Compare commits

..
6 Commits
8 changed files with 154 additions and 7 deletions
+6 -2
View File
@@ -11,8 +11,12 @@ CMD ["python", "-m", "unittest", "discover", "-s", "tests", "-v"]
FROM python:3.11-slim
ENV PYTHONDONTWRITEBYTECODE=1 PYTHONUNBUFFERED=1
RUN groupadd --gid 1001 archive-control \
&& useradd --uid 1001 --gid 1001 --no-create-home archive-control
RUN if ! getent group 1001 >/dev/null; then \
groupadd --gid 1001 archive-control; \
fi \
&& if ! getent passwd 1001 >/dev/null; then \
useradd --uid 1001 --gid 1001 --no-create-home archive-control; \
fi
COPY --from=builder /wheels /wheels
RUN pip install --no-cache-dir /wheels/*.whl && rm -rf /wheels
USER 1001:1001
+22
View File
@@ -259,6 +259,28 @@ Generated protobuf Python bindings are committed in each consumer with the
exact `archive-control-proto` tag/commit recorded. Release order is proto,
control consumer, client consumer, E2E, then image publication.
### Reproducible multi-platform Buildx lifecycle
The named Buildx builders are local acceleration/cache only; they are not a
deployment dependency and can be removed after publication. To create a fresh
builder, verify its platforms, publish a release, and remove it afterwards:
```bash
docker buildx create --name archive-control-release --driver docker-container --use
docker buildx inspect --bootstrap
docker buildx build --platform linux/amd64,linux/arm64 \
--tag sodium/archive-clients:vX.Y.Z --push .
docker buildx rm archive-control-release
```
`docker buildx inspect` must list both `linux/amd64` and `linux/arm64` before
publishing. If the host has no arm64 emulation, install/configure it according
to the host Docker distribution before the build; do not publish a partial
single-platform tag. Retain the pushed manifest digest in the release notes
and deploy the immutable tag or digest. The optional `archive-control-qemu`
builder follows the same lifecycle when it is used for an emulation smoke
build.
## Operator usage
1. Prepare local qB, sync, state, backup, config, and secret mounts with the
+1 -1
View File
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
[project]
name = "archive-clients"
version = "0.1.15"
version = "0.1.19"
requires-python = ">=3.11"
dependencies = ["protobuf==7.35.1", "websockets==16.0"]
+11 -1
View File
@@ -168,7 +168,7 @@ def _v1_files(info: dict[bytes, Any]) -> list[MetaFile]:
components = [name] + [_component(part) for part in raw_path]
result.append(MetaFile(
"/".join(components), _length(item.get(b"length")),
_padding(item),
_padding(item) or _bitcomet_padding_name(components[-1]),
))
return result
@@ -223,3 +223,13 @@ def _length(value: Any) -> int:
def _padding(value: dict[bytes, Any]) -> bool:
attributes = value.get(b"attr", b"")
return isinstance(attributes, bytes) and b"p" in attributes
def _bitcomet_padding_name(component: str) -> bool:
"""Recognize BitComet's legacy padding-file convention.
Such v1 torrents often omit the standard ``attr=p`` flag, but
qBittorrent/libtorrent still hides these synthetic entries from its file
API. The exact reserved prefix is the interoperable marker.
"""
return component.startswith("_____padding_file_")
+14 -1
View File
@@ -3,6 +3,7 @@
from __future__ import annotations
import json
import logging
import threading
import time
import uuid
@@ -16,6 +17,8 @@ from urllib import error, parse, request
from archive_clients.config import ServiceConfig
from archive_clients.resources import NormalizedResource, normalize_resource
logger = logging.getLogger(__name__)
class QBittorrentError(RuntimeError):
pass
@@ -72,7 +75,17 @@ class QBittorrentReader:
torrent_hash = torrent.get("hash")
if not isinstance(torrent_hash, str):
raise QBittorrentError("qBittorrent torrent hash is invalid")
result.append(self._normalize(torrent, torrent_hash))
try:
result.append(self._normalize(torrent, torrent_hash))
except ValueError as exc:
