Read NPY chunk at a time

[aziz-mu]

This PR solves (1) from the issue #1670, although some design changes were made from the issue:
(i) We are keeping mmap for now as it simplifies reading the header
(ii) We are using arrow::RecordBatch to match the design of ReadFile

[codecov]

Codecov Report

Patch coverage: 100.00% and no project coverage change.

Comparison is base (251d978) 91.22% compared to head (dc8745b) 91.22%.

❗ Current head dc8745b differs from pull request most recent head 94d01a1. Consider uploading reports for the commit 94d01a1 to get more accurate results

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1697   +/-   ##
=======================================
  Coverage   91.22%   91.22%           
=======================================
  Files         737      737           
  Lines       26987    26994    +7     
=======================================
+ Hits        24619    24626    +7     
  Misses       2368     2368           

Impacted Files	Coverage Δ
src/include/storage/copier/npy_reader.h	100.00% <ø> (ø)
...age/in_mem_storage_structure/in_mem_column_chunk.h	100.00% <ø> (ø)
src/storage/copier/node_copier.cpp	62.93% <100.00%> (+0.65%)	⬆️
src/storage/copier/npy_reader.cpp	90.74% <100.00%> (+2.10%)	⬆️
...e/in_mem_storage_structure/in_mem_column_chunk.cpp	81.62% <100.00%> (+0.16%)	⬆️

... and 8 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.