_path)
 
-async def save_upload_to_tmp(up: UploadFile) -> str:
+# --- Save uploaded file ---
+async def save_upload(up: UploadFile) -> str:
     if not up or not up.filename:
         raise ValueError("No file uploaded")
     dest = os.path.join(TMP_DIR, up.filename)
@@ -39,48 +36,28 @@ async def save_upload_to_tmp(up: UploadFile) -> str:
         f.write(data)
     return dest
 
+# --- Call AI (sync or async) ---
 async def call_ai(fn, *args, **kwargs):
-    """Call AI methods whether they are sync or async."""
     if fn is None:
-        raise AttributeError("Requested AI method is not implemented in multimodal_module")
+        raise AttributeError("Requested AI method not implemented")
     if inspect.iscoroutinefunction(fn):
         return await fn(*args, **kwargs)
     return await asyncio.to_thread(lambda: fn(*args, **kwargs))
 
-# ---------------- FastAPI app ----------------
-app = FastAPI(title="Multimodal Module API", version="1.0.0")
+# === FASTAPI APP ===
+app = FastAPI(title="Multimodal API")
 
-# CORS so external apps can call it
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],          # tighten for production
+    allow_origins=["*"],  # change for production
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
 )
 
-# ---- Health / root ----
-@app.get("/health", response_class=PlainTextResponse)
-async def health():
-    return "ok"
-
-@app.get("/")
-async def root():
-    return {
-        "name": "Multimodal Module API",
-        "status": "ready",
-        "docs": "/docs",
-        "gradio_ui": "/ui"
-    }
-
-# ---------------- REST Endpoints ----------------
-# Text chat
+# --- API Endpoints ---
 @app.post("/api/text")
-async def api_text(
-    text: str = Form(...),
-    user_id: Optional[int] = Form(0),
-    lang: str = Form("en"),
-):
+async def api_text(text: str = Form(...), user_id: Optional[int] = Form(0), lang: str = Form("en")):
     try:
         fn = getattr(AI, "generate_response", getattr(AI, "process_text", None))
         reply = await call_ai(fn, text, int(user_id), lang)
@@ -88,71 +65,37 @@ async def api_text(
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# Hugging Face-style predict (optional)
-@app.post("/api/predict")
-async def api_predict(
-    inputs: str = Form(...),
-    user_id: Optional[int] = Form(0),
-    lang: str = Form("en"),
-):
-    try:
-        fn = getattr(AI, "generate_response", getattr(AI, "process_text", None))
-        reply = await call_ai(fn, inputs, int(user_id), lang)
-        return {"data": [reply]}
-    except Exception as e:
-        return JSONResponse({"error": str(e)}, status_code=500)
-
-# Voice -> ASR / emotion
 @app.post("/api/voice")
-async def api_voice(
-    user_id: Optional[int] = Form(0),
-    audio_file: UploadFile = File(...),
-):
+async def api_voice(user_id: Optional[int] = Form(0), audio_file: UploadFile = File(...)):
     try:
-        path = await save_upload_to_tmp(audio_file)
+        path = await save_upload(audio_file)
         fn = getattr(AI, "process_voice_message", None)
         result = await call_ai(fn, FileWrapper(path), int(user_id))
-        return JSONResponse(result)
+        return {"status": "ok", "result": result}
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# TTS
 @app.post("/api/voice_reply")
-async def api_voice_reply(
-    user_id: Optional[int] = Form(0),
-    reply_text: str = Form(...),
-    fmt: str = Form("ogg"),
-):
+async def api_voice_reply(user_id: Optional[int] = Form(0), reply_text: str = Form(...), fmt: str = Form("ogg")):
     try:
         fn = getattr(AI, "generate_voice_reply", None)
-        out_path = await call_ai(fn, reply_text, int(user_id), fmt)
-        return {"status": "ok", "file": out_path}
+        result = await call_ai(fn, reply_text, int(user_id), fmt)
+        return {"status": "ok", "file": result}
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# Image caption
 @app.post("/api/image_caption")
-async def api_image_caption(
-    user_id: Optional[int] = Form(0),
-    image_file: UploadFile = File(...),
-):
+async def api_image_caption(user_id: Optional[int] = Form(0), image_file: UploadFile = File(...)):
     try:
-        path = await save_upload_to_tmp(image_file)
+        path = await save_upload(image_file)
         fn = getattr(AI, "process_image_message", None)
         caption = await call_ai(fn, FileWrapper(path), int(user_id))
         return {"status": "ok", "caption": caption}
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# Text-to-image
 @app.post("/api/generate_image")
-async def api_generate_image(
-    user_id: Optional[int] = Form(0),
-    prompt: str = Form(...),
-    width: int = Form(512),
-    height: int = Form(512),
-    steps: int = Form(30),
-):
+async def api_generate_image(user_id: Optional[int] = Form(0), prompt: str = Form(...), width: int = Form(512), height: int = Form(512), steps: int = Form(30)):
     try:
         fn = getattr(AI, "generate_image_from_text", None)
         out_path = await call_ai(fn, prompt, int(user_id), width, height, steps)
@@ -160,66 +103,41 @@ async def api_generate_image(
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# Image edit / inpaint
 @app.post("/api/edit_image")
-async def api_edit_image(
-    user_id: Optional[int] = Form(0),
-    image_file: UploadFile = File(...),
-    mask_file: Optional[UploadFile] = File(None),
-    prompt: str = Form(""),
-):
+async def api_edit_image(user_id: Optional[int] = Form(0), image_file: UploadFile = File(...), mask_file: Optional[UploadFile] = File(None), prompt: str = Form("")):
     try:
-        img_path = await save_upload_to_tmp(image_file)
+        img_path = await save_upload(image_file)
         mask_path = None
         if mask_file:
-            mask_path = await save_upload_to_tmp(mask_file)
+            mask_path = await save_upload(mask_file)
         fn = getattr(AI, "edit_image_inpaint", None)
-        out_path = await call_ai(
-            fn,
-            FileWrapper(img_path),
-            FileWrapper(mask_path) if mask_path else None,
-            prompt,
-            int(user_id),
-        )
+        out_path = await call_ai(fn, FileWrapper(img_path), FileWrapper(mask_path) if mask_path else None, prompt, int(user_id))
         return {"status": "ok", "file": out_path}
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# Video
 @app.post("/api/video")
-async def api_video(
-    user_id: Optional[int] = Form(0),
-    video_file: UploadFile = File(...),
-):
+async def api_video(user_id: Optional[int] = Form(0), video_file: UploadFile = File(...)):
     try:
-        path = await save_upload_to_tmp(video_file)
+        path = await save_upload(video_file)
         fn = getattr(AI, "process_video", None)
         result = await call_ai(fn, FileWrapper(path), int(user_id))
-        return JSONResponse(result)
+        return {"status": "ok", "result": result}
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# Files (PDF/DOCX/TXT)
 @app.post("/api/file")
-async def api_file(
-    user_id: Optional[int] = Form(0),
-    file_obj: UploadFile = File(...),
-):
+async def api_file(user_id: Optional[int] = Form(0), file_obj: UploadFile = File(...)):
     try:
-        path = await save_upload_to_tmp(file_obj)
+        path = await save_upload(file_obj)
         fn = getattr(AI, "process_file", None)
         result = await call_ai(fn, FileWrapper(path), int(user_id))
-        return JSONResponse(result)
+        return {"status": "ok", "result": result}
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# Code completion
 @app.post("/api/code")
-async def api_code(
-    user_id: Optional[int] = Form(0),
-    prompt: str = Form(...),
-    max_tokens: int = Form(512),
-):
+async def api_code(user_id: Optional[int] = Form(0), prompt: str = Form(...), max_tokens: int = Form(512)):
     try:
         fn = getattr(AI, "code_complete", None)
         try:
@@ -230,28 +148,20 @@ async def api_code(
     except Exception as e:
         return JSONResponse({"error": str(e)}, status_code=500)
 
