update constant module to use scipy value and remove using legacy constants

[zhichen3]

added a script to auto generate scipy constant value. Address issue #1655.

[zingale]

I think that if we update G here, then the initial models that do self-gravity would need to be regenerated, since Castro would pick up this G. This is fine, but we should coordinate this carefully.

[zhichen3]

I think that if we update G here, then the initial models that do self-gravity would need to be regenerated, since Castro would pick up this G. This is fine, but we should coordinate this carefully.

Right, the old G is not updated. Only the new G is updated, and that's in comment. So its not being used actually

[zhichen3]

I also updated the burn_cell_metal_chem unit test script to get the "correct" line number. Previously the test actually results in some integration failure, to see that grep for fail in test_*.out. With this pr, we no longer see those failure. So the reference line number changed, resulting in comparison between different quantities. Here I updated it and the test passes.

[zingale]

tests:
http://groot.astro.sunysb.edu/Microphysics/test-suite/gfortran/2025-06-07-001/index.html

[zingale]

I think we will still need to update any Castro initial models that use gamma_law, since the constants are changing.

[zhichen3]

from https://github.com/AMReX-Astro/initial_models, the only one i see is low_mass_convective_star?

[zingale]

oh indeed. I thought it would be more. I'll review this today.

[zingale]

Castro tests: http://groot.astro.sunysb.edu/Castro/test-suite/gfortran/

[zingale]

can you add a page to the docs (perhaps just in the "Microphysics overview" section, that discusses the constants and the script used to generate the header?