Files
artifactapi/tests/test_cache.py
T
unkinben 8fe4bac2b9 feat: add check_mutable_updates flag for conditional upstream revalidation
When check_mutable_updates: true is set on a remote, expired user-defined
mutable files are revalidated before re-downloading:

- On expiry a conditional HEAD is sent with If-None-Match / If-Modified-Since
- 304 Not Modified: TTL is refreshed in Redis, S3 cache is untouched
- 200 / no conditional support: cache is invalidated and file re-downloaded
- Network error: safe fallback — assume changed, re-download

ETag and Last-Modified from upstream responses are stored in Redis under
mutable:meta:<remote>:<hash> (no expiry, cleaned up on re-download or
cache flush). The flag only applies to user-configured mutable_patterns;
built-in package-type defaults (APKINDEX, repomd.xml, Docker manifests)
are always re-fetched unconditionally.

cache/flush also clears mutable:meta:* keys alongside index:* keys.
2026-04-27 11:00:09 +10:00

286 lines
13 KiB
Python

"""Tests for RedisCache, focusing on is_mutable_file with configurable patterns."""
import hashlib
from unittest.mock import ANY, MagicMock, patch
import pytest
from artifactapi.cache import RedisCache
from artifactapi.config import _PACKAGE_MUTABLE_PATTERNS
@pytest.fixture
def bare_cache():
"""RedisCache instance bypassing __init__ (no Redis needed for pure-logic tests)."""
return RedisCache.__new__(RedisCache)
@pytest.fixture
def unavailable_cache():
"""RedisCache where Redis is not reachable."""
with patch("redis.from_url", side_effect=Exception("connection refused")):
return RedisCache("redis://localhost:6379/0")
@pytest.fixture
def mock_redis_client():
return MagicMock()
@pytest.fixture
def cache_with_redis(mock_redis_client):
"""RedisCache backed by a MagicMock Redis client."""
with patch("redis.from_url", return_value=mock_redis_client):
c = RedisCache("redis://localhost:6379/0")
c.client = mock_redis_client
c.available = True
return c
# ---------------------------------------------------------------------------
# is_mutable_file — alpine patterns
# ---------------------------------------------------------------------------
class TestIsMutableFileAlpine:
def test_apkindex_tarball_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["alpine"]
assert bare_cache.is_mutable_file("alpine/v3.18/x86_64/APKINDEX.tar.gz", patterns)
def test_nested_apkindex_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["alpine"]
assert bare_cache.is_mutable_file("mirrors/dl-cdn/alpine/v3.19/community/x86_64/APKINDEX.tar.gz", patterns)
def test_apk_package_is_not_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["alpine"]
assert not bare_cache.is_mutable_file("alpine/v3.18/x86_64/musl-1.2.4-r2.apk", patterns)
def test_random_tarball_is_not_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["alpine"]
assert not bare_cache.is_mutable_file("some/path/archive.tar.gz", patterns)
def test_apkindex_signature_file_is_not_index(self, bare_cache):
# Signature file adjacent to the index should not be treated as an index
patterns = _PACKAGE_MUTABLE_PATTERNS["alpine"]
assert not bare_cache.is_mutable_file("alpine/v3.18/x86_64/APKINDEX.tar.gz.sig", patterns)
def test_apkindex_tmp_file_is_not_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["alpine"]
assert not bare_cache.is_mutable_file("alpine/v3.18/x86_64/APKINDEX.tar.gz.tmp", patterns)
# ---------------------------------------------------------------------------
# is_mutable_file — rpm patterns
# ---------------------------------------------------------------------------
class TestIsMutableFileRpm:
def test_repomd_xml_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["rpm"]
assert bare_cache.is_mutable_file("almalinux/9/x86_64/repomd.xml", patterns)
def test_repodata_primary_xml_gz_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["rpm"]
assert bare_cache.is_mutable_file("repo/repodata/primary.xml.gz", patterns)
def test_repodata_sqlite_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["rpm"]
assert bare_cache.is_mutable_file("repo/repodata/primary.sqlite", patterns)
def test_repodata_sqlite_bz2_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["rpm"]
assert bare_cache.is_mutable_file("repo/repodata/other.sqlite.bz2", patterns)
def test_repodata_yaml_xz_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["rpm"]
