blob: 6d804a135402f30d27908d5109eef2609ff2af0b (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
|
# Development tasks. Everything runs through `uv`, so there is nothing to
# install first and no virtualenv to activate by hand.
#
# make list the targets
# make test run the suite on PYTHON (default: the minimum supported)
# make lint ruff check + format check, no files touched
# make build wheel + sdist into dist/
#
# PYTHON pins the interpreter uv resolves. It defaults to 3.9 — the minimum
# this package claims to support — because that is the version most likely to
# catch a syntax or stdlib assumption that a newer one silently accepts.
PYTHON ?= 3.9
# Every 3.x this package claims to support, for `make test-all`. Keep in sync
# with requires-python in pyproject.toml.
PYTHON_VERSIONS := 3.9 3.10 3.11 3.12 3.13 3.14
UV ?= uv
RUN := $(UV) run --python $(PYTHON) --with-editable .
.DEFAULT_GOAL := help
.PHONY: help test test-all lint format build clean check
help:
@echo 'Targets:'
@echo ' test run the test suite (PYTHON=$(PYTHON))'
@echo ' test-all run it on every supported Python: $(PYTHON_VERSIONS)'
@echo ' lint ruff check + ruff format --check (read-only)'
@echo ' format apply ruff format and ruff check --fix (writes)'
@echo ' build build the wheel and sdist into dist/'
@echo ' check lint + test + build, in that order'
@echo ' clean remove build artefacts and caches'
@echo
@echo 'Override the interpreter with e.g. make test PYTHON=3.13'
test:
$(RUN) python -m unittest discover -s tests -v
# Sequential, not parallel: uv may need to download an interpreter, and
# interleaved unittest output from six runs is unreadable when one fails.
# `set -e` inside the loop is what makes a failure on 3.10 stop the run instead
# of being masked by a later success.
test-all:
@set -e; for v in $(PYTHON_VERSIONS); do \
echo "==> Python $$v"; \
$(UV) run --python $$v --with-editable . python -m unittest discover -s tests; \
done
# Read-only by design, so it is safe in CI and in a pre-commit hook: it reports
# what `make format` would change without changing anything.
lint:
$(UV)x ruff check .
$(UV)x ruff format --check .
format:
$(UV)x ruff format .
$(UV)x ruff check --fix .
build:
$(UV) build
# The order matters: lint is the fastest to fail, build the slowest, so a
# broken run reports the cheapest problem first.
check: lint test build
clean:
rm -rf dist build .ruff_cache .venv
find . -name '__pycache__' -type d -prune -exec rm -rf {} +
find . -name '*.egg-info' -type d -prune -exec rm -rf {} +
|