-# ---------------- Gradio UI (mounted at /ui) ----------------
-def _gradio_text_fn(text, user_id, lang):
+# === GRADIO UI ===
+def gradio_text_fn(text, user_id, lang):
     fn = getattr(AI, "generate_response", getattr(AI, "process_text", None))
-    if fn is None:
-        return "Error: text handler not implemented in multimodal_module"
-    # Gradio callbacks run in a worker thread, safe to create/own an event loop
-    return asyncio.run(call_ai(fn, text, int(user_id or 0), lang))
+    loop = asyncio.get_event_loop()
+    return loop.run_until_complete(call_ai(fn, text, int(user_id or 0), lang))
 
-with gr.Blocks(title="Multimodal Bot — UI") as demo:
-    gr.Markdown("# 🧠 Multimodal Bot — UI\nThis is a helper UI. Use the REST API for external apps.")
+with gr.Blocks(title="Multimodal Bot") as demo:
+    gr.Markdown("# 🧠 Multimodal Bot — UI")
     with gr.Row():
-        g_uid = gr.Textbox(label="User ID", value="0")
-        g_lang = gr.Dropdown(["en", "zh", "ja", "ko", "es", "fr", "de", "it"], value="en", label="Language")
-    g_in = gr.Textbox(lines=3, label="Message")
-    g_out = gr.Textbox(lines=6, label="Reply")
-    gr.Button("Send").click(_gradio_text_fn, [g_in, g_uid, g_lang], g_out)
-
-# Mount Gradio *into* FastAPI at /ui (does not open another port)
-app = gr.mount_gradio_app(app, demo, path="/ui")
-
-# ---------------- Entrypoint ----------------
-if __name__ == "__main__":
-    # Hugging Face Spaces (FastAPI template) sets PORT; bind exactly to it.
-    port = int(os.environ.get("PORT", "7860"))
-    uvicorn.run("app:app", host="0.0.0.0", port=port)
+        uid = gr.Textbox(label="User ID", value="0")
+        lang = gr.Dropdown(["en", "zh", "ja", "ko", "es", "fr", "de", "it"], value="en", label="Language")
+    inp = gr.Textbox(lines=3, label="Message")
+    out = gr.Textbox(lines=6, label="Reply")
+    gr.Button("Send").click(gradio_text_fn, [inp, uid, lang], out)
+
+# Mount Gradio under /ui
+app = gr.mount_gradio_app(app, demo, path="/ui")