Add close() and flush() methods to stream classes for full file-like compatibility

[khsrali]

These test on aiida-core are flaky

FAILED tests/engine/processes/calcjobs/test_calc_job.py::TestCalcJob::test_get_hash - AttributeError: 'PackedObjectReader' object has no attribute 'close'
FAILED tests/engine/processes/calcjobs/test_calc_job.py::TestCalcJob::test_get_objects_attributes - AttributeError: 'PackedObjectReader' object has no attribute 'close'
FAILED tests/engine/processes/calcjobs/test_calc_job.py::TestCalcJob::test_compute_hash_version_independent - AttributeError: 'PackedObjectReader' object has no attribute 'close'
FAILED tests/engine/processes/calcjobs/test_calc_job.py::test_monitor_version - AttributeError: 'PackedObjectReader' object has no attribute 'close'
FAILED tests/engine/processes/calcjobs/test_calc_job.py::test_monitor_result_override_exit_code - AttributeError: 'PackedObjectReader' object has no attribute 'close'
FAILED tests/engine/processes/calcjobs/test_calc_job.py::test_monitor_result_action_disable_all - AttributeError: 'PackedObjectReader' object has no attribute 'close'
FAILED tests/engine/processes/calcjobs/test_calc_job.py::test_monitor_result_action_disable_self - AttributeError: 'PackedObjectReader' object has no attribute 'close'
FAILED tests/engine/processes/calcjobs/test_calc_job.py::test_restart_after_daemon_reset - AssertionError: The process excepted: AttributeError: 'PackedObjectReader' object has no attribute 'flush'

The flakiness comes down to whether objects are loose or packed in the disk-objectstore at the time the test reads from the repository. The point is disk-objectstore packs objects automatically when certain thresholds are reached.
If the packed object are returned they still should support close() and flush(), becausein aiida-core TextIOWrapper is used as a context manager, its exit calls close(), which propagates to the underlying stream.

The commit adda no-op close() and flush() methods to PackedObjectReader, CallbackStreamWrapper, and ZlibLikeBaseStreamDecompresser.

This was discovered in aiidateam/aiida-core#7206, where async changes results in different tests ordering or something resulting to this issue to pop out.
The methods are intentionally no-ops because these readers don't own the underlying file handles they read from.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.63%. Comparing base (e5cb644) to head (71ae6bc).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #205   +/-   ##
=======================================
  Coverage   99.63%   99.63%           
=======================================
  Files           8        8           
  Lines        1932     1938    +6     
=======================================
+ Hits         1925     1931    +6     
  Misses          7        7           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[agoscinski]

I thought about implementing a protocol to enforce this somehow like it was done in #194. But in the end the bug comes from its usage in aiida-core where ignore the type error. I move the consideration of some enforcement through a protocol to PR #194 because I am not really sure what interface we want to enforce.

EDIT: Looking at aiida-core, I think we could solve the issue by just implementing io.BinaryIO Leaving it for PR #194

[khsrali]

Ok, agreed with @agoscinski to merge this now and consider the io.BinaryIO idea for #194