diff --git a/PKG-INFO b/PKG-INFO
index 0c7e81ca9b76cc80f86bf3d5e8517d914fe3a8db..2b46f3e1f28938822858cc690385b9ea7132e04d 100644
--- a/PKG-INFO
+++ b/PKG-INFO
@@ -1,6 +1,6 @@
Metadata-Version: 2.1
Name: swh.model
-Version: 0.11.0
+Version: 0.12.0
Summary: Software Heritage data model
Home-page: https://forge.softwareheritage.org/diffusion/DMOD/
Author: Software Heritage developers
diff --git a/debian/changelog b/debian/changelog
index c0a91d7be5c22b3af8e8f68ac5aa2e8a48cc9dee..78dfe95abd11f26f3b29e5aaa86236372c1f2ad2 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -1,8 +1,10 @@
-swh-model (0.11.0-1~swh1~bpo10+1) buster-swh; urgency=medium
+swh-model (0.12.0-1~swh1) unstable-swh; urgency=medium
- * Rebuild for buster-swh
+ * New upstream release 0.12.0 - (tagged by David Douard
+ <david.douard@sdfa3.org> on 2021-01-26 17:22:28 +0100)
+ * Upstream changes: - v0.12.0
- -- Software Heritage autobuilder (on jenkins-debian1) <jenkins@jenkins-debian1.internal.softwareheritage.org> Wed, 20 Jan 2021 14:36:12 +0000
+ -- Software Heritage autobuilder (on jenkins-debian1) <jenkins@jenkins-debian1.internal.softwareheritage.org> Tue, 26 Jan 2021 16:27:16 +0000
swh-model (0.11.0-1~swh1) unstable-swh; urgency=medium
diff --git a/swh.model.egg-info/PKG-INFO b/swh.model.egg-info/PKG-INFO
index 0c7e81ca9b76cc80f86bf3d5e8517d914fe3a8db..2b46f3e1f28938822858cc690385b9ea7132e04d 100644
--- a/swh.model.egg-info/PKG-INFO
+++ b/swh.model.egg-info/PKG-INFO
@@ -1,6 +1,6 @@
Metadata-Version: 2.1
Name: swh.model
-Version: 0.11.0
+Version: 0.12.0
Summary: Software Heritage data model
Home-page: https://forge.softwareheritage.org/diffusion/DMOD/
Author: Software Heritage developers
diff --git a/swh.model.egg-info/SOURCES.txt b/swh.model.egg-info/SOURCES.txt
index 41e33ce05d1343687198b11af955dfbf13c1b07d..753984d84e32b477ac56d2e3cea965093bdab96d 100644
--- a/swh.model.egg-info/SOURCES.txt
+++ b/swh.model.egg-info/SOURCES.txt
@@ -61,6 +61,7 @@ swh/model/fields/simple.py
swh/model/tests/__init__.py
swh/model/tests/generate_testdata.py
swh/model/tests/generate_testdata_from_disk.py
+swh/model/tests/swh_model_data.py
swh/model/tests/test_cli.py
swh/model/tests/test_collections.py
swh/model/tests/test_from_disk.py
@@ -70,6 +71,7 @@ swh/model/tests/test_hypothesis_strategies.py
swh/model/tests/test_identifiers.py
swh/model/tests/test_merkle.py
swh/model/tests/test_model.py
+swh/model/tests/test_swh_model_data.py
swh/model/tests/test_toposort.py
swh/model/tests/test_validators.py
swh/model/tests/data/dir-folders/sample-folder.tgz
diff --git a/swh/model/tests/swh_model_data.py b/swh/model/tests/swh_model_data.py
new file mode 100644
index 0000000000000000000000000000000000000000..8f49709606de28f66abad0225ea1f3d9b7db595f
--- /dev/null
+++ b/swh/model/tests/swh_model_data.py
@@ -0,0 +1,348 @@
+# Copyright (C) 2019-2021 The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+import datetime
+from typing import Dict, Sequence
+
+import attr
+
+from swh.model.hashutil import MultiHash, hash_to_bytes, hash_to_hex
+from swh.model.identifiers import SWHID
+from swh.model.model import (
+ BaseModel,
+ Content,
+ Directory,
+ DirectoryEntry,
+ MetadataAuthority,
+ MetadataAuthorityType,
+ MetadataFetcher,
+ MetadataTargetType,
+ ObjectType,
+ Origin,
+ OriginVisit,
+ OriginVisitStatus,
+ Person,
+ RawExtrinsicMetadata,
+ Release,
+ Revision,
+ RevisionType,
+ SkippedContent,
+ Snapshot,
+ SnapshotBranch,
+ TargetType,
+ Timestamp,
+ TimestampWithTimezone,
+)
+
+UTC = datetime.timezone.utc
+
+CONTENTS = [
+ Content(
