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

ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
OUT_DIR="$ROOT_DIR/docs/consolidation/jupyter"
NOTEBOOKS_DIR="/Users/apple/notebooks"
DRY_RUN=1
APPLY=0

usage() {
  cat <<'USAGE'
Usage:
  bash scripts/docs/jupyter_sync.sh [--dry-run] [--apply] [--notebooks-dir PATH]

Behavior:
  - default mode is --dry-run
  - --apply writes markdown and csv outputs under docs/consolidation/jupyter
USAGE
}

while [[ $# -gt 0 ]]; do
  case "$1" in
    --dry-run)
      DRY_RUN=1
      APPLY=0
      shift
      ;;
    --apply)
      APPLY=1
      DRY_RUN=0
      shift
      ;;
    --notebooks-dir)
      NOTEBOOKS_DIR="$2"
      shift 2
      ;;
    -h|--help)
      usage
      exit 0
      ;;
    *)
      echo "Unknown arg: $1" >&2
      usage
      exit 2
      ;;
  esac
done

JUPYTER_CMD=""
if command -v jupyter >/dev/null 2>&1; then
  JUPYTER_CMD="jupyter"
elif python3 -m jupyter --version >/dev/null 2>&1; then
  JUPYTER_CMD="python3 -m jupyter"
fi

server_list=""
if [[ -n "$JUPYTER_CMD" ]]; then
  server_list="$(eval "$JUPYTER_CMD server list" 2>/dev/null || true)"
fi

server_rows="$(echo "$server_list" | awk '/^http/ {print}' || true)"
server_count="$(echo "$server_rows" | sed '/^\s*$/d' | wc -l | tr -d ' ')"

ok_api=0
total_api=0
api_lines=""
if [[ "$server_count" -gt 0 ]]; then
  while IFS= read -r line; do
    [[ -z "$line" ]] && continue
    url="$(echo "$line" | awk '{print $1}')"
    token="$(echo "$url" | sed -n 's/.*token=\([^&]*\).*/\1/p')"
    base="${url%%\?*}"
    status_url="${base%/}/api/status"
    if [[ -n "$token" ]]; then
      probe_url="${status_url}?token=${token}"
    else
      probe_url="$status_url"
    fi
    code="$(curl -sS -o /tmp/jupyter_status.json -w '%{http_code}' "$probe_url" || true)"
    total_api=$((total_api + 1))
    if [[ "$code" == "200" ]]; then
      ok_api=$((ok_api + 1))
    fi
    api_lines+="- ${probe_url} -> HTTP ${code}"$'\n'
  done <<< "$server_rows"
fi

nb_count=0
if [[ -d "$NOTEBOOKS_DIR" ]]; then
  nb_count="$(find "$NOTEBOOKS_DIR" -maxdepth 1 -type f -name '*.ipynb' | wc -l | tr -d ' ')"
fi

if [[ "$DRY_RUN" -eq 1 ]]; then
  echo "[dry-run] jupyter_cmd: ${JUPYTER_CMD:-<not-found>}"
  echo "[dry-run] server_count: $server_count"
  echo "[dry-run] api_ok: $ok_api/$total_api"
  echo "[dry-run] notebooks_dir: $NOTEBOOKS_DIR"
  echo "[dry-run] notebooks_count: $nb_count"
  if [[ -n "$api_lines" ]]; then
    echo "[dry-run] api_probes:"
    printf "%s" "$api_lines"
  fi
  exit 0
fi

mkdir -p "$OUT_DIR"
STAMP="$(date +%Y%m%d-%H%M%S)"
REPORT="$OUT_DIR/JUPYTER_SYNC_${STAMP}.md"
LATEST="$OUT_DIR/JUPYTER_SYNC_LATEST.md"
INDEX="$OUT_DIR/notebooks_index_${STAMP}.csv"
INDEX_LATEST="$OUT_DIR/notebooks_index_latest.csv"

{
  echo "# Jupyter Sync Report"
  echo
  echo "Generated: $(date -u '+%Y-%m-%d %H:%M:%S UTC')"
  echo
  echo "- jupyter_cmd: ${JUPYTER_CMD:-not-found}"
  echo "- server_count: $server_count"
  echo "- api_ok: $ok_api/$total_api"
  echo "- notebooks_dir: $NOTEBOOKS_DIR"
  echo "- notebooks_count: $nb_count"
  echo
  echo "## API Probes"
  echo
  if [[ -n "$api_lines" ]]; then
    printf "%s" "$api_lines"
  else
    echo "- no active jupyter servers discovered"
  fi
} > "$REPORT"

if [[ -d "$NOTEBOOKS_DIR" ]]; then
  {
    echo "path,size_bytes,mtime_epoch"
    find "$NOTEBOOKS_DIR" -maxdepth 1 -type f -name '*.ipynb' -print0 | \
      while IFS= read -r -d '' file; do
        size="$(stat -f '%z' "$file")"
        mtime="$(stat -f '%m' "$file")"
        echo "\"$file\",$size,$mtime"
      done
  } > "$INDEX"
else
  echo "path,size_bytes,mtime_epoch" > "$INDEX"
fi

cp "$REPORT" "$LATEST"
cp "$INDEX" "$INDEX_LATEST"

echo "Wrote: $REPORT"
echo "Updated: $LATEST"
echo "Wrote: $INDEX"
echo "Updated: $INDEX_LATEST"