From c45625f61a9b8374cdcaf7ee65df331cf1e8f07a Mon Sep 17 00:00:00 2001
From: Antoine Lambert <anlambert@softwareheritage.org>
Date: Wed, 21 Feb 2024 11:36:35 +0100
Subject: [PATCH] tests: Ensure nar hash extid matches nar hash of target
Add an extra check in function fetch_extids_from_checksums to ensure
a NAR hash extid matches the NAR hash of the targeted archived object.
Related to swh/infra/sysadm-environment#5256.
---
requirements-test.txt | 1 +
swh/loader/tests/__init__.py | 30 ++++++++++++++++++++++++++++++
2 files changed, 31 insertions(+)
diff --git a/requirements-test.txt b/requirements-test.txt
index f9a1547b..ba4ed4b2 100644
--- a/requirements-test.txt
+++ b/requirements-test.txt
@@ -5,6 +5,7 @@ urllib3
swh-core[testing]
swh-scheduler[testing] >= 2.0.0
swh-storage[testing] >= 2.0.0
+swh-vault
types-click
types-python-dateutil
types-pyyaml
diff --git a/swh/loader/tests/__init__.py b/swh/loader/tests/__init__.py
index 0046cfee..119f3b18 100644
--- a/swh/loader/tests/__init__.py
+++ b/swh/loader/tests/__init__.py
@@ -7,13 +7,17 @@ from collections import defaultdict
import os
from pathlib import PosixPath
import subprocess
+import tempfile
from typing import Dict, Iterable, List, Optional, Tuple, Union
+from swh.loader.core.nar import Nar
from swh.model.hashutil import hash_to_bytes
from swh.model.model import ExtID, OriginVisitStatus, Snapshot, TargetType
+from swh.model.swhids import ObjectType
from swh.storage.algos.origin import origin_get_latest_visit_status
from swh.storage.algos.snapshot import snapshot_get_all_branches
from swh.storage.interface import StorageInterface
+from swh.vault.to_disk import DirectoryBuilder
def assert_last_visit_matches(
@@ -283,6 +287,32 @@ def fetch_extids_from_checksums(
extid = storage.extid_get_from_extid(id_type, ids, extid_version)
extids.extend(extid)
+ for extid_ in extids:
+ if extid_.extid_type.startswith("nar-"):
+ # check NAR hashes of archived directory or content match the expected ones
+ target_swhid = extid_.target
+ with tempfile.TemporaryDirectory() as tmp_dir:
+ nar = Nar(hash_names=list(checksums.keys()))
+ if target_swhid.object_type == ObjectType.DIRECTORY:
+ dir_builder = DirectoryBuilder(
+ storage=storage,
+ root=tmp_dir.encode(),
+ dir_id=target_swhid.object_id,
+ )
+ dir_builder.build()
+ path_to_hash = tmp_dir
+ else:
+ path_to_hash = os.path.join(tmp_dir, "content")
+ content_bytes = storage.content_get_data(
+ {"sha1_git": target_swhid.object_id}
+ )
+ assert content_bytes is not None
+ with open(path_to_hash, "wb") as content:
+ content.write(content_bytes)
+
+ nar.serialize(PosixPath(path_to_hash))
+ assert nar.hexdigest() == checksums
+
return extids
--
GitLab