ProxmoxVE/misc/error_handler.func

#!/usr/bin/env bash
# ------------------------------------------------------------------------------
# ERROR HANDLER - ERROR & SIGNAL MANAGEMENT
# ------------------------------------------------------------------------------
# Copyright (c) 2021-2026 community-scripts ORG
# Author: MickLesk (CanbiZ)
# License: MIT | https://github.com/community-scripts/ProxmoxVE/raw/main/LICENSE
# ------------------------------------------------------------------------------
#
# Provides comprehensive error handling and signal management for all scripts.
# Includes:
#   - Exit code explanations (shell, package managers, databases, custom codes)
#   - Error handler with detailed logging
#   - Signal handlers (EXIT, INT, TERM)
#   - Initialization function for trap setup
#
# Usage:
#   source <(curl -fsSL .../error_handler.func)
#   catch_errors
#
# ------------------------------------------------------------------------------

# ==============================================================================
# SECTION 1: EXIT CODE EXPLANATIONS
# ==============================================================================

# ------------------------------------------------------------------------------
# explain_exit_code()
#
# - Canonical version is defined in api.func (sourced before this file)
# - This section only provides a fallback if api.func was not loaded
# - See api.func SECTION 1 for the authoritative exit code mappings
# ------------------------------------------------------------------------------
if ! declare -f explain_exit_code &>/dev/null; then
  explain_exit_code() {
    local code="$1"
    case "$code" in
    1) echo "General error / Operation not permitted" ;;
    2) echo "Misuse of shell builtins (e.g. syntax error)" ;;
    3) echo "General syntax or argument error" ;;
    10) echo "Docker / privileged mode required (unsupported environment)" ;;
    4) echo "curl: Feature not supported or protocol error" ;;
    5) echo "curl: Could not resolve proxy" ;;
    6) echo "curl: DNS resolution failed (could not resolve host)" ;;
    7) echo "curl: Failed to connect (network unreachable / host down)" ;;
    8) echo "curl: Server reply error (FTP/SFTP or apk untrusted key)" ;;
    16) echo "curl: HTTP/2 framing layer error" ;;
    18) echo "curl: Partial file (transfer not completed)" ;;
    22) echo "curl: HTTP error returned (404, 429, 500+)" ;;
    23) echo "curl: Write error (disk full or permissions)" ;;
    24) echo "curl: Write to local file failed" ;;
    25) echo "curl: Upload failed" ;;
    26) echo "curl: Read error on local file (I/O)" ;;
    27) echo "curl: Out of memory (memory allocation failed)" ;;
    28) echo "curl: Operation timeout (network slow or server not responding)" ;;
    30) echo "curl: FTP port command failed" ;;
    32) echo "curl: FTP SIZE command failed" ;;
    33) echo "curl: HTTP range error" ;;
    34) echo "curl: HTTP post error" ;;
    35) echo "curl: SSL/TLS handshake failed (certificate error)" ;;
    36) echo "curl: FTP bad download resume" ;;
    39) echo "curl: LDAP search failed" ;;
    44) echo "curl: Internal error (bad function call order)" ;;
    45) echo "curl: Interface error (failed to bind to specified interface)" ;;
    46) echo "curl: Bad password entered" ;;
    47) echo "curl: Too many redirects" ;;
    48) echo "curl: Unknown command line option specified" ;;
    51) echo "curl: SSL peer certificate or SSH host key verification failed" ;;
    52) echo "curl: Empty reply from server (got nothing)" ;;
    55) echo "curl: Failed sending network data" ;;
    56) echo "curl: Receive error (connection reset by peer)" ;;
    57) echo "curl: Unrecoverable poll/select error (system I/O failure)" ;;
    59) echo "curl: Couldn't use specified SSL cipher" ;;
    61) echo "curl: Bad/unrecognized transfer encoding" ;;
    63) echo "curl: Maximum file size exceeded" ;;
    75) echo "Temporary failure (retry later)" ;;
    78) echo "curl: Remote file not found (404 on FTP/file)" ;;
    79) echo "curl: SSH session error (key exchange/auth failed)" ;;
    92) echo "curl: HTTP/2 stream error (protocol violation)" ;;
    95) echo "curl: HTTP/3 layer error" ;;
    64) echo "Usage error (wrong arguments)" ;;
    65) echo "Data format error (bad input data)" ;;
    66) echo "Input file not found (cannot open input)" ;;
    67) echo "User not found (addressee unknown)" ;;
    68) echo "Host not found (hostname unknown)" ;;
    69) echo "Service unavailable" ;;
    70) echo "Internal software error" ;;
    71) echo "System error (OS-level failure)" ;;
    72) echo "Critical OS file missing" ;;
