#!/usr/bin/env bash
set -euo pipefail

usage() {
  cat >&2 <<'EOF'
Usage:
  REAL_RASTER_PATH=/path/to/orthophoto.tif \
  REAL_REFERENCE_VECTOR_PATH=/path/to/reference-buildings.geojson \
  QUALITY_MODEL_ASSET_IDS="yolov8n-building-segmentation-pt yolov8n-pt" \
  QUALITY_TILE_SIZES="512 640" \
  QUALITY_TILE_OVERLAPS="64" \
  QUALITY_THRESHOLDS="0.50 0.25 0.15" \
  QUALITY_SAMPLE_SLUG="mol" \
  bash scripts/run_detection_quality_matrix.sh [base_url]

or:
  bash scripts/run_detection_quality_matrix.sh [base_url] /path/to/orthophoto.tif /path/to/reference-buildings.geojson

Required inputs:
  REAL_RASTER_PATH             Georeferenced .tif/.tiff/.geotiff raster.
  REAL_REFERENCE_VECTOR_PATH   EPSG-aware .geojson/.json reference vector with building polygons.

Optional environment:
  QUALITY_MODEL_ASSET_IDS      Space/comma separated local model asset IDs. Default: active configured model asset.
  QUALITY_TILE_SIZES           Space/comma separated raster tile sizes, default: 640.
  QUALITY_TILE_OVERLAPS        Space/comma separated raster tile overlaps, default: 64.
  QUALITY_THRESHOLDS           Space/comma separated confidence thresholds, default: 0.50 0.25 0.15.
  QUALITY_OUTPUT_DIR           Output directory, default: artifacts/detection-quality-matrix/<timestamp>.
  QUALITY_SAMPLE_SLUG          Optional AOI slug included in persisted project names.
  REAL_PROJECT_REGION          Persisted project region forwarded to the workflow.
  REAL_AREA_NAME               Persisted AOI name when REAL_AREA_BBOX is set.
  REAL_AREA_BBOX               Optional EPSG:4326 minx,miny,maxx,maxy AOI bounds.
  REAL_IOU_THRESHOLD           QA IoU threshold, default inherited by the underlying workflow.

This script never downloads models and never uses fixture detections. It repeats
the existing real-data upload/tile/detection/QA workflow for each matrix row.
EOF
}

ROOT="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
cd "$ROOT"

BASE_URL="${1:-${GE_INTEL_BASE_URL:-http://localhost:1202}}"
REAL_RASTER_PATH="${2:-${REAL_RASTER_PATH:-}}"
REAL_REFERENCE_VECTOR_PATH="${3:-${REAL_REFERENCE_VECTOR_PATH:-}}"
QUALITY_MODEL_ASSET_IDS="${QUALITY_MODEL_ASSET_IDS:-${REAL_MODEL_ASSET_ID:-__active__}}"
QUALITY_TILE_SIZES="${QUALITY_TILE_SIZES:-640}"
QUALITY_TILE_OVERLAPS="${QUALITY_TILE_OVERLAPS:-64}"
QUALITY_THRESHOLDS="${QUALITY_THRESHOLDS:-0.50 0.25 0.15}"
QUALITY_SAMPLE_SLUG="${QUALITY_SAMPLE_SLUG:-}"
REAL_PROJECT_REGION="${REAL_PROJECT_REGION:-Kempen}"
REAL_AREA_NAME="${REAL_AREA_NAME:-${QUALITY_SAMPLE_SLUG:-Detection quality} AOI}"
REAL_AREA_BBOX="${REAL_AREA_BBOX:-}"
QUALITY_OUTPUT_DIR="${QUALITY_OUTPUT_DIR:-artifacts/detection-quality-matrix/$(date -u +%Y%m%dT%H%M%SZ)}"

if [ "${BASE_URL}" = "-h" ] || [ "${BASE_URL}" = "--help" ]; then
  usage
  exit 0
fi

if [ -z "${REAL_RASTER_PATH}" ] || [ -z "${REAL_REFERENCE_VECTOR_PATH}" ]; then
  usage
  exit 2
fi

if ! command -v curl >/dev/null 2>&1; then
  echo "curl is required for detection quality matrix verification" >&2
  exit 1
fi

