Legacy Fortran to Python mapping¶

Upstream Fortran source¶

Primary upstream reference:

external/odt1/source1/

Current mapping¶

Fortran	Python
`Brng.f`	`pyodt1.rng.OdtRNG`
`BSeeds.f`	`pyodt1.rng.OdtRNG(seed_index=...)`
`BExp.f`	`pyodt1.advance.sample_exponential_wait()`
`BInitRun.f`	`pyodt1.advance.generate_initial_state()`
`BInitIter.f`	`pyodt1.advance.compute_initial_dt()`, `pyodt1.advance.compute_td()`
`BAdv.f`	`pyodt1.advance.advance_diffusion()`
`BEqnStep.f`	`pyodt1.advance.equation_step()`
`BLenProb.f`	`pyodt1.eddy_sampling.build_length_distribution()`
`BLength.f`	`pyodt1.eddy_sampling.sample_length()`
location sampling in `BSampleEddy.f`	`pyodt1.eddy_sampling.sample_location()`
`BsKd.f`	`pyodt1.eddy_sampling.bs_kd()`
`BProb.f`	`pyodt1.acceptance.acceptance_probability()`
`BTriplet.f`	`pyodt1.triplet.triplet_map()`
`BAddK.f`	`pyodt1.triplet.add_k()`
`BEddy.f`	`pyodt1.solver.OdtSolver.apply_eddy()`
`BSetOld.f`	`pyodt1.statistics.save_old_values()`
`BChange.f`	`pyodt1.statistics.accumulate_change()`
`BStats.f`	`pyodt1.statistics.accumulate_cstats()`
`BSeries.f`	`pyodt1.statistics.accumulate_series()`
`BSnap.f`	`pyodt1.statistics.compute_snap_outputs()`, `pyodt1.statistics.write_snap_intercomparison()`, `pyodt1.statistics.write_snap_xmgrace()`
`BRecord.f`	`pyodt1.statistics.brecord_text()`
`XRecord.f`	`pyodt1.statistics.xrecord_text()`
`BWriteSeries.f`	`pyodt1.statistics.finalize_series_variance()`, `pyodt1.statistics.write_series_text()`
`BReadOptions.f`	`pyodt1.config.read_options()`, `pyodt1.legacy.b_read_options()`
`BrngGet.f`	`pyodt1.rng.OdtRNG.get_state()`, `pyodt1.legacy.brng_get()`
`BrngPut.f`	`pyodt1.rng.OdtRNG.put_state()`, `pyodt1.legacy.brng_put()`
`BInitStats.f`	`pyodt1.legacy.b_init_stats()`
`BAddTerm.f`	`pyodt1.legacy.b_add_term()`
top-level legacy case execution in `Bodt.f`	`pyodt1.legacy.run_legacy_case()`
`BLowerdt.f`	`pyodt1.solver.OdtSolver.lower_dt()`
`BRaisedt.f`	`pyodt1.solver.OdtSolver.raise_dt()`
simple realization scheduling in `Bodt.f`	`pyodt1.solver.OdtSolver.run_realization()`
`nstat` / `ntseg` sub-interval scheduling in `Bodt.f`	`pyodt1.solver.OdtSolver.run_scheduled_realization()`
repeated realization / series aggregation path in `Bodt.f`	`pyodt1.solver.OdtSolver.run_iterations()`

Notes¶

This mapping now covers most of the numerically important Bodt.f execution path: initialization, eddy sampling, acceptance, triplet application, adaptive dt, deterministic advancement, repeated realizations, time statistics, change statistics, and major postprocessing/output helpers.

The main remaining gaps or caveats are:

top-level legacy file-opening / file-closing orchestration in Bodt.f is not mirrored as a single monolithic Python driver
top-level legacy file-opening / file-closing orchestration in Bodt.f is not mirrored line-for-line, though pyodt1.legacy.run_legacy_case() now provides a legacy-style end-to-end case runner that reads a case directory, writes the expected legacy file set, and emits fort.11
the external legacy sources in this repository now include a minimal robustness patch for the intercomparison BSnap / BRecord path; the patched path is validated, while the original unmodified negative-N variant is documented separately via scripts/investigate_bsnap_intercomparison.py as compiler-fragile legacy behavior