    73) echo "Cannot create output file" ;;
    74) echo "I/O error" ;;
    76) echo "Remote protocol error" ;;
    77) echo "Permission denied" ;;
    100) echo "APT: Package manager error (broken packages / dependency problems)" ;;
    101) echo "APT: Configuration error (bad sources.list, malformed config)" ;;
    102) echo "APT: Lock held by another process (dpkg/apt still running)" ;;
    124) echo "Command timed out (timeout command)" ;;
    125) echo "Command failed to start (Docker daemon or execution error)" ;;
    126) echo "Command invoked cannot execute (permission problem?)" ;;
    127) echo "Command not found" ;;
    128) echo "Invalid argument to exit" ;;
    129) echo "Killed by SIGHUP (terminal closed / hangup)" ;;
    130) echo "Aborted by user (SIGINT)" ;;
    131) echo "Killed by SIGQUIT (core dumped)" ;;
    132) echo "Killed by SIGILL (illegal CPU instruction)" ;;
    134) echo "Process aborted (SIGABRT - possibly Node.js heap overflow)" ;;
    137) echo "Killed (SIGKILL / Out of memory?)" ;;
    139) echo "Segmentation fault (core dumped)" ;;
    141) echo "Broken pipe (SIGPIPE - output closed prematurely)" ;;
    143) echo "Terminated (SIGTERM)" ;;
    144) echo "Killed by signal 16 (SIGUSR1 / SIGSTKFLT)" ;;
    146) echo "Killed by signal 18 (SIGTSTP)" ;;
    150) echo "Systemd: Service failed to start" ;;
    151) echo "Systemd: Service unit not found" ;;
    152) echo "Permission denied (EACCES)" ;;
    153) echo "Build/compile failed (make/gcc/cmake)" ;;
    154) echo "Node.js: Native addon build failed (node-gyp)" ;;
    160) echo "Python: Virtualenv / uv environment missing or broken" ;;
    161) echo "Python: Dependency resolution failed" ;;
    162) echo "Python: Installation aborted (permissions or EXTERNALLY-MANAGED)" ;;
    170) echo "PostgreSQL: Connection failed (server not running / wrong socket)" ;;
    171) echo "PostgreSQL: Authentication failed (bad user/password)" ;;
    172) echo "PostgreSQL: Database does not exist" ;;
    173) echo "PostgreSQL: Fatal error in query / syntax" ;;
    180) echo "MySQL/MariaDB: Connection failed (server not running / wrong socket)" ;;
    181) echo "MySQL/MariaDB: Authentication failed (bad user/password)" ;;
    182) echo "MySQL/MariaDB: Database does not exist" ;;
    183) echo "MySQL/MariaDB: Fatal error in query / syntax" ;;
    190) echo "MongoDB: Connection failed (server not running)" ;;
    191) echo "MongoDB: Authentication failed (bad user/password)" ;;
    192) echo "MongoDB: Database not found" ;;
    193) echo "MongoDB: Fatal query error" ;;
    200) echo "Proxmox: Failed to create lock file" ;;
    203) echo "Proxmox: Missing CTID variable" ;;
    204) echo "Proxmox: Missing PCT_OSTYPE variable" ;;
    205) echo "Proxmox: Invalid CTID (<100)" ;;
    206) echo "Proxmox: CTID already in use" ;;
    207) echo "Proxmox: Password contains unescaped special characters" ;;
    208) echo "Proxmox: Invalid configuration (DNS/MAC/Network format)" ;;
    209) echo "Proxmox: Container creation failed" ;;
    210) echo "Proxmox: Cluster not quorate" ;;
    211) echo "Proxmox: Timeout waiting for template lock" ;;
    212) echo "Proxmox: Storage type 'iscsidirect' does not support containers (VMs only)" ;;
    213) echo "Proxmox: Storage type does not support 'rootdir' content" ;;
    214) echo "Proxmox: Not enough storage space" ;;
    215) echo "Proxmox: Container created but not listed (ghost state)" ;;
    216) echo "Proxmox: RootFS entry missing in config" ;;
    217) echo "Proxmox: Storage not accessible" ;;
    218) echo "Proxmox: Template file corrupted or incomplete" ;;
    219) echo "Proxmox: CephFS does not support containers - use RBD" ;;
    220) echo "Proxmox: Unable to resolve template path" ;;
    221) echo "Proxmox: Template file not readable" ;;
    222) echo "Proxmox: Template download failed" ;;
    223) echo "Proxmox: Template not available after download" ;;
    224) echo "Proxmox: PBS storage is for backups only" ;;
    225) echo "Proxmox: No template available for OS/Version" ;;
    231) echo "Proxmox: LXC stack upgrade failed" ;;
    239) echo "npm/Node.js: Unexpected runtime error or dependency failure" ;;
    243) echo "Node.js: Out of memory (JavaScript heap out of memory)" ;;
    245) echo "Node.js: Invalid command-line option" ;;
    246) echo "Node.js: Internal JavaScript Parse Error" ;;
    247) echo "Node.js: Fatal internal error" ;;
    248) echo "Node.js: Invalid C++ addon / N-API failure" ;;
    249) echo "npm/pnpm/yarn: Unknown fatal error" ;;
    255) echo "DPKG: Fatal internal error" ;;
    *) echo "Unknown error" ;;
    esac
  }
fi

