VoxelForge/Standalone/Research/WorldAssemblerM3/Linux78/RunL78TimingPairs.py

#!/usr/bin/env python3
"""Run interleaved L77/L78 cold readiness pairs for the four owner strates."""
from __future__ import annotations

import json
import re
import statistics
import subprocess
import time
from pathlib import Path


ROOT = Path(__file__).resolve().parents[4]
OUT = ROOT / "Saved/Linux/l78/timing"
DATA = ROOT / "Standalone/data/owner_world.json"
BINARIES = {
    "before": ROOT / "Standalone/bin/linux/VoxelForgeCoreHeadless-l77-clang",
    "after": ROOT / "Saved/Linux/l78/bin/VoxelForgeCoreHeadless-l78-clang-profile",
}
CAMERAS = {
    0: (8.0, 16.0, -372.2626953125),
    1: (25.8369140625, 3.19921875, -737.0),
    2: (19.166015625, -14.2001953125, -753.3740234375),
    3: (-20.0625, -5.068359375, -900.353515625),
}
PIPELINE = re.compile(r"^\[StreamPipeline\] (.*)$", re.M)
FIELDS = re.compile(r"([a-zA-Z0-9_]+)=([^ ]+)")
LOD_READY = re.compile(
    r"^\[WorldAssemblerL77LODReady\] lod=(\d+) tiles=(\d+) cold_ready_s=([0-9.]+)$",
    re.M,
)
STAMP = re.compile(r"^\[VoxelForgeSourceStamp\] sha256=(\S+)$", re.M)
ORDERS = (("before", "after"), ("after", "before"), ("before", "after"))


def run(strate: int, pair: int, mode: str) -> dict:
    OUT.mkdir(parents=True, exist_ok=True)
    log = OUT / f"strate{strate}-pair{pair}-{mode}.log"
    camera = ",".join(f"{value:.12f}" for value in CAMERAS[strate]) + ",0,0"
    command = [
        "taskset", "-c", "0-11", "prlimit", "--as=6442450944", "--",
        "timeout", "1740s", str(BINARIES[mode]), str(DATA),
        "--world-assembler-prototype", "--stream-benchmark", "--stream-pipeline",
        "--workers=8", "--clip-radius=3", "--max-clip-level=8", "--max-level=8",
        f"--camera={camera}", "--profile-lod-work", "--linux77-profile",
    ]
    started = time.monotonic()
    with log.open("w", encoding="utf-8") as output:
        result = subprocess.run(command, cwd=ROOT, stdout=output,
                                stderr=subprocess.STDOUT, check=False, timeout=1740)
    content = log.read_text(encoding="utf-8", errors="replace")
    pipeline_match = PIPELINE.search(content)
    ready = {int(lod): {"tiles": int(tiles), "seconds": float(seconds)}
             for lod, tiles, seconds in LOD_READY.findall(content)}
    if result.returncode != 0 or pipeline_match is None or any(lod not in ready for lod in (0, 1, 2)):
        raise SystemExit(f"readiness run failed or LOD0-2 data missing: {log}; exit={result.returncode}")
    pipeline = dict(FIELDS.findall(pipeline_match.group(1)))
    row = {
        "strate": strate,
        "pair": pair,
        "mode": mode,
        "exit_code": result.returncode,
        "wall_s": round(time.monotonic() - started, 3),
        "lod02_ready_s": max(ready[lod]["seconds"] for lod in (0, 1, 2)),
        "lod_ready_s": {str(lod): ready[lod]["seconds"] for lod in sorted(ready)},
        "whole_view_ready_s": float(pipeline["whole_view_ready"].removesuffix("s")),
        "whole_tiles": int(pipeline["whole_tiles"].split("/")[0]),
        "sampler": next((line for line in content.splitlines()
                          if line.startswith("[L77Sampler] action=stop")), None),
        "source_stamp": STAMP.search(content).group(1) if STAMP.search(content) else None,
        "log": str(log.relative_to(ROOT)),
    }
    print("[L78TimingRun] " + json.dumps(row, sort_keys=True), flush=True)
    return row


def main() -> int:
    missing = [str(path) for path in (DATA, *BINARIES.values()) if not path.is_file()]
    if missing:
        raise SystemExit("missing input: " + ", ".join(missing))
    rows = []
    for strate in range(4):
        for pair, order in enumerate(ORDERS, 1):
            paired = {}
            for mode in order:
                row = run(strate, pair, mode)
                rows.append(row)
                paired[mode] = row
                (OUT / "runs-partial.json").write_text(
                    json.dumps(rows, indent=2) + "\n", encoding="utf-8")
            summary_row = {
                "strate": strate,
                "pair": pair,
                "before_lod02_s": paired["before"]["lod02_ready_s"],
                "after_lod02_s": paired["after"]["lod02_ready_s"],
                "before_whole_s": paired["before"]["whole_view_ready_s"],
                "after_whole_s": paired["after"]["whole_view_ready_s"],
            }
            print("[L78TimingPair] " + json.dumps(summary_row, sort_keys=True), flush=True)
    summary = []
    for strate in range(4):
        group = [row for row in rows if row["strate"] == strate]
        before_near = [row["lod02_ready_s"] for row in group if row["mode"] == "before"]
        after_near = [row["lod02_ready_s"] for row in group if row["mode"] == "after"]
        before_whole = [row["whole_view_ready_s"] for row in group if row["mode"] == "before"]
        after_whole = [row["whole_view_ready_s"] for row in group if row["mode"] == "after"]
        summary.append({
            "strate": strate,
            "pairs": len(before_near),
            "before_lod02_s": before_near,
            "after_lod02_s": after_near,
            "before_lod02_median_s": statistics.median(before_near),
            "after_lod02_median_s": statistics.median(after_near),
            "before_whole_s": before_whole,
            "after_whole_s": after_whole,
            "before_whole_median_s": statistics.median(before_whole),
            "after_whole_median_s": statistics.median(after_whole),
        })
    result = {"runs": rows, "summary": summary}
    (OUT / "summary.json").write_text(json.dumps(result, indent=2) + "\n",
                                      encoding="utf-8")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())