# ============================================================
# Rendered Frame Theory — Stabilising System Verification Panel
# Author: Liam Grinstead | RFT Systems | All Rights Reserved
# ============================================================

import json, random
from datetime import datetime
import numpy as np
import gradio as gr

# ------------------ About / Legal ---------------------------
RFT_VERSION = "v4.0-verification-panel"
RFT_DOI = "https://doi.org/10.5281/zenodo.17466722"
LEGAL_NOTICE = (
    "All Rights Reserved — RFT-IPURL v1.0 (UK / Berne). "
    "Research validation use only. No reverse-engineering without written consent."
)

# ------------------ System Profiles -------------------------
PROFILES = {
    "AI / Neural":          {"base": (0.86, 0.80), "w": (0.65, 0.35)},
    "SpaceX / Aerospace":   {"base": (0.84, 0.79), "w": (0.60, 0.40)},
    "Energy / RHES":        {"base": (0.83, 0.78), "w": (0.55, 0.45)},
    "Extreme Perturbation": {"base": (0.82, 0.77), "w": (0.50, 0.50)},
}

# ------------------ Simulation Core -------------------------
def _rng(seed: int):
    return np.random.RandomState(seed)

def simulate_step(rng, profile, sigma, dist):
    base_q, base_z = PROFILES[profile]["base"]
    wq, wz = PROFILES[profile]["w"]
    if dist == "uniform":
        qn = rng.uniform(-sigma, sigma)
        zn = rng.uniform(-sigma * 0.8, sigma * 0.8)
    else:
        qn = rng.normal(0, sigma)
        zn = rng.normal(0, sigma * 0.8)

    q = float(np.clip(base_q + wq * qn, 0.0, 0.99))
    z = float(np.clip(base_z + wz * zn, 0.0, 0.99))
    variance = abs(qn) + abs(zn)

    if variance > 0.15:
        status = "critical"
    elif variance > 0.07:
        status = "perturbed"
    else:
        status = "nominal"

    return {"σ": round(sigma, 6), "QΩ": q, "ζ_sync": z, "status": status}

# ------------------ Simulation Runner -----------------------
def run(profile, dist, sigma, seed, samples):
    rng = _rng(int(seed))
    results = [simulate_step(rng, profile, sigma, dist) for _ in range(samples)]
    q_mean = np.mean([r["QΩ"] for r in results])
    z_mean = np.mean([r["ζ_sync"] for r in results])
    majority = max(
        ["nominal", "perturbed", "critical"],
        key=lambda s: sum(1 for r in results if r["status"] == s),
    )

    summary = {
        "profile": profile,
        "noise_scale": sigma,
        "distribution": dist,
        "QΩ_mean": round(float(q_mean), 6),
        "ζ_sync_mean": round(float(z_mean), 6),
        "status_majority": majority,
        "timestamp_utc": datetime.utcnow().isoformat() + "Z",
        "rft_notice": LEGAL_NOTICE,
    }
    return summary, json.dumps(summary, indent=2)

# ------------------ File Saver -------------------------------
def save_run_log(run_json_str):
    try:
        data = json.loads(run_json_str)
        filename = f"RFT_Omega_Run_{datetime.utcnow().strftime('%Y-%m-%dT%H-%M-%SZ')}.json"
        with open(filename, "w") as f:
            json.dump(data, f, indent=2)
        return filename
    except Exception as e:
        return None

# ------------------ Gradio Interface ------------------------
with gr.Blocks(title="Rendered Frame Theory — Stabilising System Verification Panel") as demo:
    gr.Markdown(
        f"## 🧠 Rendered Frame Theory — Stabilising System Verification Panel  \n"
        f"**Version:** {RFT_VERSION}  \n"
        f"**DOI:** [{RFT_DOI}]({RFT_DOI})  \n"
        f"{LEGAL_NOTICE}"
    )

    gr.Markdown(
        """
        ### 🧩 How to Use  
        1️⃣ Select a **System Profile** (AI / Neural, SpaceX / Aerospace, Energy / RHES, Extreme Perturbation).  
        2️⃣ Choose a **Noise Distribution** (gauss or uniform).  
        3️⃣ Adjust **Noise Scale (σ)** to simulate environmental perturbations.  
        4️⃣ Press **Run Simulation** — results will show mean QΩ (stability) and ζ_sync (coherence) across samples.  
        5️⃣ Use **Save Run Log** to download your test record as a JSON file with timestamp.  

        **Interpretation:**  
        - `Nominal` → Stable harmonic equilibrium  
        - `Perturbed` → Transitional / adaptive state  
        - `Critical` → Instability threshold reached  

        Each output is time-stamped, reproducible, and protected under RFT-IPURL v1.0.  
        """
    )

    with gr.Row():
        profile = gr.Dropdown(list(PROFILES.keys()), label="System Profile", value="AI / Neural")
        dist = gr.Radio(["gauss", "uniform"], label="Noise Distribution", value="gauss")

    with gr.Row():
        sigma = gr.Slider(0.0, 0.3, value=0.05, step=0.01, label="Noise Scale (σ)")
        seed = gr.Number(value=123, label="Seed (integer)")
        samples = gr.Slider(1, 20, value=5, step=1, label="Samples per run")

    run_btn = gr.Button("Run Simulation")
    output = gr.JSON(label="Simulation Results")
    hidden_json = gr.Textbox(visible=False)
    save_btn = gr.Button("💾 Save Run Log")
    download_file = gr.File(label="Download Saved Log")

    run_btn.click(run, inputs=[profile, dist, sigma, seed, samples], outputs=[output, hidden_json])
    save_btn.click(save_run_log, inputs=[hidden_json], outputs=[download_file])

# ------------------ Launch -------------------------------
if __name__ == "__main__":
    demo.launch(server_name="0.0.0.0", share=False, ssr_mode=False)