Follow-up housekeeping from the review of #1729, which moved update_GDDLAI into src/suews/src/suews_phys_lai.f95. None of these affects behaviour, so they did not hold that PR up.
1. Duplicate entry in src/supy/meson.build
suews_phys_lai.f95 is listed twice in src_f95_all (lines 258 and 267). That list only feeds depend_files: on the suews_objects custom target, so it is a rebuild-watch list rather than a compile list and the duplicate is inert. The second entry should still go; line 258 is the right slot, since it keeps the module ahead of suews_phys_dailystate.f95.
2. Imports stranded by the move
src/suews/src/suews_phys_lai.f95 imports supy_error_flag and never uses it.src/suews/src/suews_phys_dailystate.f95 still imports set_supy_error and IEEE_IS_NAN; both left with the moved block. supy_error_flag is still used there and should stay.
gfortran does not warn on unused only: names, so nothing in CI catches these.
3. make format has been failing silently
src/suews/Makefile calls:
fprettify --config .fprettify.rc src/suews/src/*.f95 2>/dev/null || true
fprettify 0.3.7 has no --config option, so this exits with a usage error that 2>/dev/null || true swallows whole. fprettify already discovers .fprettify.rc by walking up from the file, so dropping the flag is the fix.
This is the root cause of item 4, and likely of drift elsewhere in the tree.
4. Formatting drift in suews_phys_lai.f95
With the config applied, fprettify src/suews/src/suews_phys_lai.f95 rewrites 145 of 489 lines, mostly four-space module-level indent against the configured indent = 3, and clears 17 lines of trailing whitespace. Most of this travelled with the block rather than starting in #1729: suews_phys_dailystate.f95 goes from 168 non-conforming lines to 58 once the block leaves. Worth a pass over both files once item 3 is fixed.
5. Reviewer routing is stale
dev-ref/SCIENTIFIC_REVIEWERS.md routes this code through the suews_phys_dailystate.f95 row, which no longer holds it. Needs a row for suews_phys_lai.f95.
Not in scope here
The helpers in the new module all stay nested inside update_gddlai under CONTAINS, so the module exposes one entry point. Promoting them to module-level procedures would make them independently testable, but that is a design question for whoever continues the LAI work rather than housekeeping.
Follow-up housekeeping from the review of #1729, which moved
update_GDDLAIintosrc/suews/src/suews_phys_lai.f95. None of these affects behaviour, so they did not hold that PR up.1. Duplicate entry in
src/supy/meson.buildsuews_phys_lai.f95is listed twice insrc_f95_all(lines 258 and 267). That list only feedsdepend_files:on thesuews_objectscustom target, so it is a rebuild-watch list rather than a compile list and the duplicate is inert. The second entry should still go; line 258 is the right slot, since it keeps the module ahead ofsuews_phys_dailystate.f95.2. Imports stranded by the move
src/suews/src/suews_phys_lai.f95importssupy_error_flagand never uses it.src/suews/src/suews_phys_dailystate.f95still importsset_supy_errorandIEEE_IS_NAN; both left with the moved block.supy_error_flagis still used there and should stay.gfortran does not warn on unused
only:names, so nothing in CI catches these.3.
make formathas been failing silentlysrc/suews/Makefilecalls:fprettify 0.3.7 has no
--configoption, so this exits with a usage error that2>/dev/null || trueswallows whole. fprettify already discovers.fprettify.rcby walking up from the file, so dropping the flag is the fix.This is the root cause of item 4, and likely of drift elsewhere in the tree.
4. Formatting drift in
suews_phys_lai.f95With the config applied,
fprettify src/suews/src/suews_phys_lai.f95rewrites 145 of 489 lines, mostly four-space module-level indent against the configuredindent = 3, and clears 17 lines of trailing whitespace. Most of this travelled with the block rather than starting in #1729:suews_phys_dailystate.f95goes from 168 non-conforming lines to 58 once the block leaves. Worth a pass over both files once item 3 is fixed.5. Reviewer routing is stale
dev-ref/SCIENTIFIC_REVIEWERS.mdroutes this code through thesuews_phys_dailystate.f95row, which no longer holds it. Needs a row forsuews_phys_lai.f95.Not in scope here
The helpers in the new module all stay nested inside
update_gddlaiunderCONTAINS, so the module exposes one entry point. Promoting them to module-level procedures would make them independently testable, but that is a design question for whoever continues the LAI work rather than housekeeping.