Transcription Tracking Issue

[mabruzzo]

This issue is designed to track progress and coordination of the transcription efforts.

The (provisional)¹ plan is:

• treat the gen2024 branch as a development branch (the rationale for this is provided in newchem-cpp Tracking Issue #255)
• essentially, we will make pull requests of transcribed functionality into the gen2024 branch
• at some point in the future, the gen2024 branch will be merged into main (in which case, any unmerged transcription PRs will need to be shifted to main)
• for convenience a branch with all/most transcribed functions (that haven't been merged upstream yet) can be found here

Routines Currently being Transcribed (1/32):

• reserved for: @ChristopherBignamini:

  • rate_timestep_g
    • file: solve_rate_cool_g.F
    • wrapper-function: grackle::impl::fortran_wrapper::rate_timestep_g

• reserved for @mabruzzo:

  • step_rate_g
    • file: solve_rate_cool_g.F
    • wrapper-function: grackle::impl::fortran_wrapper::step_rate_g
    • special considerations: try not to refactor this too much (We want to eventually factor out common logic duplicated by this routine and the grackle::impl::chemistry::species_density_derivatives_0d

Routines ready for transcription (5/32):

Leave a comment on this issue if you want to reserve one of these cases

Routines that will be ready after cool1d_multi_g is transcribed: (4/32)

• cool1d_cloudy_g
  • file: cool1d_cloudy_g.F
  • already existing wrapper: grackle::impl::fortran_wrapper::cool1d_multi_g
  • remaining call sites that must be wrapped: cool1d_multi_g
  • Special Considerations: An existing PR previously tried transcribing the logic ( Port calc_temp1d_cloudy_g & cool1d_cloudy_g from Fortran to C #153 ). I suspect that we don't want to try to reuse that transcribed logic.² But, if it's not too hard, we may want to repurpose the unit-test logic.
• calc_temp1d_cloudy_g
  • file: calc_temp1d_cloudy_g.F
  • already existing wrapper: grackle::impl::fortran_wrapper::calc_temp_cloudy_g
  • remaining call sites that must be wrapped: cool1d_multi_g
  • Special Considerations: An existing PR previously tried transcribing the logic ( Port calc_temp1d_cloudy_g & cool1d_cloudy_g from Fortran to C #153 ). I suspect that we don't want to try to reuse that transcribed logic.² But, if it's not too hard, we may want to repurpose the unit-test logic.
• cool1D_cloudy_old_tables_g
  • file: cool1d_cloudy_old_tables_g.F
  • call sites that must be wrapped: cool1d_multi_g
• calc_all_tdust_gasgr_1d_g
  • file: calc_all_tdust_gasgr_1d_g.F
  • already existing wrapper: grackle::impl::fortran_wrapper::calc_all_tdust_gasgr_1d_g
  • remaining call sites that must be wrapped: cool1d_multi_g.F

Routines not yet ready for transcription: (6/32)

These routines all have 1 or more Fortran call-sites (i.e. they are called from 1 or more other Fortran routines). As such they are not yet ready for transcription.

• if a routine has multiple call-sites, we won't be ready to transcribe it until a wrapper function has been created and every call-site calls the wrapper function
• if a routine only has 1 Fortran call-site, we can transcribe it immediately after transcribing the call-site.

These cases include:

• calc_grain_size_increment_1d
  • file: calc_grain_size_increment_1d.F
  • already existing wrapper: grackle::impl::fortran_wrapper::calc_grain_size_increment_1d
  • remaining call sites that must be wrapped: cool1d_multi_g.F and lookup_cool_rates1d_g
• calc_tdust_1d_g
  • file: calc_tdust_1d_g.F
  • call sites that must be wrapped: (UNKNOWN)
• calc_kappa_gr_g
  • file: calc_tdust_1d_g.F
  • call sites that must be wrapped: (UNKNOWN)
• calc_gr_balance_g
  • file: calc_tdust_1d_g.F
  • call sites that must be wrapped: (UNKNOWN)
• calc_grain_size_increment_species_1d
  • file: calc_grain_size_increment_1d.F
  • call sites that must be wrapped: (UNKNOWN)
• solve_cubic_equation
  • file: calc_grain_size_increment_1d.F
  • call sites that must be wrapped: (UNKNOWN)