# A stale or malformed qBittorrent entry must not hide every
# otherwise valid resource from Archive/Evict inventory.
# Keep it ineligible and leave an operator-visible diagnosis.
logger.warning(
"Skipping qBittorrent resource with inconsistent "
"metadata: hash=%s name=%r reason=%s",
torrent_hash, torrent.get("name"), exc,
)
return result
def get_resource(self, torrent_hash: str) -> NormalizedResource | None:
+11 -2
View File
@@ -91,12 +91,21 @@ def normalize_resource(
observed_at: datetime | None = None,
) -> NormalizedResource:
metainfo = decode_metainfo(metainfo_bytes)
if len(raw_files) != len(metainfo.files):
# qBittorrent/libtorrent may omit torrent padding files from
# /torrents/files while keeping them in the exported metainfo.
visible_metainfo_files = tuple(
item for item in metainfo.files if not item.padding
)
if len(raw_files) == len(metainfo.files):
matched_metainfo_files = metainfo.files
elif len(raw_files) == len(visible_metainfo_files):
matched_metainfo_files = visible_metainfo_files
else:
raise ResourceError("qBittorrent and metainfo file counts differ")
files = []
canonical = True
for expected_index, (raw, meta_file) in enumerate(
zip(raw_files, metainfo.files, strict=True)
zip(raw_files, matched_metainfo_files, strict=True)
):
index = _integer(raw.get("index"), "file index")
if index != expected_index:
+48
View File
@@ -205,6 +205,54 @@ class QBittorrentReaderTests(unittest.TestCase):
self.assertEqual(resources, [])
self.assertEqual(len(opener.calls), 2)
def test_listing_skips_malformed_torrent_and_keeps_valid_resources(self):
def torrent(name, content):
info = {
b"length": len(content), b"name": name.encode(),
b"piece length": 16384, b"pieces": b"x" * 20,
}
return encode({b"info": info}), hashlib.sha1(encode(info)).hexdigest()
first_bytes, first_hash = torrent("first.txt", b"one")
malformed_bytes, malformed_hash = torrent("broken.txt", b"bad")
second_bytes, second_hash = torrent("second.txt", b"two")
torrents = [
{"hash": first_hash, "name": "first.txt", "state": "uploading"},
{"hash": malformed_hash, "name": "broken.txt", "state": "stalledUP"},
{"hash": second_hash, "name": "second.txt", "state": "uploading"},
]
responses = [
b"Ok.", json.dumps(torrents).encode(),
b'[{"index":0,"name":"first.txt","size":3,"progress":1,"priority":1}]',
first_bytes,
b'[{"index":0,"name":"broken.txt","size":3,"progress":1,"priority":1},'
b'{"index":1,"name":"unexpected.txt","size":1,"progress":1,"priority":1}]',
malformed_bytes,
b'[{"index":0,"name":"second.txt","size":3,"progress":1,"priority":1}]',
second_bytes,
]
with tempfile.TemporaryDirectory() as directory:
root = Path(directory)
password = root / "password"
password.write_text("secret", encoding="utf-8")
os.chmod(password, 0o600)
config = ServiceConfig(
"http://qb", PurePosixPath("/downloads"), root,
username="admin", password_file=password,
)
opener = _Opener(responses)
with patch(
"archive_clients.qbittorrent.request.build_opener",
return_value=opener,
), self.assertLogs("archive_clients.qbittorrent", "WARNING") as logs:
resources = QBittorrentReader(config).list_resources()
self.assertEqual(
[resource.summary.display_name for resource in resources],
["first.txt", "second.txt"],
)
self.assertIn(malformed_hash, "\n".join(logs.output))
def test_stopped_add_selection_recheck_and_entry_only_delete(self):
torrent_hash = "a" * 40
responses = [
+41
View File
@@ -118,6 +118,47 @@ class ResourceTests(unittest.TestCase):
self.assertEqual(root.available_file_count, 1)
self.assertEqual(root.available_logical_bytes, 3)
def test_qbittorrent_hidden_padding_files_are_normalized(self):
info = {
b"files": [
{b"length": 3, b"path": [b"first.bin"]},
{
b"length": 5,
b"path": [b"_____padding_file_5"],
},
{b"length": 7, b"path": [b"last.bin"]},
],
b"name": b"with-padding",
b"piece length": 16384,
b"pieces": b"x" * 20,
}
metainfo = encode({b"info": info})
torrent_hash = hashlib.sha1(encode(info)).hexdigest()
normalized = normalize_resource(
{
"hash": torrent_hash,
"name": "with-padding",
"state": "stalledUP",
},
[
{
"index": 0, "name": "with-padding/first.bin",
"size": 3, "completed": 3, "priority": 1,
},
{
"index": 1, "name": "with-padding/last.bin",
"size": 7, "completed": 7, "priority": 1,
},
],
metainfo,
)
self.assertEqual(
[item.canonical_path for item in normalized.files],
["with-padding/first.bin", "with-padding/last.bin"],
)
self.assertEqual(normalized.summary.total_file_count, 2)
self.assertEqual(normalized.summary.selected_complete_bytes, 10)
def test_noncanonical_and_unsafe_paths_are_distinct(self):
info = {
b"length": 3,