#!/usr/bin/env bash

set -e

LOGDIR="/tmp/typhoon/manual"
HEAD_LOGDIR="../head_reference/logs"
TESTS="../tests"

# Suite names default to the three curated suites (see CLAUDE.md's
# head_reference protocol). Override on the command line to run other
# suites, e.g. ./run_tests.sh ecm_ringers ecm_lmr_delta
SUITES=("$@")
if [ ${#SUITES[@]} -eq 0 ]; then
    SUITES=(ecm_ringers ecm_confident_quick ecm_hard_quick)
fi

# sd (fixed depth) is the right choice when judging whether a pruning/
# ordering change makes the tree smaller or larger for the same search
# effort; sn (fixed node budget) is the right choice when judging how many
# positions solve within a fixed cost. Override via SD/SN env vars --
# this applies to every suite passed on the command line.
#
# ecm_late.ep_ (see its header comment) is a different shape of suite --
# curated for positions where the *correct* move is only confirmed well
# into the search, specifically so avg-solution-time/total-nodecount from
# the stats block below can be read as an extensions/search-quality
# metric ("does the engine confirm these tactics sooner, in aggregate")
# rather than a solve-count gate. It's sn-only (no sd sweep): the
# discovery data behind it was measured on an 8-thread box (avg ~13 ply /
# ~70M total nodes across all 8 threads) and a fixed *depth* on 1 thread,
# with no cutoff-sharing across threads to shrink the tree, can blow up
# unpredictably on the sharpest of these positions -- a fixed *node*
# budget bounds wall-clock regardless. 15M nodes (~12s/position at this
# box's ~1.3Mnps single-thread rate, ~5 min for all 25 positions) is
# picked to stay in "quick ad hoc iteration" territory while still being
# in the right ballpark of the per-thread node share implied by the
# 8-thread discovery data -- not a rigorously derived number, adjust via
# SN if it's cutting off too many positions before they confirm.
SD="${SD:-10}"
SN="${SN:-5000000}"
SN_LABEL="${SN_LABEL:-5m}"
ECM_LATE_SN=15000000
ECM_LATE_SN_LABEL=15m

suite_sn_only() {
    [ "$1" = "ecm_late" ]
}
suite_sn() {
    if [ "$1" = "ecm_late" ]; then echo "$ECM_LATE_SN"; else echo "$SN"; fi
}
suite_sn_label() {
    if [ "$1" = "ecm_late" ]; then echo "$ECM_LATE_SN_LABEL"; else echo "$SN_LABEL"; fi
}

/bin/rm -rf "${LOGDIR}"
mkdir -p "${LOGDIR}"

echo -n "Launching test suites (${SUITES[*]}) in parallel... "
_start_time=$(date +%s)
for suite in "${SUITES[@]}"; do
   suite_sn=$(suite_sn "$suite")
   if ! suite_sn_only "$suite"; then
       ./typhoon \
           --cpus 1 \
           --hash 256m \
           --logfile "$LOGDIR/sd${SD}_${suite}.log" \
           --batch \
           --command "force; book name /nonexistent.book.bin; sd ${SD}; script $TESTS/${suite}.ep_" \
           > "$LOGDIR/sd${SD}_${suite}.out" 2>&1 &
   fi
   ./typhoon \
       --cpus 1 \
       --hash 256m \
       --logfile "$LOGDIR/sn${suite_sn}_${suite}.log" \
       --batch \
       --command "force; book name /nonexistent.book.bin; sn ${suite_sn}; script $TESTS/${suite}.ep_" \
       > "$LOGDIR/sn${suite_sn}_${suite}.out" 2>&1 &
done
wait
_elapsed=$(( $(date +%s) - _start_time ))
echo "DONE! (${_elapsed}s)"
echo

print_stats_table() {
    # $1 = head .out file, $2 = candidate .out file
    # Extracts the "Final statistics" block (correct solutions .. script time)
    # from each and prints a HEAD/CANDIDATE table with a numeric diff column.
    local headfile="$1" candfile="$2"
    awk -F: '
        NR==FNR {
            n++; lbl[n]=$1; val1[n]=$2; next
        }
        {
            m++; val2[m]=$2
        }
        END {
            printf("  %-22s %-35s %-35s\n", "", "HEAD", "CANDIDATE")
            for (i = 1; i <= n; i++) {
                label = lbl[i]; gsub(/^ +| +$/, "", label)
                v1 = val1[i]; gsub(/^ +| +$/, "", v1)
                v2 = (i <= m ? val2[i] : ""); gsub(/^ +| +$/, "", v2)
                if (label == "leftover") {
                    gsub(/^[0-9.]+% alpha, +/, "", v1)
                    gsub(/^[0-9.]+% alpha, +/, "", v2)
                }
                n1 = v1 + 0; n2 = v2 + 0
                diff = n2 - n1
                if (diff == 0) {
                    dstr = ""
                } else if (diff == int(diff)) {
                    dstr = sprintf(" (%+d)", diff)
                } else {
                    dstr = sprintf(" (%+.2f)", diff)
                }
                printf("  %-22s %-35s %-35s\n", label ":", v1, v2 dstr)
            }
        }
    ' "$headfile" "$candfile"
}

for suite in "${SUITES[@]}"; do
    suite_sn=$(suite_sn "$suite")
    suite_sn_label=$(suite_sn_label "$suite")
    if ! suite_sn_only "$suite"; then
        echo "----[ $suite sd=${SD} ]------------------------------------------------------------------"
        print_stats_table \
            <(sed -n '/correct solutions/,/script time/p' "$HEAD_LOGDIR/sd${SD}_${suite}.out") \
            <(sed -n '/correct solutions/,/script time/p' "$LOGDIR/sd${SD}_${suite}.out")
        echo
    fi
    echo "----[ $suite sn=${suite_sn} ]------------------------------------------------------------------"
    print_stats_table \
        <(sed -n '/correct solutions/,/script time/p' "$HEAD_LOGDIR/sn${suite_sn_label}_${suite}.out") \
        <(sed -n '/correct solutions/,/script time/p' "$LOGDIR/sn${suite_sn}_${suite}.out")
    echo
done
echo
echo "NOTE: The logfiles for these runs are in /tmp/typhoon/manual and will be"
echo "      overwritten by the next invocation of this script; save now them if"
echo "      you want them!"