"""Re-run ONLY PaddleOCR layout detection (no Claude calls) on a saved run's
page PNG, at the default 0.5 cutoff and at a lowered threshold, and write the
all-regions debug image for each so the effect of the threshold is visible.

Usage: venv/bin/python3 _replay_regions.py <run_dir> <page_num> [threshold]
       (threshold defaults to 0.3)
"""
import sys
from collections import Counter
from pathlib import Path

from smart_extractor import detect_regions, draw_all_regions_debug


def summarize(regions):
    c = Counter(r.get("type", "unknown") for r in regions)
    return ", ".join(f"{k}={v}" for k, v in sorted(c.items(), key=lambda kv: -kv[1]))


def main(run_dir, page_num, threshold=0.3):
    run = Path(run_dir)
    png = run / "pages" / f"page_{page_num:03d}.png"
    out = Path(f"/tmp/replay_regions_{run.name}_p{page_num:03d}")
    out.mkdir(parents=True, exist_ok=True)

    print(f"=== page {png}  ->  {out} ===\n")

    # BASELINE: stock 0.5 cutoff
    print("--- detection @ default (0.5) ---")
    base = detect_regions(str(png), threshold=0.5)
    print(f"  {len(base)} regions: {summarize(base)}")
    draw_all_regions_debug(str(png), base, out, page_num)
    Path(out / f"page_{page_num:03d}_allregions_debug.png").rename(
        out / f"page_{page_num:03d}_thr050.png")

    # LOWERED: threshold under test
    print(f"\n--- detection @ {threshold} ---")
    low = detect_regions(str(png), threshold=threshold)
    print(f"  {len(low)} regions: {summarize(low)}")
    draw_all_regions_debug(str(png), low, out, page_num)
    Path(out / f"page_{page_num:03d}_allregions_debug.png").rename(
        out / f"page_{page_num:03d}_thr{int(threshold*100):03d}.png")

    # DELTA: new regions gained by lowering the cutoff
    print(f"\n--- delta: {len(low) - len(base)} more region(s) at {threshold} ---")
    base_boxes = {tuple(r["bbox"]) for r in base}
    gained = [r for r in low if tuple(r["bbox"]) not in base_boxes]
    gained.sort(key=lambda r: (r["bbox"][1], r["bbox"][0]))
    for r in gained:
        b = r["bbox"]
        print(f"  + R{r['id']:>3} {r.get('type','?'):>16}  conf={r.get('confidence',0):.2f}  "
              f"bbox={b}  ({b[2]-b[0]}x{b[3]-b[1]})")

    print(f"\nImages written to: {out}")
    for f in sorted(out.glob("*.png")):
        print(f"  {f.name}")


if __name__ == "__main__":
    thr = float(sys.argv[3]) if len(sys.argv) > 3 else 0.3
    main(sys.argv[1], int(sys.argv[2]), thr)