app.py

import os
import json
from fastapi import Request
from fastapi.responses import HTMLResponse, StreamingResponse, JSONResponse
from gradio import Server
from gradio.oauth import attach_oauth, _get_valid_oauth_info_from_session
from openai import AsyncOpenAI
from pydantic import BaseModel
from typing import List

# Initialize gradio.Server (which is a subclass of FastAPI)
app = Server()

# Delegate Hugging Face OAuth to Gradio's battle-tested implementation.
# attach_oauth(app) registers /login/huggingface, /login/callback, /logout, AND
# the SessionMiddleware. After it runs, the user's token + profile are stored
# in request.session['oauth_info'] under the keys authlib returns
# (access_token, expires_at, userinfo, ...).
attach_oauth(app)

# Define request schemas
class Message(BaseModel):
    role: str
    content: str

class ChatRequest(BaseModel):
    messages: List[Message]
    model: str = "zai-org/GLM-5.2:fireworks-ai"
    temperature: float = 0.7
    max_tokens: int = 2048


@app.get("/me")
async def me(request: Request):
    """Expose the current user's profile (or null) to the frontend."""
    oauth_info = _get_valid_oauth_info_from_session(request.session) or {}
    user_info = oauth_info.get("userinfo") or {}
    if not oauth_info or not user_info:
        return JSONResponse({"user": None})
    return JSONResponse({"user": user_info})


@app.post("/api/chat")
async def chat_endpoint(request: Request, payload: ChatRequest):
    # Determine which token pays for this inference:
    #   1. The logged-in user's OAuth access token (bills the USER — the goal).
    #   2. Fall back to the host's HF_TOKEN (bills the Space owner) if configured.
    oauth_info = _get_valid_oauth_info_from_session(request.session) or {}
    user_token = oauth_info.get("access_token", "") or ""
    host_token = os.environ.get("HF_TOKEN", "").strip()
    api_key = (user_token or host_token).strip()

    # Track who is being billed so the UI can show it.
    billed_to = "user" if user_token else ("host" if host_token else "none")

    async def error_generator(msg):
        yield "data: " + json.dumps({"error": msg}) + "\n\n"
        yield "data: [DONE]\n\n"

    headers = {
        "Cache-Control": "no-cache",
        "Connection": "keep-alive",
        "X-Accel-Buffering": "no",
    }

    if not api_key:
        return StreamingResponse(
            error_generator(
                "Not signed in. Click the Hugging Face button in the sidebar to sign in "
                "and run inference billed to your own HF account."
            ),
            media_type="text/event-stream",
            headers=headers,
        )

    # Initialize AsyncOpenAI client pointing at Hugging Face Router.
    # When authenticated with the user's token, the HF router bills that user.
    client = AsyncOpenAI(
        base_url="https://router.huggingface.co/v1",
        api_key=api_key,
    )

    async def event_generator():
        try:
            # Call completion API asynchronously with stream=True
            stream = await client.chat.completions.create(
                model="zai-org/GLM-5.2:fireworks-ai",
                messages=[{"role": m.role, "content": m.content} for m in payload.messages],
                temperature=payload.temperature,
                max_tokens=payload.max_tokens,
                stream=True,
            )

            # Yield chunks as they arrive asynchronously
            async for chunk in stream:
                if chunk.choices and len(chunk.choices) > 0:
                    delta_content = chunk.choices[0].delta.content
                    if delta_content:
                        yield f"data: {json.dumps({'content': delta_content, 'billedTo': billed_to})}\n\n"

            # Signal stream completion
            yield "data: [DONE]\n\n"

        except Exception as e:
            # Yield any error occurring during streaming
            yield f"data: {json.dumps({'error': str(e)})}\n\n"
            yield "data: [DONE]\n\n"

    # Configure headers to bypass buffer layers in reverse proxies (like Nginx/Cloudflare)
    return StreamingResponse(event_generator(), media_type="text/event-stream", headers=headers)


@app.get("/", response_class=HTMLResponse)
async def homepage():
    html_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), "index.html")
    if not os.path.exists(html_path):
        return HTMLResponse("<h3>index.html not found. Please ensure it is created in the workspace directory.</h3>", status_code=404)
    with open(html_path, "r", encoding="utf-8") as f:
        return HTMLResponse(f.read())


if __name__ == "__main__":
    # Launch Gradio Server (which binds to the FastAPI app underneath)
    app.launch(show_error=True)