Transcribed Routines Awaiting Review (16/32)

• interpolate_1D_g
  • PR: Transcribe interpolate routines brittonsmith/grackle#38³
• interpolate_2D_g
  • PR: Transcribe interpolate routines brittonsmith/grackle#38³
• interpolate_3D_g
  • PR: Transcribe interpolate routines brittonsmith/grackle#38³
• interpolate_3Dz_g
  • PR: Transcribe interpolate routines brittonsmith/grackle#38³
• interpolate_2Df3D_g
  • PR: Transcribe interpolate routines brittonsmith/grackle#38³
• interpolate_4D_g
  • PR: Transcribe interpolate routines brittonsmith/grackle#38³
• interpolate_5D_g
  • PR: Transcribe interpolate routines brittonsmith/grackle#38³
• scale_fields_g
  • file: scale_fields_g.F
  • PR: [newchem-cpp] transcribe scale_fields_g #406
• cool1d_multi_g
  • file: cool1d_multi_g.F
  • wrapper-function: grackle::impl::fortran_wrapper::cool1d_multi_g
  • PR: [newchem-cpp] transcribe cool1d_multi_g #405
• ceiling_species_g[newchem-cpp] transcribe ceiling_species #408
• make_consistent_g [newchem-cpp] transcribe make consistent #417
• gaussj_g [newchem-cpp] transcribe gaussj #393
• lookup_cool_rates1d_g [newchem-cpp] transcribe lookup_rate_1d_g #412, [newchem-cpp] cleanup lookup_cool_rate_1d part 1 #415, [newchem-cpp] lookup_cool_rate1d cleanup part 2 #416

Merged Transcriptions:

• calc_tdust_3d_g
  • PR: [newchem-cpp] transcribe calc_tdust_3d #385
• cool_multi_time_g
  • PRs: cleanup calc_tdust_3d_g brittonsmith/grackle#20 (Consistency tweaks: GrainSpeciesCollection and cool_multi_time_g brittonsmith/grackle#25 applies some subsequently established conventions)
• solve_rate_cool_g
  • PRs: transcribe: solve_rate_cool_g to C++ brittonsmith/grackle#24 (should probably be broken into pieces)
• step_rate_newton_raphson
  • PRs: initial transcription of step_rate_newton_raphson brittonsmith/grackle#29, step_rate_newton_raphson cleanup (1/2) brittonsmith/grackle#30
• lookup_cool_rates0d
  • PRs: transcribe lookup_cool_rates0d 1/2 brittonsmith/grackle#31, transcribe lookup_cool_rates0d 2/2 brittonsmith/grackle#33
• calc_temp_cloudy_g
  • PR: transcribe calc_temp_cloudy_g brittonsmith/grackle#34
• scale_fields_table_g
  • PR: transcribe scale_fields_table brittonsmith/grackle#35

Footnotes

1. It's provisional until @brittonsmith has a moment to clarify his thought on how he thinks we should handle the details of development ↩

2. The referenced PR was made before we started our current efforts. Consequently, it will take some work to extract just the logic we want. Thus, it may not be worth the effort to reuse the logic ↩ ↩²

3. As explained in its description, Transcribe interpolate routines brittonsmith/grackle#38 is a very special PR. The PR justifies why it breaks many conventions for what makes a good transcription PR. Because of its special status, it may not be merged until all other routines have been transcribed. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

[ChristopherBignamini]

@mabruzzo , thank you very much for this useful list. I pushed the transcribe of cool1d_multi_g and will start working on ceiling_species_g.