make_single_notebook.py

"""
make_single_notebook.py
Converts HSRIS_COMPLETE.py → HSRIS_FINAL.ipynb
Each  # ── CELL N :  comment starts a new notebook cell.
"""
import json, re
from pathlib import Path

src  = Path("HSRIS_COMPLETE.py").read_text(encoding="utf-8")
lines = src.splitlines()

CELL_RE = re.compile(r"^# ── CELL \d+")

def make_code_cell(code_lines):
    src = [l + "\n" for l in code_lines]
    return {"cell_type": "code", "execution_count": None,
            "metadata": {}, "outputs": [], "source": src}

def make_md_cell(text):
    return {"cell_type": "markdown", "metadata": {},
            "source": [l + "\n" for l in text.splitlines()]}

# ── split into cells ──────────────────────────────────────────
cells   = []
buf     = []

def flush(buf):
    code = "\n".join(buf).strip()
    if code:
        cells.append(make_code_cell(buf))

for line in lines:
    if CELL_RE.match(line):
        flush(buf)
        buf = [line]
    elif re.match(r"^# ={30,}", line):
        # section header → markdown
        flush(buf)
        buf = []
    else:
        buf.append(line)
flush(buf)

# ── title markdown cell ───────────────────────────────────────
title_cell = make_md_cell("""# 🚀 Hybrid Semantic Retrieval & Intelligence System (HSRIS)
**Platform:** Kaggle | **GPU:** Tesla T4 ×2 | **Framework:** PyTorch + NumPy

| Stage | Method | Details |
|---|---|---|
| Encoders | Label + One-Hot | Priority → int, Channel → binary vector |
| Sparse Retrieval | BoW + N-Grams + TF-IDF | Sparse COO tensors |
| Dense Retrieval | GloVe 300-d | TF-IDF-weighted sentence embeddings |
| Hybrid Search | α·TF-IDF + (1-α)·GloVe | α=0.4 default |
| Evaluation | Precision@5 | 50 queries, GloVe-wins analysis |
| UI | Gradio | Query box + α slider + Top-3 results |

> **Run order:** Enable GPU → Run All → Gradio link appears at the bottom.""")

cells.insert(0, title_cell)

# ── assemble notebook ─────────────────────────────────────────
nb = {
    "nbformat": 4, "nbformat_minor": 5,
    "metadata": {
        "kernelspec": {"display_name": "Python 3",
                       "language": "python", "name": "python3"},
        "language_info": {"name": "python", "version": "3.10.0"},
        "accelerator": "GPU",
    },
    "cells": cells,
}

out = Path("HSRIS_FINAL.ipynb")
out.write_text(json.dumps(nb, indent=1, ensure_ascii=False), encoding="utf-8")
size_kb = out.stat().st_size // 1024
print(f"✅  HSRIS_FINAL.ipynb  |  {len(cells)} cells  |  {size_kb} KB")