Offload more work to the underlying pip command used to install the
build requirements, so there's no need to duplicate code to handle
environment markers/extras. This is done by setting the correct options
from the finder passed in argument to `_install_build_reqs`.
- avoid hitting the index: use `common_wheels`/`script.pip_install_local`
- use `script.pip(..., use_module=True)` to simplify some tests
- improve `test_pep518_uses_build_env` parametrization
- simplify `test_pep518_with_user_pip`: we only need to check build
requirements can be installed, so no need for testing with both
`install` and `wheel` command
- fix `test_pip_wheel_with_pep518_build_reqs_no_isolation`: building
pep518-3.0 without isolation should fail if the build requirements
are not installed.
* fix build environment handling when using PyPy
* do use the build environment for all build commands
* allow installing and building a wheel of a PEP 518 enabled
package without prior installation of setuptools and/or wheels
* fix check for minimum supported requirements for PEP 518 support:
- correctly handle complex requirements
- both setuptools and wheels are needed
Mostly indentation complaints, plus an unused test parameter, fixed.
Leaves one new pylint complaint, which is about a test function naming
issue which affects every test.
The list of install options passed into the setup routine is mutable,
passed by reference, so adding items for "this package" to that list
mutates the options for all subsequent packages.
Isolate the lists before mutating them.
Includes a functional test, which has been confirmed to fail without
this fix.
Fixes#3763Fixes#4453Fixes#5089
This test relies on buggy setuptools behaviour (keeping
requirements with environment markers in install_requires,
rather than moving them to extras_require) which has been
fixed in recent setuptools versions.
The use of environment markers implies that the user expects the
packages to not be installed in some cases (eg depending on version
of Python), so the log output shouldn't be classed as a warning,
particularly since this results in it being sent to `stderr` rather
than `stdout`.
Fixes#4876.
This essentially allows me to do an overall check general check by running the tests using pytest's `-k basic` syntax. Given that I like running tests often and that, in general, I make typos more often than changes that break core functionality, I think this will reduce cycle times for me.
