app.py

import os
import base64
import json
from datetime import datetime
import traceback
from typing import Optional, Dict, Any

import gradio as gr
from huggingface_hub import HfApi, InferenceClient
from fastmcp import FastMCP


HF_DATASET_REPO = os.environ.get("HF_DATASET_REPO", "OppaAI/Robot_MCP")
HF_VLM_MODEL = os.environ.get("HF_VLM_MODEL", "Qwen/Qwen2.5-VL-7B-Instruct")

mcp = FastMCP("Robot_MCP_Server")


# -------------------------------
# Upload helper
# -------------------------------
def upload_image(image_b64: str, hf_token: str):
    try:
        image_bytes = base64.b64decode(image_b64)
        size_bytes = len(image_bytes)

        os.makedirs("/tmp", exist_ok=True)
        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
        local_path = f"/tmp/robot_img_{timestamp}.jpg"

        with open(local_path, "wb") as f:
            f.write(image_bytes)

        filename = f"robot_{timestamp}.jpg"

        api = HfApi()
        api.upload_file(
            path_or_fileobj=local_path,
            path_in_repo=f"tmp/{filename}",
            repo_id=HF_DATASET_REPO,
            repo_type="dataset",
            token=hf_token
        )

        url = f"https://huggingface.co/datasets/{HF_DATASET_REPO}/resolve/main/tmp/{filename}"
        return local_path, url, filename, size_bytes

    except Exception:
        traceback.print_exc()
        return None, None, None, 0


# -------------------------------
# Safe JSON parse
# -------------------------------
def safe_parse_json_from_text(text: str):
    if not text:
        return None
    try:
        return json.loads(text)
    except:
        pass

    cleaned = text.strip().strip("`").strip()
    if cleaned.lower().startswith("json"):
        cleaned = cleaned[4:].strip()

    try:
        start = cleaned.find("{")
        end = cleaned.rfind("}")
        return json.loads(cleaned[start:end + 1])
    except:
        return None


# -------------------------------
# TRUE CORE FUNCTION
# -------------------------------
# -------------------------------
# TRUE CORE FUNCTION (with objects)
# -------------------------------
def robot_watch_core(payload: Dict[str, Any]):
    if isinstance(payload, str):
        try:
            payload = json.loads(payload)
        except:
            return {"error": "Invalid JSON payload"}

    hf_token = payload.get("hf_token")
    if not hf_token:
        return {"error": "hf_token missing"}

    robot_id = payload.get("robot_id", "unknown")
    image_b64 = payload.get("image_b64")
    if not image_b64:
        return {"error": "image_b64 missing"}

    # Upload
    _, hf_url, _, size_bytes = upload_image(image_b64, hf_token)
    if not hf_url:
        return {"error": "Image upload failed"}

    # VLM
    system_prompt = """
Respond in STRICT JSON ONLY.
{
 "description": "...",
 "human": "...",
 "environment": "...",
 "objects": []  // list of detected objects
}
"""

    messages = [
        {"role": "system", "content": system_prompt},
        {"role": "user", "content": [
            {"type": "text", "text": "Analyze the image."},
            {"type": "image_url", "image_url": {"url": f"data:image/jpeg;base64,{image_b64}"}}
        ]}
    ]

    client = InferenceClient(token=hf_token)

    try:
        resp = client.chat.completions.create(
            model=HF_VLM_MODEL,
            messages=messages,
            max_tokens=500,
            temperature=0.1
        )
    except Exception as e:
        return {"status": "error", "message": str(e)}

    vlm_output = resp.choices[0].message.content.strip()
    parsed = safe_parse_json_from_text(vlm_output)

    if parsed is None:
        return {
            "status": "model_no_json",
            "vlm_raw": vlm_output,
            "message": "Invalid JSON returned"
        }

    # Ensure "objects" is a list
    objects = parsed.get("objects", [])
    if not isinstance(objects, list):
        objects = []

    return {
        "status": "success",
        "robot_id": robot_id,
        "file_size_bytes": size_bytes,
        "image_url": hf_url,
        "description": parsed.get("description"),
        "human": parsed.get("human"),
        "environment": parsed.get("environment"),
        "objects": objects,   # ← new field
        "vlm_raw": vlm_output
    }



# -------------------------------
# REGISTER MCP TOOL (wrapper)
# -------------------------------
@mcp.tool()
def robot_watch(payload: Dict[str, Any]):
    return robot_watch_core(payload)


# -------------------------------
# Gradio wrapper
# -------------------------------
def process_json(payload):
    return robot_watch_core(payload)


app = gr.Interface(
    fn=process_json,
    inputs=gr.JSON(label="Input JSON"),
    outputs=gr.JSON(label="Result JSON"),
    title="Robot MCP Server",
    description="JSON endpoint for robot vision pipeline.",
    api_name="predict"
)


if __name__ == "__main__":
    app.launch(mcp_server=True)