+ length=4,
+ data=f"foo{i}".encode(),
+ status="visible",
+ **MultiHash.from_data(f"foo{i}".encode()).digest(),
+ )
+ for i in range(10)
+] + [
+ Content(
+ length=14,
+ data=f"forbidden foo{i}".encode(),
+ status="hidden",
+ **MultiHash.from_data(f"forbidden foo{i}".encode()).digest(),
+ )
+ for i in range(10)
+]
+
+SKIPPED_CONTENTS = [
+ SkippedContent(
+ length=4,
+ status="absent",
+ reason=f"because chr({i}) != '*'",
+ **MultiHash.from_data(f"bar{i}".encode()).digest(),
+ )
+ for i in range(2)
+]
+
+duplicate_content1 = Content(
+ length=4,
+ sha1=hash_to_bytes("44973274ccef6ab4dfaaf86599792fa9c3fe4689"),
+ sha1_git=b"another-foo",
+ blake2s256=b"another-bar",
+ sha256=b"another-baz",
+ status="visible",
+)
+
+# Craft a sha1 collision
+sha1_array = bytearray(duplicate_content1.sha1_git)
+sha1_array[0] += 1
+duplicate_content2 = attr.evolve(duplicate_content1, sha1_git=bytes(sha1_array))
+
+
+DUPLICATE_CONTENTS = [duplicate_content1, duplicate_content2]
+
+
+COMMITTERS = [
+ Person(fullname=b"foo", name=b"foo", email=b""),
+ Person(fullname=b"bar", name=b"bar", email=b""),
+]
+
+DATES = [
+ TimestampWithTimezone(
+ timestamp=Timestamp(seconds=1234567891, microseconds=0,),
+ offset=120,
+ negative_utc=False,
+ ),
+ TimestampWithTimezone(
+ timestamp=Timestamp(seconds=1234567892, microseconds=0,),
+ offset=120,
+ negative_utc=False,
+ ),
+]
+
+REVISIONS = [
+ Revision(
+ id=hash_to_bytes("4ca486e65eb68e4986aeef8227d2db1d56ce51b3"),
+ message=b"hello",
+ date=DATES[0],
+ committer=COMMITTERS[0],
+ author=COMMITTERS[0],
+ committer_date=DATES[0],
+ type=RevisionType.GIT,
+ directory=b"\x01" * 20,
+ synthetic=False,
+ metadata=None,
+ parents=(),
+ ),
+ Revision(
+ id=hash_to_bytes("677063f5c405d6fc1781fc56379c9a9adf43d3a0"),
+ message=b"hello again",
+ date=DATES[1],
+ committer=COMMITTERS[1],
+ author=COMMITTERS[1],
+ committer_date=DATES[1],
+ type=RevisionType.MERCURIAL,
+ directory=b"\x02" * 20,
+ synthetic=False,
+ metadata=None,
+ parents=(),
+ extra_headers=((b"foo", b"bar"),),
+ ),
+]
+
+RELEASES = [
+ Release(
+ id=hash_to_bytes("8059dc4e17fcd0e51ca3bcd6b80f4577d281fd08"),
+ name=b"v0.0.1",
+ date=TimestampWithTimezone(
+ timestamp=Timestamp(seconds=1234567890, microseconds=0,),
+ offset=120,
+ negative_utc=False,
+ ),
+ author=COMMITTERS[0],
+ target_type=ObjectType.REVISION,
+ target=b"\x04" * 20,
+ message=b"foo",
+ synthetic=False,
+ ),
+]
+
+ORIGINS = [
+ Origin(url="https://somewhere.org/den/fox",),
+ Origin(url="https://overtherainbow.org/fox/den",),
+]
+
+ORIGIN_VISITS = [
+ OriginVisit(
+ origin=ORIGINS[0].url,
+ date=datetime.datetime(2013, 5, 7, 4, 20, 39, 369271, tzinfo=UTC),
+ visit=1,
+ type="git",
+ ),
+ OriginVisit(
+ origin=ORIGINS[1].url,
+ date=datetime.datetime(2014, 11, 27, 17, 20, 39, tzinfo=UTC),
+ visit=1,
+ type="hg",
+ ),
+ OriginVisit(
+ origin=ORIGINS[0].url,
+ date=datetime.datetime(2018, 11, 27, 17, 20, 39, tzinfo=UTC),
+ visit=2,
+ type="git",
+ ),
+ OriginVisit(
+ origin=ORIGINS[0].url,
+ date=datetime.datetime(2018, 11, 27, 17, 20, 39, tzinfo=UTC),
+ visit=3,
+ type="git",
+ ),
+ OriginVisit(
+ origin=ORIGINS[1].url,
+ date=datetime.datetime(2015, 11, 27, 17, 20, 39, tzinfo=UTC),
+ visit=2,
+ type="hg",
+ ),
+]
+
+# The origin-visit-status dates needs to be shifted slightly in the future from their
+# visit dates counterpart. Otherwise, we are hitting storage-wise the "on conflict"
+# ignore policy (because origin-visit-add creates an origin-visit-status with the same
+# parameters from the origin-visit {origin, visit, date}...
