matrix: inconsistent test

[SunCatMC]

The test "extract column from diff widths matrix" is not the same as the test "extract column where numbers have different widths" in problem-specification, as it has multiple spaces in between matrix values, compared to original one-space test.

I don't know how you handle differences in specification and actual exercises, but if it's ok to have them, there should be consistency, as you can have a solution that accounts for multiple spaces in columns but doesn't in rows (as the test for rows "extract row from diff widths matrix" is exactly the same as the one in specification)

[ErikSchierboom]

Tracks are in principle free to diverge from the canonical data as defined in prob-specs. This is usually only done when a test doesn't fit a track. The exercise should indeed be consistent. Would you be interested in submitting a PR to fix this?

[SunCatMC]

Sure, I can submit a PR

I guess changing it to match the specification is better, as this way you don't get a non-obvious point of failure for such a simple puzzle

[ErikSchierboom]

Agreed! That would definitely be best.