assert bare_cache.is_mutable_file("repo/repodata/comps.yaml.xz", patterns)
def test_packages_gz_pattern_matches_any_path(self, bare_cache):
# The Packages.gz$ regex is a carryover from the original hardcoded logic and
# deliberately matches any path ending in Packages.gz — including Debian-style paths.
# This test documents that intentional behaviour.
patterns = _PACKAGE_MUTABLE_PATTERNS["rpm"]
assert bare_cache.is_mutable_file("debian/dists/stable/main/binary-amd64/Packages.gz", patterns)
def test_rpm_package_is_not_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["rpm"]
assert not bare_cache.is_mutable_file("almalinux/9/x86_64/Packages/bash-5.1.8.x86_64.rpm", patterns)
def test_arbitrary_xml_outside_repodata_is_not_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["rpm"]
assert not bare_cache.is_mutable_file("some/path/config.xml", patterns)
# ---------------------------------------------------------------------------
# is_mutable_file — docker patterns
# ---------------------------------------------------------------------------
class TestIsMutableFileDocker:
def test_tag_manifest_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["docker"]
assert bare_cache.is_mutable_file("library/nginx/manifests/latest", patterns)
def test_version_tag_manifest_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["docker"]
assert bare_cache.is_mutable_file("library/nginx/manifests/1.25.3", patterns)
def test_hyphenated_tag_manifest_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["docker"]
assert bare_cache.is_mutable_file("library/nginx/manifests/latest-rc", patterns)
def test_numeric_date_tag_manifest_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["docker"]
assert bare_cache.is_mutable_file("library/nginx/manifests/20240101", patterns)
def test_digest_manifest_is_not_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["docker"]
digest = "sha256:" + "a" * 64
assert not bare_cache.is_mutable_file(f"library/nginx/manifests/{digest}", patterns)
def test_tags_list_is_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["docker"]
assert bare_cache.is_mutable_file("library/nginx/tags/list", patterns)
def test_blob_is_not_index(self, bare_cache):
patterns = _PACKAGE_MUTABLE_PATTERNS["docker"]
assert not bare_cache.is_mutable_file("library/nginx/blobs/sha256:abc123", patterns)
# ---------------------------------------------------------------------------
# is_mutable_file — edge cases
# ---------------------------------------------------------------------------
class TestIsMutableFileEdgeCases:
def test_empty_patterns_nothing_is_index(self, bare_cache):
assert not bare_cache.is_mutable_file("APKINDEX.tar.gz", [])
assert not bare_cache.is_mutable_file("repomd.xml", [])
assert not bare_cache.is_mutable_file("library/nginx/manifests/latest", [])
def test_none_patterns_nothing_is_index(self, bare_cache):
assert not bare_cache.is_mutable_file("APKINDEX.tar.gz", None)
assert not bare_cache.is_mutable_file("repomd.xml", None)
def test_custom_patterns_match(self, bare_cache):
patterns = [r"metadata\.json$", r"index\.yaml$"]
assert bare_cache.is_mutable_file("repo/metadata.json", patterns)
assert bare_cache.is_mutable_file("repo/subdir/index.yaml", patterns)
assert not bare_cache.is_mutable_file("repo/data.tar.gz", patterns)
def test_custom_pattern_does_not_match_standard_index(self, bare_cache):
patterns = [r"metadata\.json$"]
assert not bare_cache.is_mutable_file("APKINDEX.tar.gz", patterns)
# ---------------------------------------------------------------------------
# get_index_cache_key
# ---------------------------------------------------------------------------
class TestGetIndexCacheKey:
def test_key_format_is_deterministic(self, bare_cache):
# Assert against a pre-computed value to pin the hash algorithm,
# truncation length, and format string in one assertion.
path = "alpine/v3.18/x86_64/APKINDEX.tar.gz"
expected_hash = hashlib.sha256(path.encode()).hexdigest()[:16]
key = bare_cache.get_index_cache_key("alpine-test", path)
assert key == f"index:alpine-test:{expected_hash}"
def test_different_paths_produce_different_keys(self, bare_cache):
k1 = bare_cache.get_index_cache_key("alpine-test", "alpine/v3.18/x86_64/APKINDEX.tar.gz")
k2 = bare_cache.get_index_cache_key("alpine-test", "alpine/v3.19/x86_64/APKINDEX.tar.gz")
