#!/usr/bin/env bash
# Usage: verify_query_one.sh SPECIMEN
# SPECIMEN = "species--strain" (Make pattern stem)
# Output: stats/verify_query/SPECIMEN.stats (one CSV data row, no header)
set -euo pipefail

SPECIMEN="$1"
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
PYTHON="${SCRIPT_DIR}/../.venv/bin/python3"
VERIFY_PY="${SCRIPT_DIR}/verify_query.py"

species="${SPECIMEN%%--*}"
strain="${SPECIMEN#*--}"

DENSE="${SCRIPT_DIR}/query_dense/${SPECIMEN}.fasta.gz"
SPARSE="${SCRIPT_DIR}/query_sparse/${SPECIMEN}.fasta.gz"
STATS_DIR="${SCRIPT_DIR}/stats/verify_query"
STATS_FILE="${STATS_DIR}/${SPECIMEN}.stats"

mkdir -p "${STATS_DIR}"

echo "[${SPECIMEN}] verifying query (dense vs sparse)"

"${PYTHON}" "${VERIFY_PY}" \
    --species "${species}" \
    --strain  "${strain}" \
    "${DENSE}" "${SPARSE}" \
    >"${STATS_FILE}"