"""Read scheduler metadata only; never run jobs or report their arguments."""
from pathlib import Path
import re
import shlex

SAFE_PATH = re.compile(r'^[A-Za-z0-9_./+-]+$')


def cron_rows(text, source, system_file=False):
    rows = []
    for number, raw in enumerate(text.splitlines(), 1):
        line = raw.strip()
        if not line or line.startswith('#') or re.match(r'^[A-Za-z_][A-Za-z0-9_]*\s*=', line):
            continue
        parts = line.split(None, 1 if line.startswith('@') else 5)
        count = 1 if line.startswith('@') else 5
        if count == 1 and parts[0] not in {'@reboot', '@yearly', '@annually', '@monthly', '@weekly', '@daily', '@midnight', '@hourly'}:
            continue
        if len(parts) <= count:
            continue
        schedule, command = ' '.join(parts[:count]), parts[count]
        account = 'Current user'
        if system_file:
            user_command = command.split(None, 1)
            if len(user_command) != 2:
                continue
            account, command = user_command
        # A safe executable/script filename is enough to identify the workload.
        # Arguments, embedded scripts and environment assignments never leave here.
        try:
            tokens = shlex.split(command)
        except ValueError:
            tokens = []
        executable = tokens[0] if tokens and SAFE_PATH.fullmatch(tokens[0]) else '[command omitted]'
        script = executable
        if Path(executable).name.startswith(('python', 'perl', 'ruby', 'bash', 'sh')) and len(tokens) > 1:
            if SAFE_PATH.fullmatch(tokens[1]) and not tokens[1].startswith('-'):
                script = tokens[1]
        rows.append({'name': f'{Path(script).name} ({source}:{number})', 'path': source,
                     'enabled': True, 'account': account, 'executable': executable,
                     'triggers': schedule + ' (cron; timezone not resolved)',
                     'state': 'Configured schedule; execution unverified'})
    return rows


def read_system_cron(cancelled=lambda: False):
    paths = [Path('/etc/crontab')]
    directory = Path('/etc/cron.d')
    if directory.exists():
        paths.extend(sorted(directory.iterdir()))
    rows, incomplete = [], False
    incomplete = len(paths) > 256
    for path in paths[:256]:
        if cancelled():
            from samsara.collectors.linux import CollectionCancelled
            raise CollectionCancelled()
        if not path.exists():
            continue
        # Avoid following symlinks or special files. Bound every file read.
        if path.is_symlink() or not path.is_file():
            incomplete = True
            continue
        if path.name.startswith('.') or path.name.endswith(('~', '.bak', '.dpkg-old', '.rpmnew', '.rpmsave')):
            continue
        try:
            with path.open('r', encoding='utf-8', errors='replace') as source:
                text = source.read(65537)
            if len(text) > 65536:
                incomplete = True
                continue
            rows.extend(cron_rows(text, str(path), system_file=True))
        except OSError:
            incomplete = True
    return rows, incomplete


def enrich_timers(rows, command):
    names = [row['name'] for row in rows if re.fullmatch(r'[A-Za-z0-9_.@:-]+\.timer', row['name'])]
    if not names:
        return
    output = command(['systemctl', 'show', '--property=Id', '--property=TimersCalendar',
                      '--property=NextElapseUSecRealtime', '--property=Triggers', '--no-pager', '--', *names[:200]])
    details = {}
    for block in output.strip().split('\n\n'):
        properties = dict(line.split('=', 1) for line in block.splitlines() if '=' in line)
        if 'Id' in properties:
            details[properties['Id']] = properties
    for row in rows:
        properties = details.get(row['name'], {})
        schedule = properties.get('TimersCalendar', '')
        next_run = properties.get('NextElapseUSecRealtime', '')
        if schedule or next_run:
            row['triggers'] = '; '.join(filter(None, [schedule, 'Next: ' + next_run if next_run else '']))
        if properties.get('Triggers'):
            row['associated_unit'] = properties['Triggers']