Page MenuHomeSoftware Heritage
Differential D7603

[WIP] Add support for commits with no author
Changes PlannedPublicDraft
Actions

Authored by vlorentz on Apr 20 2022, 12:56 PM.

Details

Reviewers
None
Group Reviewers
Reviewers
Maniphest Tasks
T3880: Support Git commits with no angle brackets in author name
Summary

Currently stuck because this is high effort and low priority;
I am only opening this diff so I don't lose this WIP code

Resolves T3880

Diff Detail

Repository
rDLDG Git loader
Branch
no-author
Lint
No Linters Available
Unit
No Unit Test Coverage
Build Status
Buildable 28566
Build 44659: Phabricator diff pipeline on jenkinsJenkins console · Jenkins
Build 44658: arc lint + arc unit

Event Timeline

vlorentz created this revision.Apr 20 2022, 12:56 PM
Herald added a reviewer: Reviewers. · View Herald TranscriptApr 20 2022, 12:56 PM
swh-public-ci added a comment.Apr 20 2022, 12:57 PM

Build has FAILED

Patch application report for D7603 (id=27538)

Rebasing onto 4281899e4a...

Current branch diff-target is up to date.
Changes applied before test
commit 8123c03673d22b28f18fcf88995e865182724b40
Author: Valentin Lorentz <vlorentz@softwareheritage.org>
Date:   Wed Apr 20 12:55:49 2022 +0200

    [WIP] Add support for commits with no author

Link to build: https://jenkins.softwareheritage.org/job/DLDG/job/tests-on-diff/186/
See console output for more information: https://jenkins.softwareheritage.org/job/DLDG/job/tests-on-diff/186/console

Harbormaster returned this revision to the author for changes because remote builds failed.Apr 20 2022, 12:57 PM
Harbormaster failed remote builds in B28566: Diff 27538!
vlorentz added a task: Restricted Maniphest Task.Sep 27 2022, 9:10 AM
vlorentz removed a task: Restricted Maniphest Task.Sep 27 2022, 9:24 AM

Revision Contents
Changeset List

PathSize
swh/
loader/
git/
8 lines
tests/
71 lines

Diff 27538

View Options

swh/loader/git/converters.py

Show First 20 Lines • Show All 141 Lines • ▼ Show 20 Lines if dir_.compute_hash() != dir_.id:
dir_ = attr.evolve( dir_ = attr.evolve(
dir_, raw_manifest=git_object_header("tree", len(raw_string)) + raw_string dir_, raw_manifest=git_object_header("tree", len(raw_string)) + raw_string
) )
check_id(dir_) check_id(dir_)
return dir_ return dir_
def parse_author(name_email: bytes) -> Person: def parse_author(name_email: Optional[bytes]) -> Optional[Person]:
"""Parse an author line""" """Parse an author line"""
if name_email is None:
return None
return Person.from_fullname(name_email) return Person.from_fullname(name_email)
def dulwich_tsinfo_to_timestamp( def dulwich_tsinfo_to_timestamp(
timestamp, timestamp: Optional[str],
timezone: int, timezone: int,
timezone_neg_utc: bool, timezone_neg_utc: bool,
timezone_bytes: Optional[bytes], timezone_bytes: Optional[bytes],
) -> TimestampWithTimezone: ) -> TimestampWithTimezone:
"""Convert the dulwich timestamp information to a structure compatible with """Convert the dulwich timestamp information to a structure compatible with
Software Heritage.""" Software Heritage."""
if timestamp is None:
return None
ts = Timestamp( ts = Timestamp(
seconds=int(timestamp), seconds=int(timestamp),
microseconds=0, microseconds=0,
) )
if timezone_bytes is None: if timezone_bytes is None:
# Failed to parse from the raw manifest, fallback to what Dulwich managed to # Failed to parse from the raw manifest, fallback to what Dulwich managed to
# parse. # parse.
return TimestampWithTimezone.from_numeric_offset( return TimestampWithTimezone.from_numeric_offset(
▲ Show 20 Lines • Show All 157 Lines • Show Last 20 Lines
View Options

