Make pyproject.toml + hatch the sole build source of truth (PEP 517/518)

.github/workflows/test.yml

@@ -22,7 +22,9 @@ jobs:
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        python -m pip install hatch 
+        # Pin virtualenv<21: virtualenv 21.0 removed an API hatch depends on
+        # https://github.com/pypa/hatch/issues/2193
+        python -m pip install hatch 'virtualenv<21'
     - name: Lint check
       working-directory: ./python
       run: make lint LINT_PARAMS="-- --check --diff"
@@ -60,7 +62,9 @@ jobs:
       run: |
         python -m pip install --upgrade pip
         python -m pip install semver
-        python -m pip install hatch
+        # Pin virtualenv<21: virtualenv 21.0 removed an API hatch depends on
+        # https://github.com/pypa/hatch/issues/2193
+        python -m pip install hatch 'virtualenv<21'
     - name: Execute hatch envs
       working-directory: ./python
       run: make test coverage-report DBR=${{ matrix.config.dbr }}

python/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["hatchling"]
 build-backend = "hatchling.build"
 
 [project]
-name = "Tempo"
+name = "dbl-tempo"
 dynamic = ["version"]
 description = 'Tempo is timeseries manipulation for Spark. This project builds upon the capabilities of PySpark to provide a suite of abstractions and functions that make operations on timeseries data easier and highly scalable.'
 readme = "README.md"
@@ -146,6 +146,7 @@ runLint = [
   "black --check --diff ./tempo",
   "flake8 --config ./.flake8 ./tempo"
 ]
+formatBlack = ["black ./tempo ./tests"]
 
 [tool.hatch.envs.type-check]
 template = "testenv"
@@ -161,6 +162,33 @@ dependencies = [
 check = "mypy --install-types --non-interactive ./tempo"
 
 
+[tool.hatch.build.targets.wheel]
+packages = ["tempo"]
+
+[tool.hatch.build.targets.sdist]
+include = [
+    "/tempo",
+    "/tests",
+    "/README.md",
+    "/pyproject.toml",
+    "/version.py"
+]
+exclude = [
+    "/.venv*",
+    "/.tox",
+    "/.mypy_cache",
+    "/.pytest_cache",
+    "/__pycache__",
+    "*.pyc",
+    "*.pyo",
+    "*.egg-info",
+    "/dist",
+    "/build",
+    "/htmlcov",
+    "/.coverage*",
+    "/coverage.xml"
+]
+
 [tool.hatch.envs.build-dist]
 template = "testenv"
 skip-install = true
@@ -170,7 +198,7 @@ dependencies = [
 ]
 
 [tool.hatch.envs.build-dist.scripts]
-build = "python setup.py clean bdist_wheel"
+build = "hatch build"
 
 
 [tool.hatch.envs.build-docs]
@@ -218,3 +246,21 @@ run-coverage = [
   "coverage report -m",
   "coverage xml"
 ]
+
+[tool.black]
+line-length = 88
+target-version = ['py39', 'py310', 'py311']
+include = '\.pyi?$'
+extend-exclude = '''
+/(
+  # directories
+  \.eggs
+  | \.git
+  | \.hg
+  | \.mypy_cache
+  | \.tox
+  | \.venv
+  | build
+  | dist
+)/
+'''

python/tempo/ml.py

@@ -37,7 +37,7 @@ def __init__(
         timeSeriesCol: str = "event_ts",
         seriesIdCols: List[str] = [],
         gap: int = 0,
-        **other_kwargs: Any
+        **other_kwargs: Any,
     ) -> None:
         super(TimeSeriesCrossValidator, self).__init__(**other_kwargs)
         self._setDefault(timeSeriesCol="event_ts", seriesIdCols=[], gap=0)

python/tox.ini

@@ -87,10 +87,9 @@ skip_install = true
 deps =
     build
     semver
-    setuptools
-    wheel
+    hatchling
 commands =
-    python setup.py clean bdist_wheel
+    hatch build
 
 [testenv:build-docs]
 description = build documentation

python/version.py

@@ -16,15 +16,7 @@ def run_cmd(cmd):
 # fetch the most recent version tag to use as build version
 def get_latest_git_tag():
     latest_tag = run_cmd("git describe --abbrev=0 --tags")
-    build_version = re.sub(r"v\.?\s*", "", latest_tag)
-    # validate that this is a valid semantic version - will throw exception if not
-    try:
-        import semver
-        semver.VersionInfo.parse(build_version)
-    except (ModuleNotFoundError):
-        # unable to validate because semver is not installed in barebones env for hatch
-        pass
-    return build_version
+    return re.sub(r"v\.?\s*", "", latest_tag)
 
 
 # fetch the most recent build version for hatch environment creation