#!/usr/bin/env python3
"""Median across repetitions, one row per configuration. usage: summarize.py output/02-runs.txt"""
import re
import statistics
import sys
from collections import OrderedDict

rows = OrderedDict()
for line in open(sys.argv[1]):
    if "FAILED" in line or not line.strip():
        continue
    kv = dict(re.findall(r"(\w+)=([^\s]+)", line))
    rows.setdefault(kv["label"], []).append({k: float(v) for k, v in kv.items() if k not in ("label",)})

cols = [
    ("p50 ms", "p50_ms", "{:.2f}"),
    ("p99 ms", "p99_ms", "{:.1f}"),
    ("p99.9 ms", "p999_ms", "{:.1f}"),
    ("max ms", "max_ms", "{:.1f}"),
    ("peak req/s", "peak_rps", "{:.0f}"),
    ("pauses", "pauses", "{:.0f}"),
    ("pause p99 ms", "pause_p99_ms", "{:.3f}"),
    ("pause max ms", "pause_max_ms", "{:.3f}"),
    ("pause total ms", "pause_total_ms", "{:.0f}"),
    ("RSS MB", "rss_peak_mb", "{:.0f}"),
]
reps = max(len(v) for v in rows.values())
print(f"median of {reps} runs per configuration (latency from a fixed-rate open loop; throughput from a closed loop)")
print()
print(f"{'configuration':<26}" + "".join(f"{c[0]:>15}" for c in cols))
for label, runs in rows.items():
    cells = [c[2].format(statistics.median(r[c[1]] for r in runs)) for c in cols]
    print(f"{label:<26}" + "".join(f"{x:>15}" for x in cells))