fix: wk_trans inconsistent meta flags handling

[anthonynorth]

Fixes wk_trans inconsistent meta flags handling, which can result in garbage values in the m dimension when input doesn't contain m dimension and use_m = TRUE.

Fixes #216

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.85%. Comparing base (67f7013) to head (9ee8dd2).

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #217   +/-   ##
=======================================
  Coverage   98.85%   98.85%           
=======================================
  Files          85       85           
  Lines        6211     6211           
=======================================
  Hits         6140     6140           
  Misses         71       71           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[anthonynorth]

I wasn't entirely certain if transformers should see dimensions that aren't present in the output. E.g. if input contains z but use_z = FALSE, should a transformer get z = NaN, or z from input?

I've assumed the intent was to forward input as is to transformers (thus ignoring use_z/m) and omit transformed z/m if use_z/m = FALSE.

Similarly, if use_z/m = TRUE and these dimensions aren't in the input, transformer is handed a default z/m = NaN. doesn't see a junk z/m, but the next handler in the pipeline sees z/m = NaN.

[paleolimbot]

Thanks! I'm actually not entirely sure what I intended when I wrote this, so I trust your interpretation. In any case, the change you made seems to correct a bit of code that is obviously out of place from the surrounding code.