Create zarr_consolidate_metadata()

[Bisaloo]

TODO:

• Add example
• Add in vignette (deferred to Add "working with groups" section to vignette #185)
• Add attributes to consolidated metadata for v2
• Decide what to do if consolidated metadata already exists
• Add tests
• Update NEWS
• Bump version

Related to #116, HelenaLC/spatialdataR#198

Will also be useful for anndataR, as anndata python is moving towards using consolidated metadata for performance.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 15ca4f4 is merged into devel:

• ✔️pkg_load: 219ms -> 219ms [-0.98%, +0.94%]
• ❗🐌read_blosc_lz4: 231ms -> 232ms [+0.08%, +0.82%]
• ❗🐌read_boolean: 7.36ms -> 7.48ms [+0.19%, +3.02%]
• ✔️read_c_order: 13.2ms -> 13.3ms [-0.14%, +0.94%]
• ✔️read_double: 12.8ms -> 12.8ms [-0.5%, +0.47%]
• ✔️read_empty: 7.74ms -> 7.76ms [-0.57%, +1.11%]
• ✔️read_float16: 9.72ms -> 9.71ms [-1.25%, +1.12%]
• ✔️read_float32: 9.72ms -> 9.65ms [-1.45%, +0.17%]
• ✔️read_int16: 9.63ms -> 9.67ms [-0.48%, +1.21%]
• ✔️read_int32: 12.7ms -> 12.7ms [-0.24%, +0.75%]
• ✔️read_int64: 10.2ms -> 10.2ms [-0.95%, +0.52%]
• ✔️read_int8: 9.57ms -> 9.68ms [0%, +2.16%]
• ✔️read_lz4: 7.53ms -> 7.57ms [-0.29%, +1.34%]
• ✔️read_string: 3.95ms -> 3.92ms [-2.15%, +0.78%]
• ✔️read_uint32: 11.2ms -> 11.2ms [-0.35%, +1.04%]
• ✔️read_uint64: 10.1ms -> 10.1ms [-1%, +1.02%]
• ✔️read_unicode: 3.58ms -> 3.55ms [-3.34%, +1.93%]
• ✔️read_vlen_utf8: 7.86ms -> 7.88ms [-0.6%, +1.19%]
• ✔️read_zlib: 7.57ms -> 7.58ms [-0.72%, +0.95%]
• ✔️read_zstd: 247ms -> 248ms [-0.35%, +1.14%]
• ✔️write_blosc_lz4: 310ms -> 311ms [-1.76%, +2.48%]
• ✔️write_bool: 11.4ms -> 11.4ms [-0.67%, +0.51%]
• ❗🐌write_double: 28.2ms -> 28.6ms [+0.12%, +2.87%]
• ✔️write_float32: 28.2ms -> 28.4ms [-1.65%, +2.54%]
• ✔️write_int16: 28.3ms -> 28.6ms [-0.86%, +3.22%]
• ✔️write_int32: 28.2ms -> 28.2ms [-1.71%, +1.66%]
• ✔️write_int8: 28.2ms -> 28.6ms [-0.69%, +3.66%]
• 🚀write_string: 11.1ms -> 11ms [-1.48%, -0.01%]
• ❗🐌write_unicode: 10.2ms -> 10.3ms [+0.22%, +1.72%]
• 🚀write_zstd: 762ms -> 754ms [-1.97%, -0.16%]
  Further explanation regarding interpretation and methodology can be found in the documentation.