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

DATA_ROOT="${DATA_ROOT:-/app/data}"
CONFIG="${RETENTION_CONFIG:-/etc/retention/retention.json}"
GITEA_API_URL="${GITEA_API_URL:?GITEA_API_URL is required}"
GITEA_TOKEN="${GITEA_TOKEN:?GITEA_TOKEN is required}"

[ -d "$DATA_ROOT" ] || { echo "ERROR: data root missing: $DATA_ROOT" >&2; exit 1; }
[ -f "$CONFIG" ] || { echo "ERROR: config missing: $CONFIG" >&2; exit 1; }

default_max_age=$(jq -r '.default.maxAgeDays' "$CONFIG")
default_keep_min=$(jq -r '.default.keepMin' "$CONFIG")

gitea_get() {
  curl -fsS -H "Authorization: token ${GITEA_TOKEN}" \
    -H "Accept: application/json" "${GITEA_API_URL}${1}"
}

branch_names_for_repo() {
  local owner="$1" repo="$2" page=1
  while true; do
    local resp count
    resp=$(gitea_get "/api/v1/repos/${owner}/${repo}/branches?limit=50&page=${page}")
    count=$(echo "$resp" | jq 'length')
    [ "$count" -eq 0 ] && break
    echo "$resp" | jq -r '.[].name'
    [ "$count" -lt 50 ] && break
    page=$((page + 1))
  done
}

rule_max_age() {
  local branch="$1" v
  v=$(jq -r --arg b "$branch" '.branches[$b].maxAgeDays // empty' "$CONFIG")
  [ -n "$v" ] && echo "$v" || echo "$default_max_age"
}

rule_keep_min() {
  local branch="$1" v
  v=$(jq -r --arg b "$branch" '.branches[$b].keepMin // empty' "$CONFIG")
  [ -n "$v" ] && echo "$v" || echo "$default_keep_min"
}

delete_report_dir() {
  echo "DELETE $1 ($2)"
  rm -rf "$1"
}

age_days() {
  local meta="$1" published epoch_pub now
  published=$(jq -r '.published_at // empty' "$meta")
  if [ -z "$published" ]; then
    epoch_pub=$(stat -c %Y "$(dirname "$meta")" 2>/dev/null || stat -f %m "$(dirname "$meta")")
  else
    epoch_pub=$(date -u -d "$published" +%s 2>/dev/null || echo 0)
  fi
  now=$(date -u +%s)
  [ "$epoch_pub" -eq 0 ] && echo 99999 && return
  echo $(( (now - epoch_pub) / 86400 ))
}

CACHE_DIR=$(mktemp -d)
trap 'rm -rf "$CACHE_DIR"' EXIT

active_branches_file() {
  local owner="$1" repo="$2"
  local cache="${CACHE_DIR}/${owner}__${repo}.branches"
  if [ ! -f "$cache" ]; then
    branch_names_for_repo "$owner" "$repo" | sort -u > "$cache"
  fi
  echo "$cache"
}

branch_active() {
  grep -Fxq "$2" "$1"
}

parse_report_path() {
  local report_dir="$1" rel rest
  rel="${report_dir#"${DATA_ROOT}/"}"
  REPO_OWNER="${rel%%/*}"
  rest="${rel#*/}"
  REPO_NAME="${rest%%/*}"
}

# Pass 1: deleted branches (always — no config)
while IFS= read -r meta; do
  report_dir=$(dirname "$meta")
  parse_report_path "$report_dir"
  branch=$(jq -r '.branch // empty' "$meta")
  [ -n "$branch" ] || continue

  cache=$(active_branches_file "$REPO_OWNER" "$REPO_NAME")
  if ! branch_active "$cache" "$branch"; then
    delete_report_dir "$report_dir" "branch deleted: ${branch}"
  fi
done < <(find "$DATA_ROOT" -path '*/reports/*/.meta' -type f 2>/dev/null)

# Pass 2: age + keepMin for active branches (newest first)
SURVIVORS="${CACHE_DIR}/survivors.tsv"
: > "$SURVIVORS"

while IFS= read -r meta; do
  report_dir=$(dirname "$meta")
  [ -d "$report_dir" ] || continue
  parse_report_path "$report_dir"
  branch=$(jq -r '.branch // empty' "$meta")
  [ -n "$branch" ] || continue

  cache=$(active_branches_file "$REPO_OWNER" "$REPO_NAME")
  branch_active "$cache" "$branch" || continue

  printf '%s\t%s\t%s\t%s\n' "${REPO_OWNER}/${REPO_NAME}" "$branch" "$report_dir" "$(age_days "$meta")" >> "$SURVIVORS"
done < <(find "$DATA_ROOT" -path '*/reports/*/.meta' -type f 2>/dev/null)

current_key=""
kept=0

while IFS=$'\t' read -r repo_key branch report_dir days; do
  key="${repo_key}|${branch}"
  if [ "$key" != "$current_key" ]; then
    current_key="$key"
    kept=0
    max_age=$(rule_max_age "$branch")
    keep_min=$(rule_keep_min "$branch")
  fi

  if [ "$days" -gt "$max_age" ]; then
    delete_report_dir "$report_dir" "older than ${max_age}d (branch ${branch})"
    continue
  fi

  kept=$((kept + 1))
  if [ "$kept" -gt "$keep_min" ]; then
    delete_report_dir "$report_dir" "exceeds keepMin ${keep_min} (branch ${branch})"
  fi
done < <(sort -t $'\t' -k1,1 -k2,2 -k4,4n "$SURVIVORS")

echo "Retention cleanup finished."