From c773989051118788dbe2b305c1c7b33ade5f5fc1 Mon Sep 17 00:00:00 2001
From: Valentin Lorentz <vlorentz@softwareheritage.org>
Date: Fri, 3 Mar 2023 12:15:33 +0100
Subject: [PATCH] Call storage.content_get_data() with hash dicts instead of
only sha1
---
swh/vault/cookers/git_bare.py | 19 +++++++++++++++++--
swh/vault/to_disk.py | 9 ++++++---
2 files changed, 23 insertions(+), 5 deletions(-)
diff --git a/swh/vault/cookers/git_bare.py b/swh/vault/cookers/git_bare.py
index 453fc24..497876c 100644
--- a/swh/vault/cookers/git_bare.py
+++ b/swh/vault/cookers/git_bare.py
@@ -39,7 +39,18 @@ import re
import subprocess
import tarfile
import tempfile
-from typing import Any, Dict, Iterable, Iterator, List, NoReturn, Optional, Set, Tuple
+from typing import (
+ Any,
+ Dict,
+ Iterable,
+ Iterator,
+ List,
+ NoReturn,
+ Optional,
+ Set,
+ Tuple,
+ cast,
+)
import zlib
import sentry_sdk
@@ -63,6 +74,7 @@ from swh.model.model import ObjectType as ModelObjectType
from swh.model.swhids import CoreSWHID, ObjectType
from swh.storage.algos.revisions_walker import DFSRevisionsWalker
from swh.storage.algos.snapshot import snapshot_get_all_branches
+from swh.storage.interface import HashDict
from swh.vault.cookers.base import BaseVaultCooker
from swh.vault.to_disk import HIDDEN_MESSAGE, SKIPPED_MESSAGE
@@ -690,7 +702,10 @@ class GitBareCooker(BaseVaultCooker):
contents_and_data: Iterator[Tuple[Content, Optional[bytes]]]
if self.objstorage is None:
contents_and_data = (
- (content, self.storage.content_get_data(content.sha1))
+ (
+ content,
+ self.storage.content_get_data(cast(HashDict, content.hashes())),
+ )
for content in visible_contents
)
else:
diff --git a/swh/vault/to_disk.py b/swh/vault/to_disk.py
index 2721642..b2b1650 100644
--- a/swh/vault/to_disk.py
+++ b/swh/vault/to_disk.py
@@ -11,7 +11,7 @@ from typing import Any, Dict, Iterator, List
from swh.model import hashutil
from swh.model.from_disk import DentryPerms, mode_to_perms
from swh.storage.algos.dir_iterators import dir_iterator
-from swh.storage.interface import StorageInterface
+from swh.storage.interface import HashDict, StorageInterface
MISSING_MESSAGE = (
b"This content is missing from the Software Heritage archive "
@@ -48,8 +48,11 @@ def get_filtered_files_content(
for file_data in files_data:
status = file_data["status"]
if status == "visible":
- sha1 = file_data["sha1"]
- data = storage.content_get_data(sha1)
+ hashes: HashDict = {
+ "sha1": file_data["sha1"],
+ "sha1_git": file_data["sha1_git"],
+ }
+ data = storage.content_get_data(hashes)
if data is None:
content = SKIPPED_MESSAGE
else:
--
GitLab