Add media retention cleanup command
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 2026-05-27 13:04:47 +02:00
parent 3b6503a6ed
commit 507074b80e
10 changed files with 722 additions and 52 deletions

View file

@ -7,11 +7,21 @@ import os
import signal
import sys
from contextlib import suppress
from pathlib import Path
from hypercorn.asyncio import serve as hypercorn_serve
from hypercorn.config import Config as HypercornConfig
import repub.crawl as crawl_module
from repub.cleanup import DEFAULT_MEDIA_DIRS, cleanup_media
from repub.config import (
AUDIO_DIR,
FILE_DIR,
IMAGE_DIR,
VIDEO_DIR,
build_base_settings,
load_config,
)
from repub.web import SHUTDOWN_EVENT_KEY, create_app
FeedNameFilter = crawl_module.FeedNameFilter
@ -61,11 +71,39 @@ def parse_args(argv: list[str] | None = None) -> tuple[str, argparse.Namespace]:
default="repub.toml",
help="Path to runtime config TOML file",
)
cleanup_parser = subparsers.add_parser(
"cleanup-media",
help="Delete old unreferenced published media",
)
cleanup_parser.add_argument(
"-c",
"--config",
default=None,
help="Read output and media directory settings from runtime config TOML",
)
cleanup_parser.add_argument(
"--feeds-dir",
default=None,
help="Published feeds directory to clean (default: config out_dir/feeds or out/feeds)",
)
cleanup_parser.add_argument(
"--days",
type=int,
default=25,
help="Delete unreferenced media older than this many days",
)
cleanup_parser.add_argument(
"--dry-run",
action="store_true",
help="Report cleanup matches without deleting files",
)
if not raw_args:
raw_args = ["serve", "--dev-mode"]
elif raw_args[0] in {"-c", "--config"}:
raw_args = ["crawl", *raw_args]
elif raw_args[0] not in {"serve", "crawl"}:
elif raw_args[0] not in {"serve", "crawl", "cleanup-media"}:
raw_args = ["serve", "--dev-mode", *raw_args]
args = parser.parse_args(raw_args)
@ -73,6 +111,25 @@ 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, ...]]:
feeds_dir = Path(args.feeds_dir) if args.feeds_dir else Path("out/feeds")
media_dirs = DEFAULT_MEDIA_DIRS
if args.config is None:
return feeds_dir, media_dirs
config = load_config(args.config)
settings = build_base_settings(config)
media_dirs = (
str(settings.get("REPUBLISHER_IMAGE_DIR", IMAGE_DIR)),
str(settings.get("REPUBLISHER_AUDIO_DIR", AUDIO_DIR)),
str(settings.get("REPUBLISHER_VIDEO_DIR", VIDEO_DIR)),
str(settings.get("REPUBLISHER_FILE_DIR", FILE_DIR)),
)
if args.feeds_dir is None:
feeds_dir = config.out_dir / "feeds"
return feeds_dir, media_dirs
def _install_signal_handlers(stop_event: asyncio.Event) -> None:
loop = asyncio.get_running_loop()
@ -116,6 +173,29 @@ def entrypoint(argv: list[str] | None = None) -> int:
crawl_module.check_runtime = check_runtime
return crawl_module.crawl_from_config(args.config)
if command == "cleanup-media":
try:
feeds_dir, media_dirs = _cleanup_config(args)
except FileNotFoundError as error:
missing_path = (
Path(error.filename).expanduser()
if error.filename
else Path(args.config).expanduser()
)
logger.error("Config file not found: %s", missing_path)
return 2
except ValueError as error:
logger.error("Invalid config: %s", error)
return 2
result = cleanup_media(
feeds_dir=feeds_dir,
retention_days=args.days,
dry_run=bool(args.dry_run),
media_dirs=media_dirs,
)
return 1 if result.failures else 0
try:
port = int(args.port)
except ValueError: