diff --git a/swh/fuse/cache.py b/swh/fuse/cache.py
index 5c4e7ec874ee0a656d5048f79bd8bb61e07f6f27..c4299b16e709295e3c2bf8c623972e3c5627d039 100644
--- a/swh/fuse/cache.py
+++ b/swh/fuse/cache.py
@@ -23,7 +23,7 @@ from swh.fuse.fs.artifact import RevisionHistoryShardByDate
from swh.fuse.fs.entry import FuseDirEntry, FuseEntry
from swh.fuse.fs.mountpoint import CacheDir, OriginDir
from swh.model.exceptions import ValidationError
-from swh.model.identifiers import REVISION, SWHID, parse_swhid
+from swh.model.identifiers import CoreSWHID, ObjectType
from swh.web.client.client import ORIGIN_VISIT, typify_json
@@ -79,7 +79,7 @@ class FuseCache:
await self.blob.__aexit__()
await self.history.__aexit__()
- async def get_cached_swhids(self) -> AsyncGenerator[SWHID, None]:
+ async def get_cached_swhids(self) -> AsyncGenerator[CoreSWHID, None]:
""" Return a list of all previously cached SWHID """
# Use the metadata db since it should always contain all accessed SWHIDs
@@ -88,7 +88,7 @@ class FuseCache:
)
swhids = await metadata_cursor.fetchall()
for raw_swhid in swhids:
- yield parse_swhid(raw_swhid[0])
+ yield CoreSWHID.from_string(raw_swhid[0])
async def get_cached_visits(self) -> AsyncGenerator[str, None]:
""" Return a list of all previously cached visit URL """
@@ -149,14 +149,18 @@ class MetadataCache(AbstractCache):
);
"""
- async def get(self, swhid: SWHID, typify: bool = True) -> Any:
+ async def get(self, swhid: CoreSWHID, typify: bool = True) -> Any:
cursor = await self.conn.execute(
"select metadata from metadata_cache where swhid=?", (str(swhid),)
)
cache = await cursor.fetchone()
if cache:
metadata = json.loads(cache[0])
- return typify_json(metadata, swhid.object_type) if typify else metadata
+ return (
+ typify_json(metadata, swhid.object_type.name.lower())
+ if typify
+ else metadata
+ )
else:
return None
@@ -178,10 +182,10 @@ class MetadataCache(AbstractCache):
else:
return None
- async def set(self, swhid: SWHID, metadata: Any) -> None:
+ async def set(self, swhid: CoreSWHID, metadata: Any) -> None:
# Fill in the date column for revisions (used as cache for history/by-date/)
swhid_date = ""
- if swhid.object_type == REVISION:
+ if swhid.object_type == ObjectType.REVISION:
date = dateutil.parser.parse(metadata["date"])
swhid_date = RevisionHistoryShardByDate.DATE_FMT.format(
year=date.year, month=date.month, day=date.day
@@ -200,7 +204,7 @@ class MetadataCache(AbstractCache):
)
await self.conn.commit()
- async def remove(self, swhid: SWHID) -> None:
+ async def remove(self, swhid: CoreSWHID) -> None:
await self.conn.execute(
"delete from metadata_cache where swhid=?", (str(swhid),),
)
@@ -223,7 +227,7 @@ class BlobCache(AbstractCache):
);
"""
- async def get(self, swhid: SWHID) -> Optional[bytes]:
+ async def get(self, swhid: CoreSWHID) -> Optional[bytes]:
cursor = await self.conn.execute(
"select blob from blob_cache where swhid=?", (str(swhid),)
)
@@ -234,13 +238,13 @@ class BlobCache(AbstractCache):
else:
return None
- async def set(self, swhid: SWHID, blob: bytes) -> None:
+ async def set(self, swhid: CoreSWHID, blob: bytes) -> None:
await self.conn.execute(
"insert into blob_cache values (?, ?)", (str(swhid), blob)
)
await self.conn.commit()
- async def remove(self, swhid: SWHID) -> None:
+ async def remove(self, swhid: CoreSWHID) -> None:
await self.conn.execute(
"delete from blob_cache where swhid=?", (str(swhid),),
)
@@ -277,7 +281,7 @@ class HistoryCache(AbstractCache):
select * from dfs limit -1 offset 1
"""
- async def get(self, swhid: SWHID) -> Optional[List[SWHID]]:
+ async def get(self, swhid: CoreSWHID) -> Optional[List[CoreSWHID]]:
cursor = await self.conn.execute(self.HISTORY_REC_QUERY, (str(swhid),),)
cache = await cursor.fetchall()
if not cache:
@@ -286,14 +290,14 @@ class HistoryCache(AbstractCache):
for row in cache:
parent = row[0]
try:
- history.append(parse_swhid(parent))
+ history.append(CoreSWHID.from_string(parent))
except ValidationError:
logging.warning("Cannot parse object from history cache: %s", parent)
return history
async def get_with_date_prefix(
- self, swhid: SWHID, date_prefix: str
- ) -> List[Tuple[SWHID, str]]:
+ self, swhid: CoreSWHID, date_prefix: str
+ ) -> List[Tuple[CoreSWHID, str]]:
cursor = await self.conn.execute(
f"""
select swhid, date from ( {self.HISTORY_REC_QUERY} ) as history
@@ -310,7 +314,7 @@ class HistoryCache(AbstractCache):
for row in cache:
parent, date = row[0], row[1]
try:
- history.append((parse_swhid(parent), date))
+ history.append((CoreSWHID.from_string(parent), date))
except ValidationError:
logging.warning("Cannot parse object from history cache: %s", parent)
return history
diff --git a/swh/fuse/cli.py b/swh/fuse/cli.py
index 43b111e42b89a318f18f8c71fe8cfc4f69a9f763..8f5bf38b94cabb413b65f13a4c56071623cb0c8c 100644
--- a/swh/fuse/cli.py
+++ b/swh/fuse/cli.py
@@ -13,7 +13,7 @@ import click
from swh.core.cli import CONTEXT_SETTINGS
from swh.core.cli import swh as swh_cli_group
-from swh.model.cli import SWHIDParamType
+from swh.model.cli import CoreSWHIDParamType
# All generic config code should reside in swh.core.config
DEFAULT_CONFIG_PATH = os.environ.get(
@@ -97,7 +97,7 @@ def fuse(ctx, config_file):
metavar="PATH",
type=click.Path(exists=True, dir_okay=True, file_okay=False),
)
-@click.argument("swhids", nargs=-1, metavar="[SWHID]...", type=SWHIDParamType())
+@click.argument("swhids", nargs=-1, metavar="[SWHID]...", type=CoreSWHIDParamType())
@click.option(
"-f/-d",
"--foreground/--daemon",
diff --git a/swh/fuse/fs/artifact.py b/swh/fuse/fs/artifact.py
index 85651c1b0cabcd87233a3768b7a659b25e05f04d..78ddf3d3ee7aed5ef997812c174f17b6934c2443 100644
--- a/swh/fuse/fs/artifact.py
+++ b/swh/fuse/fs/artifact.py
@@ -20,7 +20,8 @@ from swh.fuse.fs.entry import (
FuseSymlinkEntry,
)
from swh.model.from_disk import DentryPerms
-from swh.model.identifiers import CONTENT, DIRECTORY, RELEASE, REVISION, SNAPSHOT, SWHID
+from swh.model.hashutil import hash_to_bytes, hash_to_hex
+from swh.model.identifiers import CoreSWHID, ObjectType
SWHID_REGEXP = r"swh:1:(cnt|dir|rel|rev|snp):[0-9a-f]{40}"
@@ -41,7 +42,7 @@ class Content(FuseFileEntry):
directory, the permissions of the `archive/SWHID` file will be arbitrary and
not meaningful (e.g., `0x644`). """
- swhid: SWHID
+ swhid: CoreSWHID
prefetch: Any = None
async def get_content(self) -> bytes:
@@ -73,7 +74,7 @@ class Directory(FuseDirEntry):
So it is possible that, in the context of a directory, a file is presented
as writable, whereas actually writing to it will fail with `EPERM`. """
- swhid: SWHID
+ swhid: CoreSWHID
async def compute_entries(self) -> AsyncIterator[FuseEntry]:
metadata = await self.fuse.get_metadata(self.swhid)
@@ -84,7 +85,7 @@ class Directory(FuseDirEntry):
# Archived permissions for directories are always set to
# 0o040000 so use a read-only permission instead
int(EntryMode.RDONLY_DIR)
- if swhid.object_type == DIRECTORY
+ if swhid.object_type == ObjectType.DIRECTORY
else entry["perms"]
)
@@ -101,7 +102,7 @@ class Directory(FuseDirEntry):
FuseSymlinkEntry, name=name, target=target,
)
# 2. Regular file
- elif swhid.object_type == CONTENT:
+ elif swhid.object_type == ObjectType.CONTENT:
yield self.create_child(
Content,
name=name,
@@ -112,12 +113,12 @@ class Directory(FuseDirEntry):
prefetch=entry,
)
# 3. Regular directory
- elif swhid.object_type == DIRECTORY:
+ elif swhid.object_type == ObjectType.DIRECTORY:
yield self.create_child(
Directory, name=name, mode=mode, swhid=swhid,
)
# 4. Submodule
- elif swhid.object_type == REVISION:
+ elif swhid.object_type == ObjectType.REVISION:
try:
# Make sure the revision metadata is fetched and create a
# symlink to distinguish it with regular directories
@@ -159,7 +160,7 @@ class Revision(FuseDirEntry):
- `meta.json`: metadata for the current node, as a symlink pointing to the
relevant `archive/<SWHID>.json` file """
- swhid: SWHID
+ swhid: CoreSWHID
async def compute_entries(self) -> AsyncIterator[FuseEntry]:
metadata = await self.fuse.get_metadata(self.swhid)
@@ -202,7 +203,7 @@ class Revision(FuseDirEntry):
class RevisionParents(FuseDirEntry):
""" Revision virtual `parents/` directory """
- parents: List[SWHID]
+ parents: List[CoreSWHID]
async def compute_entries(self) -> AsyncIterator[FuseEntry]:
root_path = self.get_relative_root_path()
@@ -218,7 +219,7 @@ class RevisionParents(FuseDirEntry):
class RevisionHistory(FuseDirEntry):
""" Revision virtual `history/` directory """
- swhid: SWHID
+ swhid: CoreSWHID
async def prefill_by_date_cache(self, by_date_dir: FuseDirEntry) -> None:
history = await self.fuse.get_history(self.swhid)
@@ -272,7 +273,7 @@ class RevisionHistory(FuseDirEntry):
class RevisionHistoryShardByDate(FuseDirEntry):
""" Revision virtual `history/by-date` sharded directory """
- history_swhid: SWHID
+ history_swhid: CoreSWHID
prefix: str = field(default="")
is_status_done: bool = field(default=False)
@@ -285,7 +286,7 @@ class RevisionHistoryShardByDate(FuseDirEntry):
name: str = field(init=False, default=".status")
mode: int = field(init=False, default=int(EntryMode.RDONLY_FILE))
- history_swhid: SWHID
+ history_swhid: CoreSWHID
def __post_init__(self):
super().__post_init__()
@@ -355,7 +356,7 @@ class RevisionHistoryShardByDate(FuseDirEntry):
class RevisionHistoryShardByHash(FuseDirEntry):
""" Revision virtual `history/by-hash` sharded directory """
- history_swhid: SWHID
+ history_swhid: CoreSWHID
prefix: str = field(default="")
SHARDING_LENGTH = 2
@@ -367,7 +368,7 @@ class RevisionHistoryShardByHash(FuseDirEntry):
if self.prefix:
root_path = self.get_relative_root_path()
for swhid in history:
- if swhid.object_id.startswith(self.prefix):
+ if swhid.object_id.startswith(hash_to_bytes(self.prefix)):
yield self.create_child(
FuseSymlinkEntry,
name=str(swhid),
@@ -377,7 +378,7 @@ class RevisionHistoryShardByHash(FuseDirEntry):
else:
sharded_dirs = set()
for swhid in history:
- next_prefix = swhid.object_id[: self.SHARDING_LENGTH]
+ next_prefix = hash_to_hex(swhid.object_id)[: self.SHARDING_LENGTH]
if next_prefix not in sharded_dirs:
sharded_dirs.add(next_prefix)
yield self.create_child(
@@ -393,7 +394,7 @@ class RevisionHistoryShardByHash(FuseDirEntry):
class RevisionHistoryShardByPage(FuseDirEntry):
""" Revision virtual `history/by-page` sharded directory """
- history_swhid: SWHID
+ history_swhid: CoreSWHID
prefix: Optional[int] = field(default=None)
PAGE_SIZE = 10_000
@@ -445,16 +446,16 @@ class Release(FuseDirEntry):
- `meta.json`: metadata for the current node, as a symlink pointing to the
relevant `archive/<SWHID>.json` file """
- swhid: SWHID
+ swhid: CoreSWHID
- async def find_root_directory(self, swhid: SWHID) -> Optional[SWHID]:
- if swhid.object_type == RELEASE:
+ async def find_root_directory(self, swhid: CoreSWHID) -> Optional[CoreSWHID]:
+ if swhid.object_type == ObjectType.RELEASE:
metadata = await self.fuse.get_metadata(swhid)
return await self.find_root_directory(metadata["target"])
- elif swhid.object_type == REVISION:
+ elif swhid.object_type == ObjectType.REVISION:
metadata = await self.fuse.get_metadata(swhid)
return metadata["directory"]
- elif swhid.object_type == DIRECTORY:
+ elif swhid.object_type == ObjectType.DIRECTORY:
return swhid
else:
return None
@@ -493,10 +494,10 @@ class Release(FuseDirEntry):
class ReleaseType(FuseFileEntry):
""" Release type virtual file """
- target_type: str
+ target_type: ObjectType
async def get_content(self) -> bytes:
- return str.encode(self.target_type + "\n")
+ return str.encode(self.target_type.name.lower() + "\n")
@dataclass
@@ -512,7 +513,7 @@ class Snapshot(FuseDirEntry):
``tags`` directory containing a ``v1.0`` symlink pointing to the branch
target SWHID. """
- swhid: SWHID
+ swhid: CoreSWHID
prefix: str = field(default="")
async def compute_entries(self) -> AsyncIterator[FuseEntry]:
@@ -627,9 +628,9 @@ class OriginVisit(FuseDirEntry):
OBJTYPE_GETTERS = {
- CONTENT: Content,
- DIRECTORY: Directory,
- REVISION: Revision,
- RELEASE: Release,
- SNAPSHOT: Snapshot,
+ ObjectType.CONTENT: Content,
+ ObjectType.DIRECTORY: Directory,
+ ObjectType.REVISION: Revision,
+ ObjectType.RELEASE: Release,
+ ObjectType.SNAPSHOT: Snapshot,
}
diff --git a/swh/fuse/fs/mountpoint.py b/swh/fuse/fs/mountpoint.py
index 9c302d88ba560a1ffadd65d8ea67a6a7e8ed4807..a894f97f41b5337b2e3d6c2eb55cc38a06342419 100644
--- a/swh/fuse/fs/mountpoint.py
+++ b/swh/fuse/fs/mountpoint.py
@@ -18,7 +18,8 @@ from swh.fuse.fs.entry import (
FuseSymlinkEntry,
)
from swh.model.exceptions import ValidationError
-from swh.model.identifiers import CONTENT, SWHID, parse_swhid
+from swh.model.hashutil import hash_to_hex
+from swh.model.identifiers import CoreSWHID, ObjectType
JSON_SUFFIX = ".json"
@@ -60,7 +61,7 @@ class ArchiveDir(FuseDirEntry):
# On the fly mounting of a new artifact
try:
if name.endswith(JSON_SUFFIX):
- swhid = parse_swhid(name[: -len(JSON_SUFFIX)])
+ swhid = CoreSWHID.from_string(name[: -len(JSON_SUFFIX)])
return self.create_child(
MetaEntry,
name=f"{swhid}{JSON_SUFFIX}",
@@ -68,14 +69,14 @@ class ArchiveDir(FuseDirEntry):
swhid=swhid,
)
else:
- swhid = parse_swhid(name)
+ swhid = CoreSWHID.from_string(name)
await self.fuse.get_metadata(swhid)
return self.create_child(
OBJTYPE_GETTERS[swhid.object_type],
name=str(swhid),
mode=int(
EntryMode.RDONLY_FILE
- if swhid.object_type == CONTENT
+ if swhid.object_type == ObjectType.CONTENT
else EntryMode.RDONLY_DIR
),
swhid=swhid,
@@ -89,7 +90,7 @@ class MetaEntry(FuseFileEntry):
""" An entry for a `archive/<SWHID>.json` file, containing all the SWHID's
metadata from the Software Heritage archive. """
- swhid: SWHID
+ swhid: CoreSWHID
async def get_content(self) -> bytes:
# Make sure the metadata is in cache
@@ -162,7 +163,7 @@ class CacheDir(FuseDirEntry):
async def compute_entries(self) -> AsyncIterator[FuseEntry]:
root_path = self.get_relative_root_path()
async for swhid in self.fuse.cache.get_cached_swhids():
- if not swhid.object_id.startswith(self.prefix):
+ if not hash_to_hex(swhid.object_id).startswith(self.prefix):
continue
yield self.create_child(
@@ -180,7 +181,7 @@ class CacheDir(FuseDirEntry):
try:
if name.endswith(JSON_SUFFIX):
name = name[: -len(JSON_SUFFIX)]
- swhid = parse_swhid(name)
+ swhid = CoreSWHID.from_string(name)
await self.fuse.cache.metadata.remove(swhid)
await self.fuse.cache.blob.remove(swhid)
except ValidationError:
@@ -189,7 +190,7 @@ class CacheDir(FuseDirEntry):
async def compute_entries(self) -> AsyncIterator[FuseEntry]:
prefixes = set()
async for swhid in self.fuse.cache.get_cached_swhids():
- prefixes.add(swhid.object_id[:2])
+ prefixes.add(hash_to_hex(swhid.object_id)[:2])
for prefix in prefixes:
yield self.create_child(
diff --git a/swh/fuse/fuse.py b/swh/fuse/fuse.py
index d18fa3bfe6a7bff1b6471555ffd264e338de57d7..ec27fb96a030021b8f0d995fe730b23a49370821 100644
--- a/swh/fuse/fuse.py
+++ b/swh/fuse/fuse.py
@@ -21,7 +21,7 @@ from swh.fuse import LOGGER_NAME
from swh.fuse.cache import FuseCache
from swh.fuse.fs.entry import FuseDirEntry, FuseEntry, FuseFileEntry, FuseSymlinkEntry
from swh.fuse.fs.mountpoint import Root
-from swh.model.identifiers import CONTENT, REVISION, SWHID
+from swh.model.identifiers import CoreSWHID, ObjectType
from swh.web.client.client import WebAPIClient
@@ -81,7 +81,7 @@ class Fuse(pyfuse3.Operations):
except KeyError:
raise pyfuse3.FUSEError(errno.ENOENT)
- async def get_metadata(self, swhid: SWHID) -> Any:
+ async def get_metadata(self, swhid: CoreSWHID) -> Any:
""" Retrieve metadata for a given SWHID using Software Heritage API """
cache = await self.cache.metadata.get(swhid)
@@ -100,11 +100,11 @@ class Fuse(pyfuse3.Operations):
self.logger.error("Cannot fetch metadata for object %s: %s", swhid, err)
raise
- async def get_blob(self, swhid: SWHID) -> bytes:
+ async def get_blob(self, swhid: CoreSWHID) -> bytes:
""" Retrieve the blob bytes for a given content SWHID using Software
Heritage API """
- if swhid.object_type != CONTENT:
+ if swhid.object_type != ObjectType.CONTENT:
raise pyfuse3.FUSEError(errno.EINVAL)
# Make sure the metadata cache is also populated with the given SWHID
@@ -126,10 +126,10 @@ class Fuse(pyfuse3.Operations):
self.logger.error("Cannot fetch blob for object %s: %s", swhid, err)
raise
- async def get_history(self, swhid: SWHID) -> List[SWHID]:
+ async def get_history(self, swhid: CoreSWHID) -> List[CoreSWHID]:
""" Retrieve a revision's history using Software Heritage Graph API """
- if swhid.object_type != REVISION:
+ if swhid.object_type != ObjectType.REVISION:
raise pyfuse3.FUSEError(errno.EINVAL)
cache = await self.cache.history.get(swhid)
@@ -332,7 +332,7 @@ class Fuse(pyfuse3.Operations):
raise pyfuse3.FUSEError(errno.ENOENT)
-async def main(swhids: List[SWHID], root_path: Path, conf: Dict[str, Any]) -> None:
+async def main(swhids: List[CoreSWHID], root_path: Path, conf: Dict[str, Any]) -> None:
""" swh-fuse CLI entry-point """
# Use pyfuse3 asyncio layer to match the rest of Software Heritage codebase
diff --git a/swh/fuse/tests/api_url.py b/swh/fuse/tests/api_url.py
index 60c47cc337f17364ee93f34b02d75f16bde823f3..5bcf2ec16decbb1ef7943d3403adabce74a61674 100644
--- a/swh/fuse/tests/api_url.py
+++ b/swh/fuse/tests/api_url.py
@@ -6,42 +6,35 @@
from enum import Enum
from typing import Union
-from swh.model.identifiers import (
- CONTENT,
- DIRECTORY,
- RELEASE,
- REVISION,
- SNAPSHOT,
- SWHID,
- parse_swhid,
-)
+from swh.model.hashutil import hash_to_hex
+from swh.model.identifiers import CoreSWHID, ObjectType
GRAPH_API_REQUEST = Enum("GRAPH_API_REQUEST", "HISTORY")
-def swhid_to_web_url(swhid: Union[SWHID, str], raw: bool = False) -> str:
+def swhid_to_web_url(swhid: Union[CoreSWHID, str], raw: bool = False) -> str:
if isinstance(swhid, str):
- swhid = parse_swhid(swhid)
+ swhid = CoreSWHID.from_string(swhid)
prefix = {
- CONTENT: "content/sha1_git:",
- DIRECTORY: "directory/",
- REVISION: "revision/",
- RELEASE: "release/",
- SNAPSHOT: "snapshot/",
+ ObjectType.CONTENT: "content/sha1_git:",
+ ObjectType.DIRECTORY: "directory/",
+ ObjectType.REVISION: "revision/",
+ ObjectType.RELEASE: "release/",
+ ObjectType.SNAPSHOT: "snapshot/",
}
- url = f"{prefix[swhid.object_type]}{swhid.object_id}/"
+ url = f"{prefix[swhid.object_type]}{hash_to_hex(swhid.object_id)}/"
if raw:
url += "raw/"
return url
def swhid_to_graph_url(
- swhid: Union[SWHID, str], request_type: GRAPH_API_REQUEST
+ swhid: Union[CoreSWHID, str], request_type: GRAPH_API_REQUEST
) -> str:
if isinstance(swhid, str):
- swhid = parse_swhid(swhid)
+ swhid = CoreSWHID.from_string(swhid)
prefix = {
GRAPH_API_REQUEST.HISTORY: "graph/visit/edges/",
diff --git a/swh/fuse/tests/data/gen-api-data.py b/swh/fuse/tests/data/gen-api-data.py
index edc7f6edb3a3dff8df45768ffd9444d59347a822..42f74c670fe72f0a877ab990985f639f6e01b2f4 100755
--- a/swh/fuse/tests/data/gen-api-data.py
+++ b/swh/fuse/tests/data/gen-api-data.py
@@ -23,15 +23,8 @@ from swh.fuse.tests.data.config import (
ORIGIN_URL,
REV_SMALL_HISTORY,
)
-from swh.model.identifiers import (
- CONTENT,
- DIRECTORY,
- RELEASE,
- REVISION,
- SNAPSHOT,
- SWHID,
- parse_swhid,
-)
+from swh.model.hashutil import hash_to_bytes
+from swh.model.identifiers import CoreSWHID, ObjectType
API_URL_real = "https://archive.softwareheritage.org/api/1"
API_URL_test = "https://invalid-test-only.archive.softwareheritage.org/api/1"
@@ -43,7 +36,7 @@ API_TOKEN = ""
MOCK_ARCHIVE: Dict[str, Any] = {}
# Temporary map (swhid -> metadata) to ease data generation
-METADATA: Dict[SWHID, Any] = {}
+METADATA: Dict[CoreSWHID, Any] = {}
def get_from_api(endpoint: str) -> str:
@@ -51,19 +44,8 @@ def get_from_api(endpoint: str) -> str:
return requests.get(f"{API_URL_real}/{endpoint}", headers=headers).text
-def get_short_type(object_type: str) -> str:
- short_type = {
- CONTENT: "cnt",
- DIRECTORY: "dir",
- REVISION: "rev",
- RELEASE: "rel",
- SNAPSHOT: "snp",
- }
- return short_type[object_type]
-
-
def generate_archive_web_api(
- swhid: SWHID, raw: bool = False, recursive: bool = False
+ swhid: CoreSWHID, raw: bool = False, recursive: bool = False
) -> None:
# Already in mock archive
if swhid in METADATA and not raw:
@@ -81,17 +63,20 @@ def generate_archive_web_api(
# Retrieve additional needed data for different artifacts (eg: content's
# blob data, release target, etc.)
if recursive:
- if swhid.object_type == CONTENT:
+ if swhid.object_type == ObjectType.CONTENT:
generate_archive_web_api(swhid, raw=True)
- elif swhid.object_type == RELEASE:
+ elif swhid.object_type == ObjectType.RELEASE:
target_type = METADATA[swhid]["target_type"]
target_id = METADATA[swhid]["target"]
- target = parse_swhid(f"swh:1:{get_short_type(target_type)}:{target_id}")
+ target = CoreSWHID(
+ object_type=ObjectType[target_type.upper()],
+ object_id=hash_to_bytes(target_id),
+ )
generate_archive_web_api(target, recursive=True)
-def generate_archive_graph_api(swhid: SWHID) -> None:
- if swhid.object_type == REVISION:
+def generate_archive_graph_api(swhid: CoreSWHID) -> None:
+ if swhid.object_type == ObjectType.REVISION:
# Empty history for all revisions (except REV_SMALL_HISTORY used in tests)
url = swhid_to_graph_url(swhid, GRAPH_API_REQUEST.HISTORY)
MOCK_ARCHIVE[url] = ""
@@ -144,7 +129,7 @@ swh:1:rev:d6b7c96c3eb29b9244ece0c046d3f372ff432d04 swh:1:rev:c01efc669f09508b55e
hist_nodes = set(
map(
- parse_swhid,
+ CoreSWHID.from_string,
[edge.split(" ")[1] for edge in history.strip().split("\n")],
)
)
@@ -163,7 +148,7 @@ def generate_origin_archive_web_api(url: str):
for entry in ALL_ENTRIES:
- swhid = parse_swhid(entry)
+ swhid = CoreSWHID.from_string(entry)
generate_archive_web_api(swhid, recursive=True)
generate_archive_graph_api(swhid)
diff --git a/swh/fuse/tests/test_cache.py b/swh/fuse/tests/test_cache.py
index b230709b1fe8226a8fe78c55fa434dea47ed6482..c650fe7ad35b979836720ef0bd74d4dd103e8518 100644
--- a/swh/fuse/tests/test_cache.py
+++ b/swh/fuse/tests/test_cache.py
@@ -6,7 +6,8 @@
import os
from swh.fuse.tests.data.config import REGULAR_FILE
-from swh.model.identifiers import parse_swhid