if [ -n "${PYTHON_BIN:-}" ]; then
  PYTHON_BIN="${PYTHON_BIN}"
else
  PYTHON_BIN=""
  for candidate in python3 python.exe python; do
    if command -v "${candidate}" >/dev/null 2>&1 && "${candidate}" -c "import json, sys" >/dev/null 2>&1; then
      PYTHON_BIN="${candidate}"
      break
    fi
  done
fi

if [ -z "${PYTHON_BIN}" ]; then
  echo "A Python interpreter is required for JSON parsing" >&2
  exit 1
fi

mkdir -p "${QUALITY_OUTPUT_DIR}"

model_requests_normalized="$(printf '%s' "${QUALITY_MODEL_ASSET_IDS}" | tr ',' ' ')"
tile_sizes_normalized="$(printf '%s' "${QUALITY_TILE_SIZES}" | tr ',' ' ')"
tile_overlaps_normalized="$(printf '%s' "${QUALITY_TILE_OVERLAPS}" | tr ',' ' ')"
thresholds_normalized="$(printf '%s' "${QUALITY_THRESHOLDS}" | tr ',' ' ')"
matrix_manifest="${QUALITY_OUTPUT_DIR}/quality_matrix_requests.tsv"

"${PYTHON_BIN}" - \
  "${model_requests_normalized}" \
  "${tile_sizes_normalized}" \
  "${tile_overlaps_normalized}" \
  "${thresholds_normalized}" \
  "${matrix_manifest}" <<'PY'
import re
import sys

models, tile_sizes, tile_overlaps, thresholds, output_path = sys.argv[1:6]


def split_values(raw: str, name: str) -> list[str]:
    values = [value.strip() for value in raw.split() if value.strip()]
    if not values:
        raise SystemExit(f"{name} must contain at least one value")
    return values


def safe_label(value: str) -> str:
    label = re.sub(r"[^A-Za-z0-9_.-]+", "_", value.strip())
    return label.replace(".", "p").replace("-", "_") or "active"


model_values = split_values(models, "QUALITY_MODEL_ASSET_IDS")
tile_size_values = split_values(tile_sizes, "QUALITY_TILE_SIZES")
tile_overlap_values = split_values(tile_overlaps, "QUALITY_TILE_OVERLAPS")
threshold_values = split_values(thresholds, "QUALITY_THRESHOLDS")

for value in tile_size_values:
    try:
        tile_size = int(value)
    except ValueError as exc:
        raise SystemExit(f"Invalid tile size: {value}") from exc
    if tile_size < 64:
        raise SystemExit(f"Tile size must be at least 64: {value}")

for value in tile_overlap_values:
    try:
        tile_overlap = int(value)
    except ValueError as exc:
        raise SystemExit(f"Invalid tile overlap: {value}") from exc
    if tile_overlap < 0:
        raise SystemExit(f"Tile overlap cannot be negative: {value}")

for value in threshold_values:
    try:
        threshold = float(value)
    except ValueError as exc:
        raise SystemExit(f"Invalid confidence threshold: {value}") from exc
    if threshold < 0.0 or threshold > 1.0:
        raise SystemExit(f"Confidence threshold must be between 0 and 1: {value}")

with open(output_path, "w", encoding="utf-8") as handle:
    for model in model_values:
        for tile_size in tile_size_values:
            for overlap in tile_overlap_values:
                for threshold in threshold_values:
                    label = "_".join(
                        [
                            f"model_{safe_label(model)}",
                            f"tile_{safe_label(tile_size)}",
                            f"overlap_{safe_label(overlap)}",
                            f"threshold_{safe_label(threshold)}",
                        ]
                    )
                    handle.write(f"{model}\t{tile_size}\t{overlap}\t{threshold}\t{label}\n")
PY

echo "== GeoIntel detection quality matrix =="
echo "Base URL: ${BASE_URL}"
echo "Raster: ${REAL_RASTER_PATH}"
echo "Reference vector: ${REAL_REFERENCE_VECTOR_PATH}"
echo "Models: ${model_requests_normalized}"
echo "Tile sizes: ${tile_sizes_normalized}"
echo "Tile overlaps: ${tile_overlaps_normalized}"
echo "Thresholds: ${thresholds_normalized}"
echo "Output: ${QUALITY_OUTPUT_DIR}"

