feat: pickle in native byteorder and pass byteorder through pickling and unpickling

[ikrommyd]

If one pickles and unpickles in-memory in a big-endian system, it currently requires two copies. We prevent that by always pickling in the native byteorder. Also in that case, a pickle file created in a system of opposite endianess, will be unpickled incorrectly. We encode the endianess information of the system that the pickle was created inside the pickle so that the system doing the unpickling, knows what byteorder to use when interpreting the bytes.

[github-actions]

The documentation preview is ready to be viewed at http://preview.awkward-array.org.s3-website.us-east-1.amazonaws.com/PR3893

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 82.63%. Comparing base (a6c3b25) to head (48403fa).

Additional details and impacted files

Files with missing lines	Coverage Δ
src/awkward/_pickle.py	70.14% <100.00%> (-8.80%)	⬇️
src/awkward/highlevel.py	78.52% <100.00%> (ø)

... and 2 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[ikrommyd]

This has to be a new "schema 2" in order to be able to unpickle files pickled with previous versions of awkward as the unpickling function is accessed by module and function name.