# ==============================================================================
# SECTION 2: ERROR HANDLERS
# ==============================================================================

# ------------------------------------------------------------------------------
# error_handler()
#
# - Main error handler triggered by ERR trap
# - Arguments: exit_code, command, line_number
# - Behavior:
#   * Returns silently if exit_code is 0 (success)
#   * Sources explain_exit_code() for detailed error description
#   * Displays error message with:
#     - Line number where error occurred
#     - Exit code with explanation
#     - Command that failed
#   * Shows last 20 lines of SILENT_LOGFILE if available
#   * Copies log to container /root for later inspection
#   * Exits with original exit code
# ------------------------------------------------------------------------------
error_handler() {
  local exit_code=${1:-$?}
  local command=${2:-${BASH_COMMAND:-unknown}}
  local line_number=${BASH_LINENO[0]:-unknown}

  command="${command//\$STD/}"

  if [[ "$exit_code" -eq 0 ]]; then
    return 0
  fi

  # Stop spinner and restore cursor FIRST — before any output
  # This prevents spinner text overlapping with error messages
  if declare -f stop_spinner >/dev/null 2>&1; then
    stop_spinner 2>/dev/null || true
  fi
  printf "\e[?25h"

  local explanation
  explanation="$(explain_exit_code "$exit_code")"

  # ALWAYS report failure to API immediately - don't wait for container checks
  # This ensures we capture failures that occur before/after container exists
  if declare -f post_update_to_api &>/dev/null; then
    post_update_to_api "failed" "$exit_code" 2>/dev/null || true
  else
    # Container context: post_update_to_api not available (api.func not sourced)
    # Send status directly via curl so container failures are never lost
    _send_abort_telemetry "$exit_code" 2>/dev/null || true
  fi

  # Use msg_error if available, fallback to echo
  if declare -f msg_error >/dev/null 2>&1; then
    msg_error "in line ${line_number}: exit code ${exit_code} (${explanation}): while executing command ${command}"
  else
    echo -e "\n${RD}[ERROR]${CL} in line ${RD}${line_number}${CL}: exit code ${RD}${exit_code}${CL} (${explanation}): while executing command ${YWB}${command}${CL}\n"
  fi

  if [[ -n "${DEBUG_LOGFILE:-}" ]]; then
    {
      echo "------ ERROR ------"
      echo "Timestamp : $(date '+%Y-%m-%d %H:%M:%S')"
      echo "Exit Code : $exit_code ($explanation)"
      echo "Line      : $line_number"
      echo "Command   : $command"
      echo "-------------------"
    } >>"$DEBUG_LOGFILE"
  fi

  # Get active log file (BUILD_LOG or INSTALL_LOG)
  local active_log=""
  if declare -f get_active_logfile >/dev/null 2>&1; then
    active_log="$(get_active_logfile)"
  elif [[ -n "${SILENT_LOGFILE:-}" ]]; then
    active_log="$SILENT_LOGFILE"
  fi

  # If active_log points to a container-internal path that doesn't exist on host,
  # fall back to BUILD_LOG (host-side log)
  if [[ -n "$active_log" && ! -s "$active_log" && -n "${BUILD_LOG:-}" && -s "${BUILD_LOG}" ]]; then
    active_log="$BUILD_LOG"
  fi

