Use Rf_dimgets() directly do set dims

[Bisaloo]

No description provided.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 1fcee50 is merged into main:

• ✔️pkg_load: 4.93ms -> 4.94ms [-1.02%, +1.37%]
• ✔️read_bigendian: 11.4ms -> 11.5ms [-0.19%, +1.21%]
• ❗🐌read_bool: 11.3ms -> 11.4ms [+0.15%, +2%]
• ✔️read_empty: 11.5ms -> 11.4ms [-0.92%, +0.46%]
• ✔️read_float32: 11.3ms -> 11.4ms [-0.69%, +1.08%]
• ✔️read_float64: 11.5ms -> 11.5ms [-0.83%, +0.53%]
• ✔️read_int16: 11.4ms -> 11.4ms [-0.67%, +1.28%]
• ✔️read_int32: 11.4ms -> 11.3ms [-0.86%, +0.36%]
• ✔️read_int64: 11.9ms -> 11.9ms [-0.2%, +1%]
• ✔️read_int8: 11.5ms -> 11.4ms [-1.72%, +0.52%]
• ✔️read_npz: 13.2ms -> 13.2ms [-0.74%, +0.91%]
• ✔️read_string: 11.3ms -> 11.3ms [-0.55%, +0.63%]
• ✔️read_structured: 12.2ms -> 12.1ms [-2.69%, +0.58%]
• ✔️read_uint16: 11.4ms -> 11.4ms [-1%, +0.84%]
• ✔️read_uint32: 11.7ms -> 11.7ms [-0.85%, +0.96%]
• ✔️read_uint64: 11.9ms -> 12ms [-0.22%, +1.38%]
• ✔️read_uint8: 11.3ms -> 11.3ms [-0.76%, +0.4%]
• ✔️read_unicode: 11.3ms -> 11.4ms [-0.32%, +0.8%]
  Further explanation regarding interpretation and methodology can be found in the documentation.

[github-actions]

This is how benchmark results would change (along with a 95% confidence interval in relative change) if 418e4fa is merged into main:

• ✔️pkg_load: 5.47ms -> 5.53ms [-1.21%, +3.47%]
• ✔️read_bigendian: 12.1ms -> 12ms [-1.19%, +0.15%]
• ✔️read_bool: 12.2ms -> 12.2ms [-0.95%, +0.97%]
• ✔️read_empty: 12ms -> 12ms [-1.03%, +0.65%]
• ✔️read_float32: 12.2ms -> 12.2ms [-0.7%, +0.73%]
• ✔️read_float64: 12.2ms -> 12.2ms [-0.53%, +0.89%]
• ✔️read_int16: 12.4ms -> 12.3ms [-1.73%, +0.74%]
• ✔️read_int32: 12.5ms -> 12.5ms [-0.43%, +1.47%]
• ✔️read_int64: 13ms -> 12.7ms [-4.74%, +0.75%]
• ✔️read_int8: 12.4ms -> 12.3ms [-3%, +0.58%]
• ✔️read_npz: 13.7ms -> 13.8ms [-0.15%, +1.2%]
• ✔️read_string: 12.2ms -> 12.2ms [-1.38%, +0.92%]
• ✔️read_structured: 12.8ms -> 12.8ms [-0.34%, +0.93%]
• ✔️read_uint16: 12.6ms -> 12.6ms [-0.32%, +1.35%]
• ✔️read_uint32: 12.7ms -> 12.6ms [-1.51%, +0.56%]
• ✔️read_uint64: 12.9ms -> 12.9ms [-1.34%, +0.33%]
• ✔️read_uint8: 12.8ms -> 12.8ms [-0.88%, +1.02%]
• ✔️read_unicode: 12.2ms -> 12.2ms [-1.13%, +0.52%]
  Further explanation regarding interpretation and methodology can be found in the documentation.