tests: switch to new GRASS test project with srid

[metzm]

The previous GRASS test project nc_spm_full_v2alpha2 did not have a spatial reference ID (SRID) and had thus an incomplete CRS definition with incorrect datum. The official sample datasets nc_basic_spm_grass7 and nc_spm_08_grass7 do have a SRID set (EPSG:3358). A side effect is that import of data into the test project that have been exported from the sample datasets would fail with strict testing for matching CRSs. This PR uses the new nc_spm_full_v2beta1 test project created and uploaded by @neteler and me.

[metzm]

Test of v.import fails now with

FAIL: test_import_gpkg_wgs84 (__main__.TestVImport.test_import_gpkg_wgs84)
Import GPKG in same proj, default params
[...]
mismatch values (key, reference, actual): [('east', 644450.6, 644450.838504622), ('north', 227744.8, 227743.97201618), ('south', 215259.6, 215258.834830312), ('west', 631257.4, 631257.659484619)]

the easiest fix would be to adjust the reference values in the test?

[wenzeslaus]

Thanks!

[metzm]

Related tests have been adjusted, but two tests are failing

on ubuntu i.smap, test_with_goodness_map

mismatch values (key, reference, actual): [('max', 4.495414, 3.60080480575562)]

his could be related to recent parallelization of r.mapcalc which seems to be a bit unstable and is not related to the modified CRS of the test dataset.

on macOS r.proj, test_print_output_json
For north the test got 35.80961644859372 but expected 35.80961644859374

This is the 16th siginifcant digit and thus at the border of floating point precision limits, the precision for comparison should be reduced.

@wenzeslaus can you please approve the latest changes? Thanks!

[echoix]

i.smap is flaky for months, just a rerun would work. If the 16th digit is the problem with these changes, just adjust the expected precision

[metzm]

All checks and tests have been passed, finally! Don't touch anymore, approve please, then it will be merged and I can proceed with other PRs depending on this one. This was much more effort than initially expected.