build wheels using Github actions, add test

[vertiond]

Part 2 of breaking up #1

This is set up to build wheels on push/pull and deploy to PyPI on tags starting with v using pypi_password in secrets

12 wheels produced in 13 minutes:
  ltc_scrypt-1.0-cp310-cp310-win32.whl
  ltc_scrypt-1.0-cp310-cp310-win_amd64.whl
  ltc_scrypt-1.0-cp36-cp36m-win32.whl
  ltc_scrypt-1.0-cp36-cp36m-win_amd64.whl
  ltc_scrypt-1.0-cp37-cp37m-win32.whl
  ltc_scrypt-1.0-cp37-cp37m-win_amd64.whl
  ltc_scrypt-1.0-cp38-cp38-win32.whl
  ltc_scrypt-1.0-cp38-cp38-win_amd64.whl
  ltc_scrypt-1.0-cp39-cp39-win32.whl
  ltc_scrypt-1.0-cp39-cp39-win_amd64.whl
  ltc_scrypt-1.0-pp37-pypy37_pp73-win_amd64.whl
  ltc_scrypt-1.0-pp38-pypy38_pp73-win_amd64.whl

python test.py 
('ltc_scrypt', '53232d90adebc88ab0855434cfaf6822fce47c62b2bdd7e8d84f667868a1ee40')

[xanimo]

Concept ACK. test.py works as expected (from reviewing previous PR) and CI run was successful despite no pypi module deployment: https://github.com/xanimo/ltc-scrypt/actions/runs/2771596763

Outside of that I'm curious on whether or not this needs multi-arch support (noting the build on windows-2019) and/or if you've any thoughts regarding that subject?

concept ACK != approval ❤️

[xanimo]

Concept ACK. test.py works as expected (from reviewing previous PR) and CI run was successful despite no pypi module deployment: https://github.com/xanimo/ltc-scrypt/actions/runs/2771596763

Outside of that I'm curious on whether or not this needs multi-arch support (noting the build on windows-2019) and/or if you've any thoughts regarding that subject?

Thanks for mitigating my mistake @patricklodder 😸 ❤️

That aside, I've done some research and experimentation's with regards to multi-arch builds and have done a couple CI runs with your (@vertiond ) other changes (PR #7) cherry-picked on top and then with some amendments to the .yml itself and have successfully built for win, macos and 'manylinux'. It also appears pypa/cibuildwheel supports Linux ARM as well so perhaps we could potentially integrate that as well. More on those changes can be found here.

7 wheels produced in 9 minutes:
  ltc_scrypt-1.0-cp310-cp310-macosx_10_9_x86_64.whl
  ltc_scrypt-1.0-cp36-cp36m-macosx_10_9_x86_64.whl
  ltc_scrypt-1.0-cp37-cp37m-macosx_10_9_x86_64.whl
  ltc_scrypt-1.0-cp38-cp38-macosx_10_9_x86_64.whl
  ltc_scrypt-1.0-cp39-cp39-macosx_10_9_x86_64.whl
  ltc_scrypt-1.0-pp37-pypy37_pp73-macosx_10_9_x86_64.whl
  ltc_scrypt-1.0-pp38-pypy38_pp73-macosx_10_9_x86_64.whl

...

12 wheels produced in 10 minutes:
  ltc_scrypt-1.0-cp310-cp310-win32.whl
  ltc_scrypt-1.0-cp310-cp310-win_amd64.whl
  ltc_scrypt-1.0-cp36-cp36m-win32.whl
  ltc_scrypt-1.0-cp36-cp36m-win_amd64.whl
  ltc_scrypt-1.0-cp37-cp37m-win32.whl
  ltc_scrypt-1.0-cp37-cp37m-win_amd64.whl
  ltc_scrypt-1.0-cp38-cp38-win32.whl
  ltc_scrypt-1.0-cp38-cp38-win_amd64.whl
  ltc_scrypt-1.0-cp39-cp39-win32.whl
  ltc_scrypt-1.0-cp39-cp39-win_amd64.whl
  ltc_scrypt-1.0-pp37-pypy37_pp73-win_amd64.whl
  ltc_scrypt-1.0-pp38-pypy38_pp73-win_amd64.whl

...

24 wheels produced in 6 minutes:
  ltc_scrypt-1.0-cp310-cp310-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl
  ltc_scrypt-1.0-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl
  ltc_scrypt-1.0-cp310-cp310-musllinux_1_1_i686.whl
  ltc_scrypt-1.0-cp310-cp310-musllinux_1_1_x86_64.whl
  ltc_scrypt-1.0-cp36-cp36m-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl
  ltc_scrypt-1.0-cp36-cp36m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl
  ltc_scrypt-1.0-cp36-cp36m-musllinux_1_1_i686.whl
  ltc_scrypt-1.0-cp36-cp36m-musllinux_1_1_x86_64.whl
  ltc_scrypt-1.0-cp37-cp37m-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl
  ltc_scrypt-1.0-cp37-cp37m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl
  ltc_scrypt-1.0-cp37-cp37m-musllinux_1_1_i686.whl
  ltc_scrypt-1.0-cp37-cp37m-musllinux_1_1_x86_64.whl
  ltc_scrypt-1.0-cp38-cp38-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl
  ltc_scrypt-1.0-cp38-cp38-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl
  ltc_scrypt-1.0-cp38-cp38-musllinux_1_1_i686.whl
  ltc_scrypt-1.0-cp38-cp38-musllinux_1_1_x86_64.whl
  ltc_scrypt-1.0-cp39-cp39-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl
  ltc_scrypt-1.0-cp39-cp39-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl
  ltc_scrypt-1.0-cp39-cp39-musllinux_1_1_i686.whl
  ltc_scrypt-1.0-cp39-cp39-musllinux_1_1_x86_64.whl
  ltc_scrypt-1.0-pp37-pypy37_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl
  ltc_scrypt-1.0-pp37-pypy37_pp73-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl
  ltc_scrypt-1.0-pp38-pypy38_pp73-manylinux_2_5_i686.manylinux1_i686.manylinux_2_17_i686.manylinux2014_i686.whl
  ltc_scrypt-1.0-pp38-pypy38_pp73-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Outside that I attempted to bump to 3.10 and use pypa and am a bit perplexed into how long it's taking for Windows 2019 but considering we'd probably only have to build once in a blue moon, it's not that big of a deal.

With that said, I'd like to hear both your thoughts on this and if we should get them in now or just support Windows at this current juncture? Also in consideration of the fact my questions about multi arch are irrelevant for use in core since we'll be fetching an archive and building ourselves, I don't necessarily think this should be blocking, but the feature in and of itself could be more robust I suppose.