swh/loader/git/tests/test_converters.py

Show First 20 Lines • Show All 320 Lines • ▼ Show 20 Lines def test_commit_to_revision_with_extra_headers(self):
offset_bytes=b"+0200", offset_bytes=b"+0200",
), ),
parents=(bytes.fromhex("c730509025c6e81947102b2d77bc4dc1cade9489"),), parents=(bytes.fromhex("c730509025c6e81947102b2d77bc4dc1cade9489"),),
synthetic=False, synthetic=False,
) )
assert revision == expected_revision assert revision == expected_revision
def test_commit_without_author(self):
target = b"641fb6e08ddb2e4fd096dcf18e80b894bf7e25ce"
message = b"some commit message"
author = None
raw_manifest = (
b"tree 641fb6e08ddb2e4fd096dcf18e80b894bf7e25ce\n" b"some commit message"
)
commit = dulwich.objects.Commit.from_raw_string(b"commit", raw_manifest)
assert converters.dulwich_commit_to_revision(commit) == Revision(
message=b"some commit message",
author=author,
committer=author,
date=None,
committer_date=None,
type=RevisionType.GIT,
directory=hash_to_bytes(target.decode()),
synthetic=False,
metadata=None,
parents=(),
)
def test_commit_without_author_name(self):
target = b"641fb6e08ddb2e4fd096dcf18e80b894bf7e25ce"
message = b"some commit message"
author = Person(fullname=b"foo@example.org", name=b"", email=b"")
raw_manifest = (
b"tree 641fb6e08ddb2e4fd096dcf18e80b894bf7e25ce\n"
b"author foo@example.org 1640191028 +0200\n"
b"committer foo@example.org 1640191028 +0200\n\n"
b"some commit message"
)
commit = dulwich.objects.Commit.from_raw_string(b"commit", raw_manifest)
assert converters.dulwich_commit_to_revision(commit) == Revision(
message=b"some commit message",
author=author,
committer=author,
date=None,
committer_date=None,
type=RevisionType.GIT,
directory=hash_to_bytes(target.decode()),
synthetic=False,
metadata=None,
parents=(),
)
def test_commit_without_dates(self):
target = b"641fb6e08ddb2e4fd096dcf18e80b894bf7e25ce"
message = b"some commit message"
author = Person(
fullname=b"Foo <foo@example.org>", name=b"Foo", email=b"foo@example.org"
)
raw_manifest = (
b"tree 641fb6e08ddb2e4fd096dcf18e80b894bf7e25ce\n"
b"author Foo <foo@example.org>\n"
b"committer Foo <foo@example.org>\n\n"
b"some commit message"
)
commit = dulwich.objects.Commit.from_raw_string(b"commit", raw_manifest)
assert converters.dulwich_commit_to_revision(commit) == Revision(
message=b"some commit message",
author=author,
committer=author,
date=None,
committer_date=None,
type=RevisionType.GIT,
directory=hash_to_bytes(target.decode()),
synthetic=False,
metadata=None,
parents=(),
)
def test_commit_without_manifest(self): def test_commit_without_manifest(self):
"""Tests a Release can still be produced when the manifest is not understood """Tests a Release can still be produced when the manifest is not understood
by the custom parser in dulwich_commit_to_revision.""" by the custom parser in dulwich_commit_to_revision."""
target = b"641fb6e08ddb2e4fd096dcf18e80b894bf7e25ce" target = b"641fb6e08ddb2e4fd096dcf18e80b894bf7e25ce"
message = b"some commit message" message = b"some commit message"
author = Person( author = Person(
fullname=b"Foo <foo@example.org>", name=b"Foo", email=b"foo@example.org" fullname=b"Foo <foo@example.org>", name=b"Foo", email=b"foo@example.org"
) )
▲ Show 20 Lines • Show All 534 Lines • Show Last 20 Lines
About · Developer information · Legal