+from swh.model.hashutil import hash_to_hex
+from swh.model.identifiers import CoreSWHID
def test_cache_artifact(fuse_mntdir):
@@ -14,8 +15,11 @@ def test_cache_artifact(fuse_mntdir):
(fuse_mntdir / "archive" / REGULAR_FILE).is_file()
- swhid = parse_swhid(REGULAR_FILE)
- assert os.listdir(fuse_mntdir / "cache") == [swhid.object_id[:2], "origin"]
+ swhid = CoreSWHID.from_string(REGULAR_FILE)
+ assert os.listdir(fuse_mntdir / "cache") == [
+ hash_to_hex(swhid.object_id)[:2],
+ "origin",
+ ]
def test_purge_artifact(fuse_mntdir):
@@ -27,7 +31,7 @@ def test_purge_artifact(fuse_mntdir):
(fuse_mntdir / "archive" / REGULAR_FILE).is_file()
assert os.listdir(fuse_mntdir / "cache") != DEFAULT_CACHE_CONTENT
# ... and remove it from cache
- swhid = parse_swhid(REGULAR_FILE)
- os.unlink(fuse_mntdir / "cache" / swhid.object_id[:2] / str(swhid))
+ swhid = CoreSWHID.from_string(REGULAR_FILE)
+ os.unlink(fuse_mntdir / "cache" / hash_to_hex(swhid.object_id)[:2] / str(swhid))
assert os.listdir(fuse_mntdir / "cache") == DEFAULT_CACHE_CONTENT
diff --git a/swh/fuse/tests/test_mountpoint.py b/swh/fuse/tests/test_mountpoint.py
index 09c0e7532c6c14421d3f27b89bbe5bf5092da5ce..3a259d2b8b6eca702b0ce6e406f902a2f306dac8 100644
--- a/swh/fuse/tests/test_mountpoint.py
+++ b/swh/fuse/tests/test_mountpoint.py
@@ -6,7 +6,7 @@
import os
from swh.fuse.tests.data.config import ORIGIN_URL_ENCODED, REGULAR_FILE
-from swh.model.identifiers import parse_swhid
+from swh.model.identifiers import CoreSWHID
def test_mountpoint(fuse_mntdir):
@@ -21,7 +21,7 @@ def test_on_the_fly_mounting(fuse_mntdir):
assert os.listdir(fuse_mntdir / "origin") == []
assert (fuse_mntdir / "origin" / ORIGIN_URL_ENCODED).is_dir()
- sharded_dir = parse_swhid(REGULAR_FILE).object_id[:2]
+ sharded_dir = CoreSWHID.from_string(REGULAR_FILE).object_id.hex()[:2]
assert os.listdir(fuse_mntdir / "cache") == [sharded_dir, "origin"]
assert os.listdir(fuse_mntdir / "cache" / sharded_dir) == [
REGULAR_FILE,
diff --git a/swh/fuse/tests/test_revision.py b/swh/fuse/tests/test_revision.py
index d88215162c22eb25d5b1e088c43de2d81b13df1e..b610b886e33d2569f69471fdfa074f78fb83c988 100644
--- a/swh/fuse/tests/test_revision.py
+++ b/swh/fuse/tests/test_revision.py
@@ -12,7 +12,8 @@ from swh.fuse.tests.common import (
get_data_from_web_archive,
)
from swh.fuse.tests.data.config import REV_SMALL_HISTORY, ROOT_DIR, ROOT_REV
-from swh.model.identifiers import parse_swhid
+from swh.model.hashutil import hash_to_hex
+from swh.model.identifiers import CoreSWHID
def test_access_meta(fuse_mntdir):
@@ -53,12 +54,12 @@ def test_list_history(fuse_mntdir):
# Only keep second node in the edge because first node is redundant
# information or the root node (hence not an ancestor)
expected = set(
- map(parse_swhid, [edge.split(" ")[1] for edge in history.split("\n")])
+ map(CoreSWHID.from_string, [edge.split(" ")[1] for edge in history.split("\n")])
)
dir_by_hash = dir_path / "by-hash"
for swhid in expected:
- depth1 = swhid.object_id[:2]
+ depth1 = hash_to_hex(swhid.object_id)[:2]
depth2 = str(swhid)
assert (dir_by_hash / depth1).exists()
assert depth2 in (os.listdir(dir_by_hash / depth1))