run_index=0
while IFS=$'\t' read -r model_request tile_size tile_overlap threshold run_label; do
  run_index=$((run_index + 1))
  run_log="${QUALITY_OUTPUT_DIR}/${run_label}.log"
  quality_response="${QUALITY_OUTPUT_DIR}/${run_label}_quality_checks.json"
  detection_run_response="${QUALITY_OUTPUT_DIR}/${run_label}_detection_run.json"
  run_summary="${QUALITY_OUTPUT_DIR}/${run_label}_summary.json"
  model_env="${model_request}"
  if [ "${model_request}" = "__active__" ]; then
    model_env=""
  fi
  project_sample_label=""
  if [ -n "${QUALITY_SAMPLE_SLUG}" ]; then
    project_sample_label=" sample ${QUALITY_SAMPLE_SLUG}"
  fi

  echo "-- Matrix run ${run_index}: model=${model_request} tile=${tile_size} overlap=${tile_overlap} threshold=${threshold} --"
  if ! REAL_PROJECT_NAME="GeoIntel Detection Quality Matrix${project_sample_label} ${model_request} tile ${tile_size} overlap ${tile_overlap} threshold ${threshold}" \
      REAL_PROJECT_REGION="${REAL_PROJECT_REGION}" \
      REAL_AREA_NAME="${REAL_AREA_NAME}" \
      REAL_AREA_BBOX="${REAL_AREA_BBOX}" \
      REAL_MODEL_ASSET_ID="${model_env}" \
      REAL_TILE_SIZE="${tile_size}" \
      REAL_TILE_OVERLAP="${tile_overlap}" \
      REAL_CONFIDENCE_THRESHOLD="${threshold}" \
      bash scripts/verify_real_data_detection_qa_workflow.sh "${BASE_URL}" "${REAL_RASTER_PATH}" "${REAL_REFERENCE_VECTOR_PATH}" \
      >"${run_log}" 2>&1; then
    echo "Detection quality matrix run failed. Log: ${run_log}" >&2
    tail -n 80 "${run_log}" >&2 || true
    exit 1
  fi

  project_id="$(sed -n 's/^Project: //p' "${run_log}" | tail -n 1)"
  area_id="$(sed -n 's/^Area: //p' "${run_log}" | tail -n 1)"
  raster_dataset_id="$(sed -n 's/^Raster dataset: //p' "${run_log}" | tail -n 1)"
  reference_dataset_id="$(sed -n 's/^Reference dataset: //p' "${run_log}" | tail -n 1)"
  selected_model_asset_id="$(sed -n 's/^Model asset: //p' "${run_log}" | tail -n 1)"
  manifest_path="$(sed -n 's/^Manifest: //p' "${run_log}" | tail -n 1)"
  analysis_run_id="$(sed -n 's/^Analysis run: //p' "${run_log}" | tail -n 1)"
  quality_check_id="$(sed -n 's/^Quality check: //p' "${run_log}" | tail -n 1)"
  detection_count="$(sed -n 's/^Detections: //p' "${run_log}" | tail -n 1)"
  export_id="$(sed -n 's/^Detection export: //p' "${run_log}" | tail -n 1)"

  if [ -z "${project_id}" ] || [ -z "${analysis_run_id}" ] || [ -z "${quality_check_id}" ]; then
    echo "Matrix run did not print project/run/quality ids. Log: ${run_log}" >&2
    exit 1
  fi

  curl -fsS "${BASE_URL%/}/api/v1/projects/${project_id}/quality-checks?limit=200" > "${quality_response}"
  curl -fsS "${BASE_URL%/}/api/v1/detection/runs/${analysis_run_id}" > "${detection_run_response}"

  "${PYTHON_BIN}" - \
    "${quality_response}" \
    "${detection_run_response}" \
    "${run_summary}" \
    "${model_request}" \
    "${selected_model_asset_id}" \
    "${tile_size}" \
    "${tile_overlap}" \
    "${threshold}" \
    "${project_id}" \
    "${area_id}" \
    "${raster_dataset_id}" \
    "${reference_dataset_id}" \
    "${manifest_path}" \
    "${analysis_run_id}" \
    "${quality_check_id}" \
    "${detection_count}" \
    "${export_id}" \
    "${run_log}" <<'PY'
import json
import sys

