This is an automated email from the ASF dual-hosted git repository.
sbp pushed a commit to branch sbp
in repository https://gitbox.apache.org/repos/asf/tooling-trusted-releases.git
The following commit(s) were added to refs/heads/sbp by this push:
new e5c186ad Add a variant of the license header
e5c186ad is described below
commit e5c186ad2000bd2c4f6f2bb756772fa1cddf870e
Author: Sean B. Palmer <[email protected]>
AuthorDate: Fri Mar 13 14:53:20 2026 +0000
Add a variant of the license header
---
atr/tasks/checks/license.py | 32 ++++++++---
tests/unit/test_checks_license.py | 117 ++++++++++++++++++++------------------
2 files changed, 85 insertions(+), 64 deletions(-)
diff --git a/atr/tasks/checks/license.py b/atr/tasks/checks/license.py
index abef74e7..d3e00cec 100644
--- a/atr/tasks/checks/license.py
+++ b/atr/tasks/checks/license.py
@@ -47,6 +47,26 @@ HTTP_APACHE_LICENSE_HEADER: Final[bytes] = (
HTTPS_APACHE_LICENSE_HEADER: Final[bytes] =
HTTP_APACHE_LICENSE_HEADER.replace(b" http ", b" https ")
+HTTP_APACHE_LICENSE_HEADER_GENERIC: Final[bytes] = (
+ b"Licensed under the Apache License Version 2 0 the License you m"
+ b"ay not use this file except in compliance with the License You "
+ b"may obtain a copy of the License at http www apache org license"
+ b"s LICENSE 2 0 Unless required by applicable law or agreed to in"
+ b" writing software distributed under the License is distributed "
+ b"on an AS IS BASIS WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND "
+ b"either express or implied See the License for the specific lang"
+ b"uage governing permissions and limitations under the License"
+)
+
+HTTPS_APACHE_LICENSE_HEADER_GENERIC: Final[bytes] =
HTTP_APACHE_LICENSE_HEADER_GENERIC.replace(b" http ", b" https ")
+
+ACCEPTED_APACHE_LICENSE_HEADERS: Final[tuple[bytes, ...]] = (
+ HTTP_APACHE_LICENSE_HEADER.lower(),
+ HTTPS_APACHE_LICENSE_HEADER.lower(),
+ HTTP_APACHE_LICENSE_HEADER_GENERIC.lower(),
+ HTTPS_APACHE_LICENSE_HEADER_GENERIC.lower(),
+)
+
# Patterns for files to include in license header checks
# Ordered by their popularity in the Stack Overflow Developer Survey 2024
INCLUDED_PATTERNS: Final[list[str]] = [
@@ -227,20 +247,14 @@ def headers_validate(content: bytes, _filename: str) ->
tuple[bool, str | None]:
if pattern in content:
return True, None
- r_span = re.compile(rb"Licensed to the.*?under the License", re.MULTILINE)
r_words = re.compile(rb"[A-Za-z0-9]+")
# Normalise the content
content = re.sub(rb"[ \t\r\n]+", b" ", content)
- # For each matching heuristic span...
- for span in r_span.finditer(content):
- # Get only the words in the span
- words = r_words.findall(span.group(0))
- joined = b" ".join(words).lower()
- if joined == HTTP_APACHE_LICENSE_HEADER.lower():
- return True, None
- elif joined == HTTPS_APACHE_LICENSE_HEADER.lower():
+ joined = b" ".join(r_words.findall(content)).lower()
+ for accepted_header in ACCEPTED_APACHE_LICENSE_HEADERS:
+ if accepted_header in joined:
return True, None
return False, "Could not find Apache License header"
diff --git a/tests/unit/test_checks_license.py
b/tests/unit/test_checks_license.py
index 4353da73..abd7d4ad 100644
--- a/tests/unit/test_checks_license.py
+++ b/tests/unit/test_checks_license.py
@@ -40,6 +40,51 @@ NOTICE_VALID: str = (
)
+class BinaryRecorder(checks.Recorder):
+ def __init__(self, path: pathlib.Path) -> None:
+ super().__init__(
+ checker="tests.unit.test_checks_license",
+ inputs_hash=None,
+ project_name="test",
+ version_name="0.2.0",
+ revision_number="00001",
+ primary_rel_path="apache-test-0.2-bin.zip",
+ member_rel_path=None,
+ afresh=False,
+ )
+ self._path = path
+ self.messages: list[tuple[str, str, dict | None]] = []
+
+ async def abs_path(self, rel_path: str | None = None) -> pathlib.Path |
None:
+ return self._path if rel_path is None else self._path / rel_path
+
+ async def primary_path_is_binary(self) -> bool:
+ return True
+
+ async def _add(
+ self,
+ status: sql.CheckResultStatus,
+ message: str,
+ data: object,
+ primary_rel_path: str | None = None,