  # Show last log lines if available
  if [[ -n "$active_log" && -s "$active_log" ]]; then
    echo -e "\n${TAB}--- Last 20 lines of log ---"
    tail -n 20 "$active_log"
    echo -e "${TAB}-----------------------------------\n"
  fi

  # Detect context: Container (INSTALL_LOG set + inside container /root) vs Host
  if [[ -n "${INSTALL_LOG:-}" && -f "${INSTALL_LOG:-}" && -d /root ]]; then
    # CONTAINER CONTEXT: Copy log and create flag file for host
    local container_log="/root/.install-${SESSION_ID:-error}.log"
    cp "${INSTALL_LOG}" "$container_log" 2>/dev/null || true

    # Create error flag file with exit code for host detection
    echo "$exit_code" >"/root/.install-${SESSION_ID:-error}.failed" 2>/dev/null || true
    # Log path is shown by host as combined log - no need to show container path
  else
    # HOST CONTEXT: Show local log path and offer container cleanup
    if [[ -n "$active_log" && -s "$active_log" ]]; then
      if declare -f msg_custom >/dev/null 2>&1; then
        msg_custom "📋" "${YW}" "Full log: ${active_log}"
      else
        echo -e "${YW}Full log:${CL} ${BL}${active_log}${CL}"
      fi
    fi

    # Offer to remove container if it exists (build errors after container creation)
    if [[ -n "${CTID:-}" ]] && command -v pct &>/dev/null && pct status "$CTID" &>/dev/null; then
      echo ""
      if declare -f msg_custom >/dev/null 2>&1; then
        echo -en "${TAB}❓${TAB}${YW}Remove broken container ${CTID}? (Y/n) [auto-remove in 60s]: ${CL}"
      else
        echo -en "${YW}Remove broken container ${CTID}? (Y/n) [auto-remove in 60s]: ${CL}"
      fi

      if read -t 60 -r response; then
        if [[ -z "$response" || "$response" =~ ^[Yy]$ ]]; then
          echo ""
          if declare -f msg_info >/dev/null 2>&1; then
            msg_info "Removing container ${CTID}"
          else
            echo -e "${YW}Removing container ${CTID}${CL}"
          fi
          pct stop "$CTID" &>/dev/null || true
          pct destroy "$CTID" &>/dev/null || true
          if declare -f msg_ok >/dev/null 2>&1; then
            msg_ok "Container ${CTID} removed"
          else
            echo -e "${GN}✔${CL} Container ${CTID} removed"
          fi
        elif [[ "$response" =~ ^[Nn]$ ]]; then
          echo ""
          if declare -f msg_warn >/dev/null 2>&1; then
            msg_warn "Container ${CTID} kept for debugging"
          else
            echo -e "${YW}Container ${CTID} kept for debugging${CL}"
          fi
        fi
      else
        # Timeout - auto-remove
        echo ""
        if declare -f msg_info >/dev/null 2>&1; then
          msg_info "No response - removing container ${CTID}"
        else
          echo -e "${YW}No response - removing container ${CTID}${CL}"
        fi
        pct stop "$CTID" &>/dev/null || true
        pct destroy "$CTID" &>/dev/null || true
        if declare -f msg_ok >/dev/null 2>&1; then
          msg_ok "Container ${CTID} removed"
        else
          echo -e "${GN}✔${CL} Container ${CTID} removed"
        fi
      fi

      # Force one final status update attempt after cleanup
      # This ensures status is updated even if the first attempt failed (e.g., HTTP 400)
      if declare -f post_update_to_api &>/dev/null; then
        post_update_to_api "failed" "$exit_code" "force"
      fi
    fi
  fi

  exit "$exit_code"
}

# ==============================================================================
# SECTION 3: TELEMETRY & CLEANUP HELPERS FOR SIGNAL HANDLERS
# ==============================================================================

# ------------------------------------------------------------------------------
# _send_abort_telemetry()
#
# - Sends failure/abort status to telemetry API
# - Works in BOTH host context (post_update_to_api available) and
#   container context (only curl available, api.func not sourced)
# - Container context is critical: without this, container-side failures
#   and signal exits are never reported, leaving records stuck in
#   "installing" or "configuring" forever
# - Arguments: $1 = exit_code
# ------------------------------------------------------------------------------
_send_abort_telemetry() {
  local exit_code="${1:-1}"
  # Try full API function first (host context - api.func sourced)
  if declare -f post_update_to_api &>/dev/null; then
    post_update_to_api "failed" "$exit_code" 2>/dev/null || true
    return
  fi
  # Fallback: direct curl (container context - api.func NOT sourced)
  # This is the ONLY way containers can report failures to telemetry
  command -v curl &>/dev/null || return 0
  [[ "${DIAGNOSTICS:-no}" == "no" ]] && return 0
  [[ -z "${RANDOM_UUID:-}" ]] && return 0