(
    quality_path,
    detection_run_path,
    output_path,
    model_request,
    selected_model_asset_id,
    tile_size,
    tile_overlap,
    threshold,
    project_id,
    area_id,
    raster_dataset_id,
    reference_dataset_id,
    manifest_path,
    analysis_run_id,
    quality_check_id,
    detection_count,
    export_id,
    run_log,
) = sys.argv[1:19]

with open(quality_path, "r", encoding="utf-8") as handle:
    payload = json.load(handle)
if "data" not in payload:
    raise SystemExit("Quality-check list is not a canonical GeoIntel data envelope")
with open(detection_run_path, "r", encoding="utf-8") as handle:
    detection_run_payload = json.load(handle)
if "data" not in detection_run_payload:
    raise SystemExit("Detection run detail is not a canonical GeoIntel data envelope")
detection_run = detection_run_payload["data"]
run_result = detection_run.get("result_json") or {}
items = payload["data"].get("items") or []
quality_check = next((item for item in items if str(item.get("id")) == quality_check_id), None)
if quality_check is None:
    raise SystemExit(f"Quality check not found in project list: {quality_check_id}")
metrics = {
    metric.get("metric_key"): metric.get("metric_value")
    for metric in quality_check.get("metrics", [])
    if metric.get("metric_key")
}
findings = quality_check.get("findings_json") or {}
coverage = findings.get("coverage") or {}
diagnostics = findings.get("box_to_footprint_diagnostics") or {}
reference_raw_count = coverage.get("reference_raw_count")
reference_evaluated_count = coverage.get("reference_evaluated_count")
reference_coverage_ratio = None
if isinstance(reference_raw_count, int) and reference_raw_count > 0 and isinstance(reference_evaluated_count, int):
    reference_coverage_ratio = reference_evaluated_count / reference_raw_count
summary = {
    "model_request": model_request,
    "model_asset_id": selected_model_asset_id,
    "tile_size": int(tile_size),
    "tile_overlap": int(tile_overlap),
    "threshold": float(threshold),
    "project_id": project_id,
    "area_id": area_id or None,
    "raster_dataset_id": raster_dataset_id,
    "reference_dataset_id": reference_dataset_id,
    "manifest_path": manifest_path,
    "analysis_run_id": analysis_run_id,
    "quality_check_id": quality_check_id,
    "detection_count": int(detection_count),
    "raw_detection_count": run_result.get("raw_detection_count", int(detection_count)),
    "suppressed_detection_count": run_result.get("suppressed_detection_count", 0),
    "duplicate_iou_threshold": run_result.get("duplicate_iou_threshold"),
    "quality_status": quality_check.get("status"),
    "quality_score": quality_check.get("score"),
    "precision": metrics.get("precision"),
    "recall": metrics.get("recall"),
    "f1_score": metrics.get("f1_score", metrics.get("f1")),
    "mean_iou": metrics.get("mean_iou"),
    "matches": findings.get("matches"),
    "false_positives": findings.get("false_positives"),
    "false_negatives": findings.get("false_negatives"),
    "coverage_applied": coverage.get("applied", False),
    "coverage_mode": coverage.get("mode"),
    "coverage_tile_count": coverage.get("tile_count", 0),
    "coverage_source_crs_values": coverage.get("source_crs_values") or [],
    "candidate_raw_count": coverage.get("candidate_raw_count"),
    "candidate_evaluated_count": coverage.get("candidate_evaluated_count"),
    "candidate_excluded_outside_count": coverage.get("candidate_excluded_outside_count"),
    "candidate_clipped_boundary_count": coverage.get("candidate_clipped_boundary_count"),
    "reference_raw_count": reference_raw_count,
    "reference_evaluated_count": reference_evaluated_count,
    "reference_excluded_outside_count": coverage.get("reference_excluded_outside_count"),
    "reference_clipped_boundary_count": coverage.get("reference_clipped_boundary_count"),
    "reference_coverage_ratio": reference_coverage_ratio,
    "diagnostic_only": diagnostics.get("diagnostic_only"),
    "diagnostic_method": diagnostics.get("diagnostic_method"),
    "strict_matches": diagnostics.get("strict_matches"),
    "envelope_matches": diagnostics.get("envelope_matches"),
    "possible_box_to_footprint_mismatch_count": diagnostics.get("possible_box_to_footprint_mismatch_count"),
    "envelope_precision": diagnostics.get("envelope_precision"),
    "envelope_recall": diagnostics.get("envelope_recall"),
    "envelope_f1_score": diagnostics.get("envelope_f1_score"),
    "export_id": export_id,
    "run_log": run_log,
}
with open(output_path, "w", encoding="utf-8") as handle:
    json.dump(summary, handle, indent=2, sort_keys=True)
print(
    "model={model} tile={tile} overlap={overlap} threshold={threshold} detections={detections} raw={raw} suppressed={suppressed} "
    "score={score} precision={precision} recall={recall} f1={f1} matches={matches} fp={fp} fn={fn} "
    "coverage={coverage} diagnostic_gap={diagnostic_gap}".format(
        model=summary["model_asset_id"],
        tile=summary["tile_size"],
        overlap=summary["tile_overlap"],
        threshold=summary["threshold"],
        detections=summary["detection_count"],
        raw=summary["raw_detection_count"],
        suppressed=summary["suppressed_detection_count"],
        score=summary["quality_score"],
        precision=summary["precision"],
        recall=summary["recall"],
        f1=summary["f1_score"],
        matches=summary["matches"],
        fp=summary["false_positives"],
        fn=summary["false_negatives"],
        coverage=summary["reference_coverage_ratio"],
        diagnostic_gap=summary["possible_box_to_footprint_mismatch_count"],
    )
)
PY
done < "${matrix_manifest}"