assert k1 != k2
def test_different_remotes_produce_different_keys(self, bare_cache):
k1 = bare_cache.get_index_cache_key("remote-a", "path/to/APKINDEX.tar.gz")
k2 = bare_cache.get_index_cache_key("remote-b", "path/to/APKINDEX.tar.gz")
assert k1 != k2
def test_key_starts_with_index_prefix_and_remote(self, bare_cache):
key = bare_cache.get_index_cache_key("myremote", "some/path")
assert key.startswith("index:myremote:")
def test_key_hash_segment_is_16_chars(self, bare_cache):
key = bare_cache.get_index_cache_key("myremote", "some/path/file.xml")
# Format: index:<remote>:<16-char hash> — the fixed length matters for key-space hygiene
parts = key.split(":")
assert len(parts) == 3
assert len(parts[2]) == 16
# ---------------------------------------------------------------------------
# mark_index_cached / is_index_valid
# ---------------------------------------------------------------------------
class TestIndexValidity:
def test_mark_index_cached_calls_setex_with_correct_ttl(self, cache_with_redis, mock_redis_client):
cache_with_redis.mark_index_cached("remote", "path/APKINDEX.tar.gz", 300)
expected_key = cache_with_redis.get_index_cache_key("remote", "path/APKINDEX.tar.gz")
mock_redis_client.setex.assert_called_once_with(expected_key, 300, ANY)
def test_present_key_is_valid(self, cache_with_redis, mock_redis_client):
mock_redis_client.exists.return_value = 1
assert cache_with_redis.is_index_valid("remote", "path/APKINDEX.tar.gz")
def test_missing_key_is_not_valid(self, cache_with_redis, mock_redis_client):
mock_redis_client.exists.return_value = 0
assert not cache_with_redis.is_index_valid("remote", "path/APKINDEX.tar.gz")
def test_unavailable_redis_is_not_valid(self, unavailable_cache):
assert not unavailable_cache.is_index_valid("remote", "some/path")
def test_mark_cached_no_op_when_unavailable(self, unavailable_cache):
# client is None when Redis is unavailable — setex cannot be called
assert unavailable_cache.client is None
unavailable_cache.mark_index_cached("remote", "some/path", 300) # must not raise
# ---------------------------------------------------------------------------
# mutable meta (ETag / Last-Modified storage)
# ---------------------------------------------------------------------------
class TestMutableMeta:
def test_meta_key_format(self, bare_cache):
path = "repo/metadata.json"
expected_hash = hashlib.sha256(path.encode()).hexdigest()[:16]
assert bare_cache.get_mutable_meta_key("myremote", path) == f"mutable:meta:myremote:{expected_hash}"
def test_meta_key_hash_is_16_chars(self, bare_cache):
key = bare_cache.get_mutable_meta_key("remote", "some/path/file.json")
assert len(key.split(":")[-1]) == 16
def test_store_and_retrieve_etag(self, cache_with_redis, mock_redis_client):
mock_redis_client.hgetall.return_value = {"etag": '"abc123"'}
cache_with_redis.store_mutable_meta("remote", "path/meta.json", '"abc123"', None)
mock_redis_client.hset.assert_called_once()
meta = cache_with_redis.get_mutable_meta("remote", "path/meta.json")
assert meta["etag"] == '"abc123"'
def test_store_and_retrieve_last_modified(self, cache_with_redis, mock_redis_client):
lm = "Mon, 01 Jan 2024 00:00:00 GMT"
mock_redis_client.hgetall.return_value = {"last_modified": lm}
cache_with_redis.store_mutable_meta("remote", "path/meta.json", None, lm)
meta = cache_with_redis.get_mutable_meta("remote", "path/meta.json")
assert meta["last_modified"] == lm
def test_store_no_op_when_both_none(self, cache_with_redis, mock_redis_client):
cache_with_redis.store_mutable_meta("remote", "path/meta.json", None, None)
mock_redis_client.hset.assert_not_called()
def test_store_no_op_when_unavailable(self, unavailable_cache):
unavailable_cache.store_mutable_meta("remote", "path", "etag", None) # must not raise
def test_get_returns_empty_when_unavailable(self, unavailable_cache):
assert unavailable_cache.get_mutable_meta("remote", "path") == {}
def test_delete_removes_meta_key(self, cache_with_redis, mock_redis_client):
expected_key = cache_with_redis.get_mutable_meta_key("remote", "path/meta.json")
cache_with_redis.delete_mutable_meta("remote", "path/meta.json")
mock_redis_client.delete.assert_called_once_with(expected_key)
def test_delete_no_op_when_unavailable(self, unavailable_cache):
unavailable_cache.delete_mutable_meta("remote", "path") # must not raise