dtf-system/deploy/production_preflight.py

"""Fail closed until the application and non-secret production inputs are ready."""
import os
from pathlib import Path
import re
import sys
from urllib.parse import urlparse

ROOT = Path(__file__).resolve().parent.parent
APPROVALS = (
    'PRODUCTION_DEPLOY_ENABLED',
    'PRODUCTION_INPUTS_APPROVED',
    'PRODUCTION_SECURITY_REVIEW_APPROVED',
    'PRODUCTION_RESTORE_REHEARSED',
)
REQUIRED = (
    'API_IMAGE', 'WEB_IMAGE', 'POSTGRES_IMAGE', 'CLAMAV_IMAGE',
    'IMAGE_TAG', 'PUBLIC_ORIGIN', 'PUBLIC_HOST', 'KANBAN_HOST',
    'SITE_PORT', 'KANBAN_PORT',
    'R2_ENDPOINT', 'R2_PUBLIC_ENDPOINT', 'R2_BUCKET',
    'POSTGRES_DB', 'POSTGRES_USER', 'APP_DB_USER', 'POSTGRES_VOLUME', 'OPERATOR_EMAIL',
    'STORAGE_QUOTA_BYTES', 'OWNER_UPLOAD_QUOTA_BYTES', 'MAX_UPLOAD_BYTES',
    'UPLOAD_PART_BYTES', 'MAX_PENDING_UPLOADS', 'SCAN_MAX_BYTES',
    'PAYMENT_ADAPTER', 'FREIGHT_ADAPTER', 'TINY_ADAPTER', 'WHATSAPP_ADAPTER',
    'DATABASE_URL_SECRET', 'DATABASE_ADMIN_URL_SECRET',
    'DB_ADMIN_PASSWORD_SECRET', 'APP_DB_PASSWORD_SECRET',
    'R2_ACCESS_KEY_ID_SECRET', 'R2_SECRET_ACCESS_KEY_SECRET',
    'OPERATOR_PASSWORD_SECRET', 'PAYMENT_TOKEN_SECRET',
    'PAYMENT_WEBHOOK_SECRET', 'TINY_TOKEN_SECRET', 'WHATSAPP_TOKEN_SECRET',
)
IMMUTABLE_IMAGE = re.compile(r'^[a-z0-9][a-z0-9._:/-]*@sha256:([0-9a-f]{64})$')
IMAGE_REPOSITORY = re.compile(
    r'^[a-z0-9][a-z0-9.-]*(?::[0-9]{1,5})?(?:/[a-z0-9][a-z0-9._-]*)+$')
DNS = re.compile(r'^(?=.{1,253}$)(?:[a-z0-9](?:[a-z0-9-]{0,61}[a-z0-9])?\.)+[a-z]{2,63}$')
NAME = re.compile(r'^[a-zA-Z0-9][a-zA-Z0-9_.-]{2,127}$')
SOURCE_BLOCKERS = {
    # Markers must name something that is still true, or the gate weakens without
    # failing. Four entries here described a local-only runtime and stopped
    # matching when R2 support landed; they were removed rather than left to rot.
    # What remains is the real blocker: no production payment or messaging adapter
    # exists, so these lines must change before a release can be meaningful.
    'local/runtime.py': (
        'payment = FakePayment()',
    ),
    'local/worker.py': (
        "adapters = {'tiny': FakeTiny(), 'whatsapp': FakeWhatsApp()}",
    ),
}


def source_errors(root=ROOT):
    errors = []
    for relative, markers in SOURCE_BLOCKERS.items():
        text = (root / relative).read_text()
        for marker in markers:
            if marker in text:
                errors.append(f'{relative} remains local-only: {marker}')
    errors.extend(secret_loading_errors(root))
    return errors


def secret_loading_errors(root=ROOT):
    """Exercise the secret loader instead of grepping it.

    Searching for a string passes as soon as someone writes that string, and
    fails when a working implementation happens to spell it differently. Load the
    module and make it resolve a real file.
    """
    import importlib.util
    import tempfile

    module_path = root / 'local' / 'secrets.py'
    if not module_path.exists():
        return ['local runtime does not load the production Docker secret *_FILE settings']
    try:
        spec = importlib.util.spec_from_file_location('_preflight_secrets', module_path)
        module = importlib.util.module_from_spec(spec)
        spec.loader.exec_module(module)
    except Exception as exc:
        return [f'local/secrets.py could not be loaded: {exc}']

    stack_names = set()
    stack = root / 'deploy' / 'stack.yaml'
    if stack.exists():
        for line in stack.read_text().splitlines():
            if '_FILE:' in line:
                key = line.split(':')[0].strip()
                # The database image consumes this one; the application does not.
                if key and key != 'POSTGRES_PASSWORD_FILE':
                    stack_names.add(key[:-len('_FILE')])

    failures = []
    with tempfile.TemporaryDirectory() as directory:
        for name in sorted(stack_names):
            path = Path(directory) / name
            path.write_text('resolved-value\n', encoding='utf-8')
            environ = {f'{name}_FILE': str(path)}
            try:
                module.load(environ)
            except Exception as exc:
                failures.append(f'{name}_FILE is not resolved by local/secrets.py: {exc}')
                continue
            if environ.get(name) != 'resolved-value':
                failures.append(f'{name}_FILE did not produce {name}')
        # A missing secret must stop the service, never start it unconfigured.
        try:
            module.load({'DATABASE_URL_FILE': str(Path(directory) / 'absent')})
        except Exception:
            pass
        else:
            failures.append('local/secrets.py does not fail closed on an unreadable secret')
    return failures