+ member_rel_path: str | None = None,
+ ) -> sql.CheckResult:
+ self.messages.append((status.value, message, data if isinstance(data,
dict) else None))
+ return sql.CheckResult(
+ id=0,
+ release_name=self.release_name,
+ revision_number=self.revision_number,
+ checker=self.checker,
+ primary_rel_path=primary_rel_path,
+ member_rel_path=member_rel_path,
+ created=datetime.datetime.now(datetime.UTC),
+ status=status,
+ message=message,
+ data=data,
+ inputs_hash=None,
+ )
+
+
def test_files_missing_cache_dir():
results =
list(license._files_check_core_logic(pathlib.Path("/nonexistent"),
is_podling=False))
assert len(results) == 1
@@ -89,61 +134,6 @@ def
test_files_single_root_with_stray_top_level_file(tmp_path):
assert sql.CheckResultStatus.SUCCESS in statuses
-def test_files_valid_license_and_notice(tmp_path):
- cache_dir = _cache_with_root(tmp_path)
- root = cache_dir / "apache-test-0.2"
- (root / "LICENSE").write_text(constants.APACHE_LICENSE_2_0)
- (root / "NOTICE").write_text(NOTICE_VALID)
- results = list(license._files_check_core_logic(cache_dir,
is_podling=False))
- artifact_results = [r for r in results if isinstance(r,
license.ArtifactResult)]
- assert all(r.status == sql.CheckResultStatus.SUCCESS for r in
artifact_results)
-
-
-class BinaryRecorder(checks.Recorder):
- def __init__(self, path: pathlib.Path) -> None:
- super().__init__(
- checker="tests.unit.test_checks_license",
- inputs_hash=None,
- project_name="test",
- version_name="0.2.0",
- revision_number="00001",
- primary_rel_path="apache-test-0.2-bin.zip",
- member_rel_path=None,
- afresh=False,
- )
- self._path = path
- self.messages: list[tuple[str, str, dict | None]] = []
-
- async def abs_path(self, rel_path: str | None = None) -> pathlib.Path |
None:
- return self._path if rel_path is None else self._path / rel_path
-
- async def primary_path_is_binary(self) -> bool:
- return True
-
- async def _add(
- self,
- status: sql.CheckResultStatus,
- message: str,
- data: object,
- primary_rel_path: str | None = None,
- member_rel_path: str | None = None,
- ) -> sql.CheckResult:
- self.messages.append((status.value, message, data if isinstance(data,
dict) else None))
- return sql.CheckResult(
- id=0,
- release_name=self.release_name,
- revision_number=self.revision_number,
- checker=self.checker,
- primary_rel_path=primary_rel_path,
- member_rel_path=member_rel_path,
- created=datetime.datetime.now(datetime.UTC),
- status=status,
- message=message,
- data=data,
- inputs_hash=None,
- )
-
-
@pytest.mark.asyncio
async def test_files_skip_binary_artifacts(tmp_path: pathlib.Path) -> None:
recorder = BinaryRecorder(tmp_path / "apache-test-0.2-bin.zip")
@@ -164,6 +154,16 @@ async def test_files_skip_binary_artifacts(tmp_path:
pathlib.Path) -> None:
assert recorder.messages == []
+def test_files_valid_license_and_notice(tmp_path):
+ cache_dir = _cache_with_root(tmp_path)
+ root = cache_dir / "apache-test-0.2"
+ (root / "LICENSE").write_text(constants.APACHE_LICENSE_2_0)
+ (root / "NOTICE").write_text(NOTICE_VALID)
+ results = list(license._files_check_core_logic(cache_dir,
is_podling=False))
+ artifact_results = [r for r in results if isinstance(r,
license.ArtifactResult)]
+ assert all(r.status == sql.CheckResultStatus.SUCCESS for r in
artifact_results)
+
+
def test_headers_check_data_fields_match_model(tmp_path):
cache_dir = _extract_test_archive(tmp_path)
results = list(license._headers_check_core_logic(cache_dir,
TEST_ARCHIVE_BASENAME, [], "none"))
@@ -209,6 +209,13 @@ def
test_headers_check_includes_excludes_source_policy(tmp_path):
assert final_result.data["excludes_source"] == "policy"
+def test_headers_validate_accepts_generic_apache_license_header():
+ content = pathlib.Path("license-normal.txt").read_bytes()
+ is_valid, error = license.headers_validate(content,
"publish-on-homebrew.sh")
+ assert is_valid is True
+ assert error is None
+
+
def _cache_with_root(tmp_path: pathlib.Path) -> pathlib.Path:
cache_dir = tmp_path / "cache"
cache_dir.mkdir()
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
