Combined CDSAPI requests will return a single combined netcdf; set
cleanup=false to keep the "_tmp_multi_<datetime>_*" files

Note: inpainting is now turned OFF for 2D data -- the reanalysis data
should be complete and turning on inpainting would artificially fill in
data that should be masked (e.g., ocean quantities over land)

Tested for 2D (single levels) and 3D (pressure levels)

to disambiguate from ERA5PressureLevels*

…rly data

…to eq/era5_pressure_levels

…th.jl into eq/era5_pressure_levels

…to eq/era5_pressure_levels

Adopt the bbox→region refactor (NumericalEarth#142): rename `metadata.bounding_box` to
`metadata.region` across ERA5 + CDSAPI extension code, examples, and tests.
Drop the obsolete `location(::ERA5Metadata)` override in favor of the new
generic `dataset_location` dispatch, and the redundant per-Metadatum
`metadata_filename` override (the generic version returns the stored
filename). Adopt `reversed_latitude_axis(::ERA5Dataset) = true` for the new
column-field path's φ-coord flip; ERA5's `retrieve_data` overrides keep
their manual `reverse(data, dims=2)` since they bypass the generic path.
Restore the CDS API environment-setup docstring on `download_dataset(::ERA5Metadatum)`.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Upstream's column-field test used `ERA5Hourly`, which this branch
renamed to `ERA5HourlySingleLevel` (reserving the simpler name for the
abstract dataset family). Update the test accordingly.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

…to eq/era5_pressure_levels

Two stopgaps to keep CI green when CDSAPI credentials aren't configured:

- Surface-level tests: extend `download_dataset_with_fallback` with a
  fallback-of-the-fallback. If both CDS and the NumericalEarthArtifacts
  download fail (e.g., the artifact key was renamed when ERA5 dataset
  types were renamed), it now returns `false` instead of throwing.
  Callers `|| return` to skip the testset cleanly with an `@info` rather
  than cascading file-not-found errors.

- Pressure-level tests: two subtests (`Download and 3D Field`,
  `Geopotential height conversion`) still call `download_dataset`
  directly without the fallback wrapper, so guard them with a
  credentials check that early-returns with an `@info`. TODO comment
  references the long-term fix (upload pressure-level test files to
  NumericalEarthArtifacts and switch to the fallback wrapper).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Surface-level testsets each downloaded `2m_temperature` then removed it,
forcing the next testset to re-download the same bytes — three CDS
round-trips for one fixture. Keep the pre-clean only in the testset
that's testing the download path itself; let downstream testsets reuse
the file and run cleanup in the last consumer.

Pressure-level "Geopotential height conversion" pre-cleaned the
`geopotential_...nc` that the previous testset's `z_interfaces`
side-effect leaves on disk, then re-downloaded it. Drop the pre-clean
and the redundant explicit `download_dataset(meta_z)` call (Field()
already downloads if needed).

Net: surface round-trips drop 3 → 1, geopotential round-trips drop 2 → 1.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Previous commit (c6653b2) folded the skip behavior directly into
download_dataset_with_fallback by changing its return type from the
result of download_fn() to a Bool. That broke other callers — JRA55
tests do `fts = download_dataset_with_fallback(...) do ... end;
@test isfile(fts.path)`, which silently NPEs when fts becomes Bool.

Restore download_dataset_with_fallback to its original semantics
(returns download_fn() result, throws on double-failure). Add a new
try_download_or_skip wrapper that catches the double-failure case,
emits @info + CI warning, and returns false — used only by the CDS
surface tests via `try_download_or_skip(...) do ... end || return`.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>