VoxelForge/Standalone/Research/WorldAssemblerM3/Linux103/RunL103Readiness.py

#!/usr/bin/env python3
"""L80-protocol near and whole-view readiness at W8/W12 for all owner strata."""
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/l103/readiness"
DATA = ROOT / "Standalone/data/owner_world.json"
BINARY = ROOT / "Saved/Linux/l101/bin/VoxelForgeCoreHeadless-l101-clang"
STRATA = {
    "SW": {"camera": (8.0, 16.0, -372.2626953125), "near": 1408, "whole": 2850},
    "FI": {"camera": (25.8369140625, 3.19921875, -737.0), "near": 1216, "whole": 2249},
    "TN": {"camera": (19.166015625, -14.2001953125, -753.3740234375), "near": 1344, "whole": 2233},
    "VS": {"camera": (-20.0625, -5.068359375, -900.353515625), "near": 1344, "whole": 2329},
}
NEAR = re.compile(r"^\[WorldAssemblerNearLODReady\] lods=0-2 tiles=(\d+)/(\d+) cold_ready_s=([0-9.]+)$", re.M)
PUBLISHED = re.compile(r"^\[WorldAssemblerNearLODPublished\] lods=0-2 tiles=(\d+)/(\d+) cold_ready_s=([0-9.]+)$", re.M)
PIPELINE_READY = re.compile(r"^\[WorldAssemblerNearLODPipelineReady\] lods=0-2 tiles=(\d+)/(\d+) cold_ready_s=([0-9.]+)$", re.M)
PIPELINE = re.compile(r"^\[StreamPipeline\] (.*)$", re.M)
CONFIG = re.compile(r"^\[HeadlessConfig\] (.*)$", re.M)
STAMP = re.compile(r"^\[VoxelForgeSourceStamp\] sha256=(\S+)$", re.M)
FIELDS = re.compile(r"([a-zA-Z0-9_]+)=([^ ]+)")


def run(stratum: str, workers: int, phase: str, replicate: int) -> dict:
    OUT.mkdir(parents=True, exist_ok=True)
    suffix = f"-r{replicate}" if phase == "measured" else ""
    log = OUT / f"{stratum.lower()}-w{workers}-{phase}{suffix}.log"
    camera = ",".join(f"{value:.12f}" for value in STRATA[stratum]["camera"]) + ",0,0"
    command = [
        "taskset", "-c", "0-11", "prlimit", "--as=6442450944", "--", "timeout", "1740s",
        str(BINARY), str(DATA), "--world-assembler-prototype", "--stream-benchmark",
        "--stream-pipeline", f"--workers={workers}", "--clip-radius=3",
        "--max-clip-level=8", "--max-level=8", f"--camera={camera}",
        "--profile-lod-work", "--stream-scheduler-policy=3",
    ]
    started = time.monotonic()
    with log.open("w", encoding="utf-8") as stream:
        result = subprocess.run(command, cwd=ROOT, stdout=stream, stderr=subprocess.STDOUT,
                                check=False, timeout=1740)
    content = log.read_text(encoding="utf-8", errors="replace")
    near, published, pipeline_ready = (NEAR.search(content), PUBLISHED.search(content),
                                       PIPELINE_READY.search(content))
    pipeline, config, stamp = PIPELINE.search(content), CONFIG.search(content), STAMP.search(content)
    if result.returncode or not all((near, published, pipeline_ready, pipeline, config, stamp)):
        raise RuntimeError(f"readiness run failed or markers missing: {log}; exit={result.returncode}")
    fields = dict(FIELDS.findall(pipeline.group(1)))
    config_fields = dict(FIELDS.findall(config.group(1)))
    spec = STRATA[stratum]
    if int(config_fields["workers"]) != workers:
        raise RuntimeError(f"worker mismatch: {log}")
    if stamp.group(1) != "c5f49a45f51856b857a1ee8877b7d777208e67daf64281caefc7ef84e8006410":
        raise RuntimeError(f"source stamp mismatch: {log}: {stamp.group(1)}")
    for marker in (near, published, pipeline_ready):
        if (int(marker.group(1)), int(marker.group(2))) != (spec["near"], spec["near"]):
            raise RuntimeError(f"incomplete near set: {log}: {marker.group(1)}/{marker.group(2)}")
    whole_done, whole_total = (int(value) for value in fields["whole_tiles"].split("/"))
    if (whole_done, whole_total) != (spec["whole"], spec["whole"]):
        raise RuntimeError(f"incomplete whole set: {log}: {whole_done}/{whole_total}")
    row = {
        "stratum": stratum, "workers": workers, "phase": phase, "replicate": replicate,
        "exit_code": result.returncode, "source_stamp": stamp.group(1),
        "near_tiles": spec["near"], "whole_tiles": spec["whole"],
        "lod02_ready_s": float(near.group(3)),
        "lod02_published_s": float(published.group(3)),
        "lod02_pipeline_ready_s": float(pipeline_ready.group(3)),
        "whole_view_ready_s": float(fields["whole_view_ready"].removesuffix("s")),
        "shape_s": float(fields["shape"].removesuffix("s")),
        "far_shape_s": float(fields["far_shape"].removesuffix("s")),
        "wall_s": round(time.monotonic() - started, 3),
        "log": str(log.relative_to(ROOT)),
    }
    print("[L103ReadinessRun] " + json.dumps(row, sort_keys=True), flush=True)
    return row


def main() -> int:
    if not DATA.is_file() or not BINARY.is_file():
        raise SystemExit(f"missing input: {DATA} or {BINARY}")
    for stratum in ("VS", "TN", "FI", "SW"):
        for workers in (12, 8):
            run(stratum, workers, "warmup", 0)
    schedules = (
        (("SW", 8), ("SW", 12), ("FI", 8), ("FI", 12), ("TN", 8), ("TN", 12), ("VS", 8), ("VS", 12)),
        (("VS", 12), ("VS", 8), ("TN", 12), ("TN", 8), ("FI", 12), ("FI", 8), ("SW", 12), ("SW", 8)),
        (("FI", 8), ("FI", 12), ("SW", 8), ("SW", 12), ("VS", 8), ("VS", 12), ("TN", 8), ("TN", 12)),
    )
    rows = []
    for replicate, schedule in enumerate(schedules, 1):
        for stratum, workers in schedule:
            rows.append(run(stratum, workers, "measured", replicate))
            (OUT / "runs-partial.json").write_text(json.dumps(rows, indent=2) + "\n",
                                                   encoding="utf-8")
    summary = []
    for stratum in STRATA:
        for workers in (8, 12):
            group = [row for row in rows if row["stratum"] == stratum and row["workers"] == workers]
            if len(group) != 3:
                raise RuntimeError(f"expected three measured runs: {stratum} W{workers}")
            summary.append({
                "stratum": stratum, "workers": workers, "measured_runs": 3,
                **{metric: [row[metric] for row in group]
                   for metric in ("lod02_ready_s", "lod02_published_s",
                                  "lod02_pipeline_ready_s", "whole_view_ready_s")},
                **{f"{metric}_median": statistics.median(row[metric] for row in group)
                   for metric in ("lod02_ready_s", "lod02_published_s",
                                  "lod02_pipeline_ready_s", "whole_view_ready_s")},
            })
    (OUT / "summary.json").write_text(json.dumps({"runs": rows, "summary": summary}, indent=2)
                                        + "\n", encoding="utf-8")
    return 0


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