  # Collect last 20 log lines for error diagnosis (best-effort)
  local error_text=""
  if [[ -n "${INSTALL_LOG:-}" && -s "${INSTALL_LOG}" ]]; then
    error_text=$(tail -n 20 "$INSTALL_LOG" 2>/dev/null | sed 's/\x1b\[[0-9;]*[a-zA-Z]//g; s/\\/\\\\/g; s/"/\\"/g; s/\r//g' | tr '\n' '|' | sed 's/|$//' | tr -d '\000-\010\013\014\016-\037\177') || true
  fi

  # Calculate duration if start time is available
  local duration=""
  if [[ -n "${DIAGNOSTICS_START_TIME:-}" ]]; then
    duration=$(($(date +%s) - DIAGNOSTICS_START_TIME))
  fi

  # Build JSON payload with error context
  local payload
  payload="{\"random_id\":\"${RANDOM_UUID}\",\"execution_id\":\"${EXECUTION_ID:-${RANDOM_UUID}}\",\"type\":\"${TELEMETRY_TYPE:-lxc}\",\"nsapp\":\"${NSAPP:-${app:-unknown}}\",\"status\":\"failed\",\"exit_code\":${exit_code}"
  [[ -n "$error_text" ]] && payload="${payload},\"error\":\"${error_text}\""
  [[ -n "$duration" ]] && payload="${payload},\"duration\":${duration}"
  payload="${payload}}"

  local api_url="${TELEMETRY_URL:-https://telemetry.community-scripts.org/telemetry}"

  # 2 attempts (retry once on failure) — original had no retry
  local attempt
  for attempt in 1 2; do
    if curl -fsS -m 5 -X POST "$api_url" \
      -H "Content-Type: application/json" \
      -d "$payload" &>/dev/null; then
      return 0
    fi
    [[ $attempt -eq 1 ]] && sleep 1
  done
  return 0
}

# ------------------------------------------------------------------------------
# _stop_container_if_installing()
#
# - Stops the LXC container if we're in the install phase
# - Prevents orphaned container processes when the host exits due to a signal
#   (SSH disconnect, Ctrl+C, SIGTERM) — without this, the container keeps
#   running and may send "configuring" status AFTER the host already sent
#   "failed", leaving records permanently stuck in "configuring"
# - Only acts when:
#   * CONTAINER_INSTALLING flag is set (during lxc-attach in build_container)
#   * CTID is set (container was created)
#   * pct command is available (we're on the Proxmox host, not inside a container)
# - Does NOT destroy the container — just stops it for potential debugging
# ------------------------------------------------------------------------------
_stop_container_if_installing() {
  [[ "${CONTAINER_INSTALLING:-}" == "true" ]] || return 0
  [[ -n "${CTID:-}" ]] || return 0
  command -v pct &>/dev/null || return 0
  pct stop "$CTID" 2>/dev/null || true
}

# ==============================================================================
# SECTION 4: SIGNAL HANDLERS
# ==============================================================================

# ------------------------------------------------------------------------------
# on_exit()
#
# - EXIT trap handler — runs on EVERY script termination
# - Catches orphaned "installing"/"configuring" records:
#   * If post_to_api sent "installing" but post_update_to_api never ran
#   * Reports final status to prevent records stuck forever
# - Best-effort log collection for failed installs
# - Stops orphaned container processes on failure
# - Cleans up lock files
# ------------------------------------------------------------------------------
on_exit() {
  local exit_code=$?

