#!/usr/bin/env bash
# batch-outcomes.sh — Summarize per-seed outcomes in a batch directory.
#
# Usage:
#   tools/batch-outcomes.sh <batch_dir>
#   tools/batch-outcomes.sh apricot:<remote_batch_dir>   # runs over ssh
#
# Output (one line per seed, tab-separated):
#   seedN  outcome  turn  p0_cities  p1_cities  wall_clock_sec
#
# The batch_dir must contain one or more `game_*/turn_stats.jsonl` files.
# Works for smoke/clan single-mode batches. For gpu-walltime and other
# multi-mode batches, run once per mode subdir:
#   tools/batch-outcomes.sh apricot:~/.cache/mc-batches/20260418_080214/gpu-true

set -euo pipefail

TARGET="${1:?usage: tools/batch-outcomes.sh <batch_dir | apricot:/path>}"

# The query is identical local vs remote — just the execution host differs.
# Packing it as a here-doc keeps escaping sane.
read -r -d '' QUERY <<'EOF' || true
set -e
: "${DIR:?DIR must be set}"
for d in "$DIR"/game_*; do
    [ -d "$d" ] || continue
    seed=$(basename "$d" | grep -oE 'seed[0-9]+')
    stats="$d/turn_stats.jsonl"
    [ -s "$stats" ] || { printf '%s\tNO-STATS\t-\t-\t-\t-\n' "$seed"; continue; }
    last=$(tail -1 "$stats")
    python3 - "$last" "$seed" <<'PY'
import json, sys
line, seed = sys.argv[1], sys.argv[2]
try:
    d = json.loads(line)
except Exception as e:
    print(f"{seed}\tPARSE-ERR\t-\t-\t-\t-")
    sys.exit(0)
outcome = d.get("outcome", "?")
turn    = d.get("turn", "?")
wc      = d.get("wall_clock_sec", "?")
ps      = d.get("player_stats", {})
p0c     = ps.get("0", {}).get("cities", "-")
p1c     = ps.get("1", {}).get("cities", "-")
wcs = f"{wc:.1f}" if isinstance(wc, (int, float)) else str(wc)
print(f"{seed}\t{outcome}\t{turn}\t{p0c}\t{p1c}\t{wcs}")
PY
done | sort -V
EOF

if [[ "$TARGET" == apricot:* ]]; then
    REMOTE_PATH="${TARGET#apricot:}"
    ssh apricot "DIR='${REMOTE_PATH}' bash -s" <<< "$QUERY"
else
    DIR="$TARGET" bash -c "$QUERY"
fi