def config_errors(values):
    errors = []
    for name in APPROVALS:
        if values.get(name) != 'approved':
            errors.append(f'{name} must equal approved')
    for name in REQUIRED:
        value = values.get(name, '')
        if not value or value.lower() in {'tbd', 'todo', 'replace-me', 'unconfirmed'}:
            errors.append(f'{name} is missing or unresolved')
    for name in ('API_IMAGE', 'WEB_IMAGE'):
        if not IMAGE_REPOSITORY.fullmatch(values.get(name, '')):
            errors.append(f'{name} must be a lowercase registry repository without a tag')
    for name in ('POSTGRES_IMAGE', 'CLAMAV_IMAGE'):
        match = IMMUTABLE_IMAGE.fullmatch(values.get(name, ''))
        if not match or match.group(1) == '0' * 64:
            errors.append(f'{name} must be an immutable non-placeholder digest reference')
    image_tag = values.get('IMAGE_TAG', '')
    if image_tag != 'latest' and not re.fullmatch(r'(?:[0-9a-f]{40}|[0-9a-f]{64})', image_tag):
        errors.append('IMAGE_TAG must be latest or a full commit SHA published by Gitea')
    origin = urlparse(values.get('PUBLIC_ORIGIN', ''))
    if (origin.scheme != 'https' or not origin.hostname or origin.path not in ('', '/')
            or origin.params or origin.query or origin.fragment):
        errors.append('PUBLIC_ORIGIN must be an HTTPS origin without a path')
    for name in ('PUBLIC_HOST', 'KANBAN_HOST'):
        if not DNS.fullmatch(values.get(name, '')):
            errors.append(f'{name} must be a valid lowercase DNS hostname')
    if origin.hostname and values.get('PUBLIC_HOST') != origin.hostname:
        errors.append('PUBLIC_ORIGIN hostname must equal PUBLIC_HOST')
    for name in ('R2_ENDPOINT', 'R2_PUBLIC_ENDPOINT'):
        endpoint = urlparse(values.get(name, ''))
        host = endpoint.hostname or ''
        if (endpoint.scheme != 'https' or not host.endswith('.r2.cloudflarestorage.com')
                or endpoint.path not in ('', '/') or endpoint.query or endpoint.fragment):
            errors.append(f'{name} must be an HTTPS Cloudflare R2 S3 API endpoint')
    bucket = values.get('R2_BUCKET', '')
    if bucket and not re.fullmatch(r'[a-z0-9][a-z0-9.-]{1,61}[a-z0-9]', bucket):
        errors.append('R2_BUCKET must be a valid S3 bucket name')
    for name in ('PAYMENT_ADAPTER', 'FREIGHT_ADAPTER', 'TINY_ADAPTER', 'WHATSAPP_ADAPTER'):
        if values.get(name, '').lower() in {'', 'fake', 'mock', 'local'}:
            errors.append(f'{name} must select an approved non-fake implementation')
    for name in REQUIRED:
        if name.endswith('_SECRET') and values.get(name) and not NAME.fullmatch(values[name]):
            errors.append(f'{name} must name a pre-provisioned external Swarm secret')
    secret_names = [values.get(name, '') for name in REQUIRED if name.endswith('_SECRET')]
    populated_secret_names = [name for name in secret_names if name]
    if len(populated_secret_names) != len(set(populated_secret_names)):
        errors.append('Every production secret setting must use a distinct external Swarm secret')
    if values.get('POSTGRES_USER') == values.get('APP_DB_USER'):
        errors.append('Database administrator and runtime user must differ')
    if values.get('PUBLIC_HOST') == values.get('KANBAN_HOST'):
        errors.append('Site and Kanban hosts must differ')
    numeric = {}
    for name in (
            'STORAGE_QUOTA_BYTES', 'OWNER_UPLOAD_QUOTA_BYTES', 'MAX_UPLOAD_BYTES',
            'UPLOAD_PART_BYTES', 'MAX_PENDING_UPLOADS', 'SCAN_MAX_BYTES'):
        try:
            numeric[name] = int(values.get(name, '0'))
            if numeric[name] <= 0:
                raise ValueError
        except ValueError:
            errors.append(f'{name} must be a positive integer')
    if numeric.get('OWNER_UPLOAD_QUOTA_BYTES', 0) > numeric.get('STORAGE_QUOTA_BYTES', 0):
        errors.append('OWNER_UPLOAD_QUOTA_BYTES cannot exceed STORAGE_QUOTA_BYTES')
    if numeric.get('SCAN_MAX_BYTES', 0) > numeric.get('MAX_UPLOAD_BYTES', 0):
        errors.append('SCAN_MAX_BYTES cannot exceed MAX_UPLOAD_BYTES')
    ports = {}
    for name in ('SITE_PORT', 'KANBAN_PORT'):
        try:
            ports[name] = int(values.get(name, '0'))
            if not 1024 <= ports[name] <= 65535:
                raise ValueError
        except ValueError:
            errors.append(f'{name} must be an unprivileged TCP port from 1024 to 65535')
    if ports.get('SITE_PORT') == ports.get('KANBAN_PORT'):
        errors.append('SITE_PORT and KANBAN_PORT must differ')
    return errors


def main(source_only=False):
    errors = source_errors()
    if not source_only:
        errors.extend(config_errors(os.environ))
    if errors:
        print('BLOCKED: production preflight failed:')
        for error in errors:
            print(f'- {error}')
        return 2
    print('PASS: production source and non-secret deployment metadata passed preflight.')
    print('This does not replace staging acceptance, image scanning, or human approval.')
    return 0


if __name__ == '__main__':
    if len(sys.argv) > 2 or (len(sys.argv) == 2 and sys.argv[1] != '--source-only'):
        raise SystemExit('usage: python deploy/production_preflight.py [--source-only]')
    raise SystemExit(main(len(sys.argv) == 2))