diff --git a/PKG-INFO b/PKG-INFO
index 1d0b3f063793879a0213fde8fc01b5eb6aa27a13..c5e6a7f0c79a058b0222de1f126fbe93293c6f80 100644
--- a/PKG-INFO
+++ b/PKG-INFO
@@ -1,6 +1,6 @@
Metadata-Version: 2.1
Name: swh.model
-Version: 0.0.52
+Version: 0.0.53
Summary: Software Heritage data model
Home-page: https://forge.softwareheritage.org/diffusion/DMOD/
Author: Software Heritage developers
diff --git a/swh.model.egg-info/PKG-INFO b/swh.model.egg-info/PKG-INFO
index 1d0b3f063793879a0213fde8fc01b5eb6aa27a13..c5e6a7f0c79a058b0222de1f126fbe93293c6f80 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: 0.0.52
+Version: 0.0.53
Summary: Software Heritage data model
Home-page: https://forge.softwareheritage.org/diffusion/DMOD/
Author: Software Heritage developers
diff --git a/swh/model/cli.py b/swh/model/cli.py
index ec33310b05722dc1fe1325e3bcb489737a42547a..581bb45b892cae89ab2bb6415fb8d608d905784b 100644
--- a/swh/model/cli.py
+++ b/swh/model/cli.py
@@ -45,6 +45,11 @@ def pid_of_file(path):
return pids.persistent_identifier(pids.CONTENT, object)
+def pid_of_file_content(data):
+ object = Content.from_bytes(mode=644, data=data).get_data()
+ return pids.persistent_identifier(pids.CONTENT, object)
+
+
def pid_of_dir(path):
object = Directory.from_disk(path=path).get_data()
return pids.persistent_identifier(pids.DIRECTORY, object)
@@ -85,7 +90,7 @@ def pid_of_git_repo(path):
def identify_object(obj_type, follow_symlinks, obj):
if obj_type == 'auto':
- if os.path.isfile(obj):
+ if obj == '-' or os.path.isfile(obj):
obj_type = 'content'
elif os.path.isdir(obj):
obj_type = 'directory'
@@ -101,7 +106,10 @@ def identify_object(obj_type, follow_symlinks, obj):
pid = None
- if obj_type in ['content', 'directory']:
+ if obj == '-':
+ content = sys.stdin.buffer.read()
+ pid = pid_of_file_content(content)
+ elif obj_type in ['content', 'directory']:
path = obj.encode(sys.getfilesystemencoding())
if follow_symlinks and os.path.islink(obj):
path = os.path.realpath(obj)
@@ -134,7 +142,7 @@ def identify_object(obj_type, follow_symlinks, obj):
help='type of object to identify (default: auto)')
@click.option('--verify', '-v', metavar='PID', type=PidParamType(),
help='reference identifier to be compared with computed one')
-@click.argument('objects', nargs=-1, required=True)
+@click.argument('objects', nargs=-1)
def identify(obj_type, verify, show_filename, follow_symlinks, objects):
"""Compute the Software Heritage persistent identifier (PID) for the given
source code object(s).
@@ -163,6 +171,9 @@ def identify(obj_type, verify, show_filename, follow_symlinks, objects):
swh:1:snp:510aa88bdc517345d258c1fc2babcd0e1f905e93 helloworld.git
""" # NoQA # overlong lines in shell examples are fine
+ if not objects:
+ objects = ['-']
+
if verify and len(objects) != 1:
raise click.BadParameter('verification requires a single object')
diff --git a/swh/model/hypothesis_strategies.py b/swh/model/hypothesis_strategies.py
index 15e3bdd94f44c882cd65d581926dc9ebac64bfc7..fcf67ca43136823cdc56502659cc5be285b498f7 100644
--- a/swh/model/hypothesis_strategies.py
+++ b/swh/model/hypothesis_strategies.py
@@ -200,6 +200,7 @@ def snapshots(draw, *, min_size=0, max_size=100, only_objects=False):
for alias in unresolved_aliases:
branches[alias] = draw(branch_targets(only_objects=True))
+ # Ensure no cycles between aliases
while True:
try:
id_ = snapshot_identifier({
diff --git a/swh/model/model.py b/swh/model/model.py
index f551706d31cd491e034ab2c72fead75fab3065be..4a308a4045eaf1b66e010fb0a9e9b477a46af0de 100644
--- a/swh/model/model.py
+++ b/swh/model/model.py
@@ -119,7 +119,7 @@ class TimestampWithTimezone(BaseModel):
@classmethod
def from_dict(cls, d):
"""Builds a TimestampWithTimezone from any of the formats
- accepted by :py:`swh.model.normalize_timestamp`."""
+ accepted by :func:`swh.model.normalize_timestamp`."""
