Allow Pydantic >= 2.5.3

[camillol]

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 100.00%. Comparing base (6be6136) to head (9d7d71a). Report is 1 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff            @@
##            master     #1388   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files          207       207           
  Lines        14969     14942   -27     
=========================================
- Hits         14969     14942   -27     

Files with missing lines	Coverage Δ
ormar/models/mixins/save_mixin.py	100.00% <100.00%> (ø)
tests/test_fastapi/test_binary_fields.py	100.00% <100.00%> (ø)
...definition/test_overwriting_pydantic_field_type.py	100.00% <ø> (ø)

... and 12 files with indirect coverage changes

[codspeed-hq[bot]]

CodSpeed Performance Report

Merging #1388 will degrade performances by 18.61%

_{Comparing camillol:unpinpydantic (9d7d71a) with master (6be6136)}

Summary

⚡ 2 improvements
❌ 5 (👁 5) regressions
✅ 77 untouched benchmarks

Benchmarks breakdown

	Benchmark	master	camillol:unpinpydantic	Change
⚡	test_min[250]	3.6 ms	2.8 ms	+28.63%
👁	test_deleting_all[500]	2.4 ms	2.9 ms	-18.61%
👁	test_deleting_individually[10]	8.7 ms	9.8 ms	-12.12%
⚡	test_get_all_with_related_models[10]	14.1 ms	12.5 ms	+12.55%
👁	test_get_one[1000]	3.4 ms	4.1 ms	-16.94%
👁	test_get_or_create_when_get[1000]	3.5 ms	4.1 ms	-14.81%
👁	test_saving_models_individually[10]	12.9 ms	14.4 ms	-10.8%

[collerek]

That's still installing 2.5.3 in tests, might be cross dependency with pydantic-extra or something else. As it mingles with pydantic internals each upper version should be checked.

Edit: see that you didn't run poetry lock that's why

[camillol]

Ran poetry lock!

[camillol]

@collerek there is one broken test, tests/test_fastapi/test_binary_fields.py::test_read_main. It's not clear to me that the behavior expected by the test is correct; it looks like it wants the JSON response to represent the LargeBinary field as its binary contents decoded into a string with the default encoding (utf-8), but why should it? Could you take care of it?