+ORIGIN_VISIT_STATUSES = [
+ OriginVisitStatus(
+ origin=ORIGINS[0].url,
+ date=datetime.datetime(2013, 5, 7, 4, 20, 39, 432222, tzinfo=UTC),
+ visit=1,
+ type="git",
+ status="ongoing",
+ snapshot=None,
+ metadata=None,
+ ),
+ OriginVisitStatus(
+ origin=ORIGINS[1].url,
+ date=datetime.datetime(2014, 11, 27, 17, 21, 12, tzinfo=UTC),
+ visit=1,
+ type="hg",
+ status="ongoing",
+ snapshot=None,
+ metadata=None,
+ ),
+ OriginVisitStatus(
+ origin=ORIGINS[0].url,
+ date=datetime.datetime(2018, 11, 27, 17, 20, 59, tzinfo=UTC),
+ visit=2,
+ type="git",
+ status="ongoing",
+ snapshot=None,
+ metadata=None,
+ ),
+ OriginVisitStatus(
+ origin=ORIGINS[0].url,
+ date=datetime.datetime(2018, 11, 27, 17, 20, 49, tzinfo=UTC),
+ visit=3,
+ type="git",
+ status="full",
+ snapshot=hash_to_bytes("17d0066a4a80aba4a0e913532ee8ff2014f006a9"),
+ metadata=None,
+ ),
+ OriginVisitStatus(
+ origin=ORIGINS[1].url,
+ date=datetime.datetime(2015, 11, 27, 17, 22, 18, tzinfo=UTC),
+ visit=2,
+ type="hg",
+ status="partial",
+ snapshot=hash_to_bytes("8ce268b87faf03850693673c3eb5c9bb66e1ca38"),
+ metadata=None,
+ ),
+]
+
+
+DIRECTORIES = [
+ Directory(id=hash_to_bytes("4b825dc642cb6eb9a060e54bf8d69288fbee4904"), entries=()),
+ Directory(
+ id=hash_to_bytes("21416d920e0ebf0df4a7888bed432873ed5cb3a7"),
+ entries=(
+ DirectoryEntry(
+ name=b"file1.ext",
+ perms=0o644,
+ type="file",
+ target=CONTENTS[0].sha1_git,
+ ),
+ DirectoryEntry(
+ name=b"dir1",
+ perms=0o755,
+ type="dir",
+ target=hash_to_bytes("4b825dc642cb6eb9a060e54bf8d69288fbee4904"),
+ ),
+ DirectoryEntry(
+ name=b"subprepo1", perms=0o160000, type="rev", target=REVISIONS[1].id,
+ ),
+ ),
+ ),
+]
+
+
+SNAPSHOTS = [
+ Snapshot(
+ id=hash_to_bytes("17d0066a4a80aba4a0e913532ee8ff2014f006a9"),
+ branches={
+ b"master": SnapshotBranch(
+ target_type=TargetType.REVISION, target=REVISIONS[0].id
+ )
+ },
+ ),
+ Snapshot(
+ id=hash_to_bytes("8ce268b87faf03850693673c3eb5c9bb66e1ca38"),
+ branches={
+ b"target/revision": SnapshotBranch(
+ target_type=TargetType.REVISION, target=REVISIONS[0].id,
+ ),
+ b"target/alias": SnapshotBranch(
+ target_type=TargetType.ALIAS, target=b"target/revision"
+ ),
+ b"target/directory": SnapshotBranch(
+ target_type=TargetType.DIRECTORY, target=DIRECTORIES[0].id,
+ ),
+ b"target/release": SnapshotBranch(
+ target_type=TargetType.RELEASE, target=RELEASES[0].id
+ ),
+ b"target/snapshot": SnapshotBranch(
+ target_type=TargetType.SNAPSHOT,
+ target=hash_to_bytes("17d0066a4a80aba4a0e913532ee8ff2014f006a9"),
+ ),
+ },
+ ),
+]
+
+
+METADATA_AUTHORITIES = [
+ MetadataAuthority(
+ type=MetadataAuthorityType.FORGE, url="http://example.org/", metadata={},
+ ),
+]
+
+METADATA_FETCHERS = [
+ MetadataFetcher(name="test-fetcher", version="1.0.0", metadata={},)
+]
+
+RAW_EXTRINSIC_METADATA = [
+ RawExtrinsicMetadata(
+ type=MetadataTargetType.ORIGIN,
