[pya]

Close #2878

On Windows, MFSimulation.load() fails for a model in a subdirectory when the
OPEN/CLOSE paths of its external files use forward slashes, e.g. models written
by iMOD Python. Flopy prepends the model directory a second time and cannot find
the file (level2\level2/dat/gwf.dis_botm.txt). MODFLOW 6 accepts these files, and
flopy loads them on macOS and Linux.

Cause

The check added in #1298 (fix for #1289), which decides whether the model relative
path still has to be prepended, split paths on os.path.sep. On Windows that is \,
so a path with / comes back as a single item and never matches the model directory.

Changes

• flopy/mf6/data/mfdatastorage.py (array data) and
  flopy/mf6/data/mfdataplist.py (list data, two places): split both paths with
  PureWindowsPath(...).parts, which accepts / and \ on every OS, and prepend
  the model relative path only if the file path does not start with all of its
  parts:

  fp_parts = PureWindowsPath(fp_relative).parts
  rp_parts = PureWindowsPath(rel_path).parts
  if fp_parts[: len(rp_parts)] != rp_parts:
      fp_relative = os.path.join(rel_path, fp_relative)

  Comparing the leading parts as a whole also fixes two edge cases of the old loop
  with multi-level model relative paths: an IndexError when the file path has
  fewer parts than the model relative path (e.g. k.dat with models/gwf), and
  prepending only some components (e.g. gwf/k.dat with models/gwf became
  models/gwf/k.dat).

• autotest/test_mf6.py: new test_load_subdir_ext_files_with_any_sep,
  parametrized with posix and win separators. It writes a model with
  model_rel_path="level2" and external array (DIS) and list (CHD) data, rewrites
  the OPEN/CLOSE paths to level2<sep>dat<sep><file>, reloads the simulation and
  checks botm and the CHD stress period data. The win case is skipped on macOS
  and Linux, where \ is not a path separator.

Testing

On Windows 11, Python 3.14, MODFLOW 6.8.2:

• Without the fix, test_load_subdir_ext_files_with_any_sep[posix] fails with the
  doubled model directory; [win] passes. With the fix, both pass.
• Related existing tests pass: test_subdir, test_basic_gwf,
  test_load_sim_when_namefile_uses_rel_paths.
• Smoke tests (pytest -n auto -S): the tests that fail with this change fail
  identically without it. They need executables missing in my environment
  (e.g. gridgen, mf2005).
• ruff check . and ruff format --check . pass.

Not tested on macOS or Linux; CI will cover these.

[codecov]

Codecov Report

❌ Patch coverage is 71.42857% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 72.7%. Comparing base (556c088) to head (a10000f).
⚠️ Report is 249 commits behind head on develop.

Files with missing lines	Patch %	Lines
flopy/mf6/data/mfdataplist.py	55.5%	4 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #2879      +/-   ##
===========================================
+ Coverage     55.5%    72.7%   +17.1%     
===========================================
  Files          644      659      +15     
  Lines       124135   132499    +8364     
===========================================
+ Hits         68947    96373   +27426     
+ Misses       55188    36126   -19062     

Files with missing lines	Coverage Δ
flopy/mf6/data/mfdatastorage.py	70.7% <100.0%> (-3.3%)	⬇️
flopy/mf6/data/mfdataplist.py	72.2% <55.5%> (-6.7%)	⬇️

... and 584 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[wpbonelli]

Maybe worth a small function somewhere for the identical code? No strong feelings

I also wonder if we should really be comparing case sensitively, but that is a separate question from this bug

Thanks for this @pya

[wpbonelli]

the CI failures are unrelated

[pya]

@wpbonelli Good points.

I though about a separate function that takes care of such kind of path handling. Maybe there are other places that do similar things. I was not sure if the reuse potential is high enough yet. Maybe worth some refactoring that makes all path handling more consistent.

As for case sensitiv, as far as I know, only Linux is truly case sensitive whereas Windows and macOS are only case preserving. I think using pathlib as much as possible should help to take care of this. Another TODO for the refactoring. ;)