Prune old job executions
All checks were successful
buildbot/nix-eval Build done.
buildbot/nix-build Build done.
buildbot/nix-effects Build done.

This commit is contained in:
Abel Luck 2026-06-02 11:31:39 +02:00
parent 813f19f355
commit 710ac76192
6 changed files with 552 additions and 11 deletions

View file

@ -22,6 +22,11 @@ from repub.config import (
build_base_settings,
load_config,
)
from repub.job_retention import (
DEFAULT_SUCCESSFUL_EXECUTION_RETENTION_DAYS,
DEFAULT_UNSUCCESSFUL_EXECUTION_RETENTION_DAYS,
cleanup_job_executions,
)
from repub.web import SHUTDOWN_EVENT_KEY, create_app
FeedNameFilter = crawl_module.FeedNameFilter
@ -97,6 +102,11 @@ def parse_args(argv: list[str] | None = None) -> tuple[str, argparse.Namespace]:
default=None,
help="Published feeds directory to clean (default: config out_dir/feeds or out/feeds)",
)
cleanup_parser.add_argument(
"--log-dir",
default=None,
help="Job execution log directory to clean (default: config out_dir/logs or alongside feeds)",
)
cleanup_parser.add_argument(
"--days",
type=int,
@ -121,11 +131,12 @@ def parse_args(argv: list[str] | None = None) -> tuple[str, argparse.Namespace]:
return command, args
def _cleanup_config(args: argparse.Namespace) -> tuple[Path, tuple[str, ...]]:
def _cleanup_config(args: argparse.Namespace) -> tuple[Path, Path, tuple[str, ...]]:
feeds_dir = Path(args.feeds_dir) if args.feeds_dir else Path("out/feeds")
log_dir = Path(args.log_dir) if args.log_dir else feeds_dir.parent / "logs"
media_dirs = DEFAULT_MEDIA_DIRS
if args.config is None:
return feeds_dir, media_dirs
return feeds_dir, log_dir, media_dirs
config = load_config(args.config)
settings = build_base_settings(config)
@ -137,7 +148,9 @@ def _cleanup_config(args: argparse.Namespace) -> tuple[Path, tuple[str, ...]]:
)
if args.feeds_dir is None:
feeds_dir = config.out_dir / "feeds"
return feeds_dir, media_dirs
if args.log_dir is None:
log_dir = config.out_dir / "logs"
return feeds_dir, log_dir, media_dirs
def _install_signal_handlers(stop_event: asyncio.Event) -> None:
@ -187,7 +200,7 @@ def entrypoint(argv: list[str] | None = None) -> int:
if command == "cleanup-media":
try:
feeds_dir, media_dirs = _cleanup_config(args)
feeds_dir, log_dir, media_dirs = _cleanup_config(args)
except FileNotFoundError as error:
missing_path = (
Path(error.filename).expanduser()
@ -206,7 +219,13 @@ def entrypoint(argv: list[str] | None = None) -> int:
dry_run=bool(args.dry_run),
media_dirs=media_dirs,
)
return 1 if result.failures else 0
job_result = cleanup_job_executions(
log_dir=log_dir,
successful_days=DEFAULT_SUCCESSFUL_EXECUTION_RETENTION_DAYS,
unsuccessful_days=DEFAULT_UNSUCCESSFUL_EXECUTION_RETENTION_DAYS,
dry_run=bool(args.dry_run),
)
return 1 if result.failures or job_result.failures else 0
try:
port = int(args.port)