"${PYTHON_BIN}" - "${QUALITY_OUTPUT_DIR}" "${BASE_URL}" "${matrix_manifest}" <<'PY'
import glob
import json
import os
import sys
from datetime import datetime, timezone

output_dir, base_url, manifest_path = sys.argv[1:4]
items = []
for path in sorted(glob.glob(os.path.join(output_dir, "*_summary.json"))):
    if path.endswith("quality_matrix_summary.json"):
        continue
    with open(path, "r", encoding="utf-8") as handle:
        items.append(json.load(handle))

if not items:
    raise SystemExit("No detection quality matrix summaries were produced")


def ranked_with(metric: str):
    return [item for item in items if item.get(metric) is not None]


best_by_score = max(ranked_with("quality_score"), key=lambda item: item["quality_score"], default=None)
best_by_recall = max(ranked_with("recall"), key=lambda item: item["recall"], default=None)
best_by_precision = max(ranked_with("precision"), key=lambda item: item["precision"], default=None)
summary = {
    "generated_at": datetime.now(timezone.utc).isoformat(),
    "base_url": base_url,
    "request_manifest_path": manifest_path,
    "run_count": len(items),
    "best_by_score": best_by_score,
    "best_by_recall": best_by_recall,
    "best_by_precision": best_by_precision,
    "items": items,
}
summary_path = os.path.join(output_dir, "quality_matrix_summary.json")
with open(summary_path, "w", encoding="utf-8") as handle:
    json.dump(summary, handle, indent=2, sort_keys=True)

print("")
print("Detection quality matrix summary")
print("model\ttile\toverlap\tthreshold\tdetections\traw\tsuppressed\tscore\tprecision\trecall\tf1\tmatches\tfp\tfn\tcoverage\tdiagnostic_gap")
for item in items:
    print(
        "{model_asset_id}\t{tile_size}\t{tile_overlap}\t{threshold:.2f}\t{detection_count}\t{raw_detection_count}\t{suppressed_detection_count}\t{quality_score}\t{precision}\t{recall}\t{f1_score}\t{matches}\t{false_positives}\t{false_negatives}\t{reference_coverage_ratio}\t{possible_box_to_footprint_mismatch_count}".format(
            **item
        )
    )
print("")
print(f"Summary: {summary_path}")
if best_by_score:
    print(
        "best_by_score model={model_asset_id} tile={tile_size} overlap={tile_overlap} threshold={threshold:.2f} score={quality_score} f1={f1_score}".format(
            **best_by_score
        )
    )
if best_by_recall:
    print(
        "best_by_recall model={model_asset_id} tile={tile_size} overlap={tile_overlap} threshold={threshold:.2f} recall={recall} detections={detection_count}".format(
            **best_by_recall
        )
    )
if best_by_precision:
    print(
        "best_by_precision model={model_asset_id} tile={tile_size} overlap={tile_overlap} threshold={threshold:.2f} precision={precision} detections={detection_count}".format(
            **best_by_precision
        )
    )
PY