+ target="http://example.org/foo.git",
+ discovery_date=datetime.datetime(2020, 7, 30, 17, 8, 20, tzinfo=UTC),
+ authority=attr.evolve(METADATA_AUTHORITIES[0], metadata=None),
+ fetcher=attr.evolve(METADATA_FETCHERS[0], metadata=None),
+ format="json",
+ metadata=b'{"foo": "bar"}',
+ ),
+ RawExtrinsicMetadata(
+ type=MetadataTargetType.CONTENT,
+ target=SWHID(
+ object_type="content", object_id=hash_to_hex(CONTENTS[0].sha1_git)
+ ),
+ discovery_date=datetime.datetime(2020, 7, 30, 17, 8, 20, tzinfo=UTC),
+ authority=attr.evolve(METADATA_AUTHORITIES[0], metadata=None),
+ fetcher=attr.evolve(METADATA_FETCHERS[0], metadata=None),
+ format="json",
+ metadata=b'{"foo": "bar"}',
+ ),
+]
+
+
+TEST_OBJECTS: Dict[str, Sequence[BaseModel]] = {
+ "content": CONTENTS,
+ "directory": DIRECTORIES,
+ "metadata_authority": METADATA_AUTHORITIES,
+ "metadata_fetcher": METADATA_FETCHERS,
+ "origin": ORIGINS,
+ "origin_visit": ORIGIN_VISITS,
+ "origin_visit_status": ORIGIN_VISIT_STATUSES,
+ "raw_extrinsic_metadata": RAW_EXTRINSIC_METADATA,
+ "release": RELEASES,
+ "revision": REVISIONS,
+ "snapshot": SNAPSHOTS,
+ "skipped_content": SKIPPED_CONTENTS,
+}
diff --git a/swh/model/tests/test_swh_model_data.py b/swh/model/tests/test_swh_model_data.py
new file mode 100644
index 0000000000000000000000000000000000000000..7b50e60e54d348d50c88b607aeab6f005239a5f1
--- /dev/null
+++ b/swh/model/tests/test_swh_model_data.py
@@ -0,0 +1,35 @@
+# Copyright (C) 2021 The Software Heritage developers
+# See the AUTHORS file at the top-level directory of this distribution
+# License: GNU General Public License version 3, or any later version
+# See top-level LICENSE file for more information
+
+import attr
+import pytest
+
+from swh.model.tests.swh_model_data import TEST_OBJECTS
+
+
+@pytest.mark.parametrize("object_type, objects", TEST_OBJECTS.items())
+def test_swh_model_data(object_type, objects):
+ """checks model objects in swh_model_data are in correct shape"""
+ assert objects
+ for obj in objects:
+ assert obj.object_type == object_type
+ attr.validate(obj)
+
+
+def test_ensure_visit_visit_status_date_consistency():
+ """ensure origin-visit-status dates are more recent than their visit counterpart
+
+ The origin-visit-status dates needs to be shifted slightly in the future from their
+ visit dates counterpart. Otherwise, we are hitting storage-wise the "on conflict"
+ ignore policy (because origin-visit-add creates an origin-visit-status with the same
+ parameters from the origin-visit {origin, visit, date}...
+
+ """
+ visits = TEST_OBJECTS["origin_visit"]
+ visit_statuses = TEST_OBJECTS["origin_visit_status"]
+ for visit, visit_status in zip(visits, visit_statuses):
+ assert visit.origin == visit_status.origin
+ assert visit.visit == visit_status.visit
+ assert visit.date < visit_status.date