  # Report orphaned "installing" records to telemetry API
  # Catches ALL exit paths: errors, signals, AND clean exits where
  # post_to_api was called but post_update_to_api was never called
  if [[ "${POST_TO_API_DONE:-}" == "true" && "${POST_UPDATE_DONE:-}" != "true" ]]; then
    if [[ $exit_code -ne 0 ]]; then
      _send_abort_telemetry "$exit_code"
    elif declare -f post_update_to_api >/dev/null 2>&1; then
      post_update_to_api "done" "0" 2>/dev/null || true
    fi
  fi

  # Best-effort log collection on failure (non-critical, telemetry already sent)
  if [[ $exit_code -ne 0 ]] && declare -f ensure_log_on_host >/dev/null 2>&1; then
    ensure_log_on_host 2>/dev/null || true
  fi

  # Stop orphaned container if we're in the install phase and exiting with error
  if [[ $exit_code -ne 0 ]]; then
    _stop_container_if_installing
  fi

  [[ -n "${lockfile:-}" && -e "$lockfile" ]] && rm -f "$lockfile"
  exit "$exit_code"
}

# ------------------------------------------------------------------------------
# on_interrupt()
#
# - SIGINT (Ctrl+C) trap handler
# - Reports status FIRST (time-critical: container may be dying)
# - Stops orphaned container to prevent "configuring" ghost records
# - Exits with code 130 (128 + SIGINT=2)
# ------------------------------------------------------------------------------
on_interrupt() {
  # Stop spinner and restore cursor before any output
  if declare -f stop_spinner >/dev/null 2>&1; then
    stop_spinner 2>/dev/null || true
  fi
  printf "\e[?25h" 2>/dev/null || true

  _send_abort_telemetry "130"
  _stop_container_if_installing
  if declare -f msg_error >/dev/null 2>&1; then
    msg_error "Interrupted by user (SIGINT)" 2>/dev/null || true
  else
    echo -e "\n${RD}Interrupted by user (SIGINT)${CL}" 2>/dev/null || true
  fi
  exit 130
}

# ------------------------------------------------------------------------------
# on_terminate()
#
# - SIGTERM trap handler
# - Reports status FIRST (time-critical: process being killed)
# - Stops orphaned container to prevent "configuring" ghost records
# - Exits with code 143 (128 + SIGTERM=15)
# ------------------------------------------------------------------------------
on_terminate() {
  # Stop spinner and restore cursor before any output
  if declare -f stop_spinner >/dev/null 2>&1; then
    stop_spinner 2>/dev/null || true
  fi
  printf "\e[?25h" 2>/dev/null || true

  _send_abort_telemetry "143"
  _stop_container_if_installing
  if declare -f msg_error >/dev/null 2>&1; then
    msg_error "Terminated by signal (SIGTERM)" 2>/dev/null || true
  else
    echo -e "\n${RD}Terminated by signal (SIGTERM)${CL}" 2>/dev/null || true
  fi
  exit 143
}

# ------------------------------------------------------------------------------
# on_hangup()
#
# - SIGHUP trap handler (SSH disconnect, terminal closed)
# - CRITICAL: This was previously MISSING from catch_errors(), causing
#   container processes to become orphans on SSH disconnect — the #1 cause
#   of records stuck in "installing" and "configuring" states
# - Reports status via direct curl (terminal is already closed, no output)
# - Stops orphaned container to prevent ghost records
# - Exits with code 129 (128 + SIGHUP=1)
# ------------------------------------------------------------------------------
on_hangup() {
  # Stop spinner (no cursor restore needed — terminal is already gone)
  if declare -f stop_spinner >/dev/null 2>&1; then
    stop_spinner 2>/dev/null || true
  fi

  _send_abort_telemetry "129"
  _stop_container_if_installing
  exit 129
}

# ==============================================================================
# SECTION 5: INITIALIZATION
# ==============================================================================

# ------------------------------------------------------------------------------
# catch_errors()
#
# - Initializes error handling and signal traps
# - Enables strict error handling:
#   * set -Ee: Exit on error, inherit ERR trap in functions
#   * set -o pipefail: Pipeline fails if any command fails
#   * set -u: (optional) Exit on undefined variable (if STRICT_UNSET=1)
# - Sets up traps:
#   * ERR  → error_handler (script errors)
#   * EXIT → on_exit      (any termination — cleanup + orphan detection)
#   * INT  → on_interrupt  (Ctrl+C)
#   * TERM → on_terminate  (kill / systemd stop)
#   * HUP  → on_hangup     (SSH disconnect / terminal closed)
# - Call this function early in every script
# ------------------------------------------------------------------------------
catch_errors() {
  set -Ee -o pipefail
  if [ "${STRICT_UNSET:-0}" = "1" ]; then
    set -u
  fi

  trap 'error_handler' ERR
  trap on_exit EXIT
  trap on_interrupt INT
  trap on_terminate TERM
  trap on_hangup HUP
}