d = normalize_timestamp(d)
return cls(
timestamp=Timestamp.from_dict(d['timestamp']),
@@ -296,8 +296,8 @@ class Revision(BaseModel, HashableObject):
message = attr.ib(type=bytes)
author = attr.ib(type=Person)
committer = attr.ib(type=Person)
- date = attr.ib(type=TimestampWithTimezone)
- committer_date = attr.ib(type=TimestampWithTimezone)
+ date = attr.ib(type=Optional[TimestampWithTimezone])
+ committer_date = attr.ib(type=Optional[TimestampWithTimezone])
type = attr.ib(type=RevisionType)
directory = attr.ib(type=Sha1Git)
synthetic = attr.ib(type=bool)
@@ -314,12 +314,20 @@ class Revision(BaseModel, HashableObject):
@classmethod
def from_dict(cls, d):
d = d.copy()
+ date = d.pop('date')
+ if date:
+ date = TimestampWithTimezone.from_dict(date)
+
+ committer_date = d.pop('committer_date')
+ if committer_date:
+ committer_date = TimestampWithTimezone.from_dict(
+ committer_date)
+
return cls(
author=Person.from_dict(d.pop('author')),
committer=Person.from_dict(d.pop('committer')),
- date=TimestampWithTimezone.from_dict(d.pop('date')),
- committer_date=TimestampWithTimezone.from_dict(
- d.pop('committer_date')),
+ date=date,
+ committer_date=committer_date,
type=RevisionType(d.pop('type')),
**d)
diff --git a/swh/model/tests/test_cli.py b/swh/model/tests/test_cli.py
index f20da7a83a6d9a77d185033fe77c52b6b5d8ed82..4d4ff0116efe58fe26fe6cc5733462c4b58af533 100644
--- a/swh/model/tests/test_cli.py
+++ b/swh/model/tests/test_cli.py
@@ -37,6 +37,15 @@ class TestIdentify(DataMixin, unittest.TestCase):
self.assertPidOK(result,
'swh:1:cnt:' + hash_to_hex(content['sha1_git']))
+ def test_content_id_from_stdin(self):
+ """identify file content"""
+ self.make_contents(self.tmpdir_name)
+ for _, content in self.contents.items():
+ result = self.runner.invoke(cli.identify,
+ input=content['data'])
+ self.assertPidOK(result,
+ 'swh:1:cnt:' + hash_to_hex(content['sha1_git']))
+
def test_directory_id(self):
"""identify an entire directory"""
self.make_from_tarball(self.tmpdir_name)
diff --git a/swh/model/tests/test_model.py b/swh/model/tests/test_model.py
index a5719ca0a86f32b3a38cd3f9653fcc706785d766..5560127ed87e1f7e961c25e00e9d8c39db492d07 100644
--- a/swh/model/tests/test_model.py
+++ b/swh/model/tests/test_model.py
@@ -66,49 +66,49 @@ def test_directory_model_id_computation():
dir_dict = dict(directory_example)
del dir_dict['id']
- dir_model = Directory(**dir_dict)
- assert dir_model.id
- assert dir_model.id == hash_to_bytes(directory_identifier(dir_dict))
-
- dir_model = Directory.from_dict(dir_dict)
- assert dir_model.id
- assert dir_model.id == hash_to_bytes(directory_identifier(dir_dict))
+ dir_id = hash_to_bytes(directory_identifier(dir_dict))
+ for dir_model in [Directory(**dir_dict), Directory.from_dict(dir_dict)]:
+ assert dir_model.id == dir_id
def test_revision_model_id_computation():
rev_dict = dict(revision_example)
del rev_dict['id']
- rev_model = Revision(**rev_dict)
- assert rev_model.id
- assert rev_model.id == hash_to_bytes(revision_identifier(rev_dict))
+ rev_id = hash_to_bytes(revision_identifier(rev_dict))
+ for rev_model in [Revision(**rev_dict), Revision.from_dict(rev_dict)]:
+ assert rev_model.id == rev_id
+
+
+def test_revision_model_id_computation_with_no_date():
+ """We can have revision with date to None
+
+ """
+ rev_dict = dict(revision_example)
+ rev_dict['date'] = None
+ rev_dict['committer_date'] = None
+ del rev_dict['id']
- rev_model = Revision.from_dict(rev_dict)
- assert rev_model.id
- assert rev_model.id == hash_to_bytes(revision_identifier(rev_dict))
+ rev_id = hash_to_bytes(revision_identifier(rev_dict))
+ for rev_model in [Revision(**rev_dict), Revision.from_dict(rev_dict)]:
+ assert rev_model.date is None
+ assert rev_model.committer_date is None
+ assert rev_model.id == rev_id
def test_release_model_id_computation():
rel_dict = dict(release_example)
del rel_dict['id']
- rel_model = Release(**rel_dict)
- assert rel_model.id
- assert rel_model.id == hash_to_bytes(release_identifier(rel_dict))
-
- rel_model = Release.from_dict(rel_dict)
- assert rel_model.id
- assert rel_model.id == hash_to_bytes(release_identifier(rel_dict))
+ rel_id = hash_to_bytes(release_identifier(rel_dict))
+ for rel_model in [Release(**rel_dict), Release.from_dict(rel_dict)]:
+ assert rel_model.id == hash_to_bytes(rel_id)
def test_snapshot_model_id_computation():
snp_dict = dict(snapshot_example)
del snp_dict['id']
- snp_model = Snapshot(**snp_dict)
- assert snp_model.id
- assert snp_model.id == hash_to_bytes(snapshot_identifier(snp_dict))
-
- snp_model = Snapshot.from_dict(snp_dict)
- assert snp_model.id
- assert snp_model.id == hash_to_bytes(snapshot_identifier(snp_dict))
+ snp_id = hash_to_bytes(snapshot_identifier(snp_dict))
+ for snp_model in [Snapshot(**snp_dict), Snapshot.from_dict(snp_dict)]:
+ assert snp_model.id == snp_id
diff --git a/version.txt b/version.txt
index c9e30f8952411f8a87cf2ee9ca7cb14a2cae870d..235911d5ec143fc28371a75af99adde637188aaa 100644
--- a/version.txt
+++ b/version.txt
@@ -1 +1 @@
-v0.0.52-0-g4e4c4ff
\ No newline at end of file
+v0.0.53-0-g4b779e1
\ No newline at end of file