Memory usage optimization via reuse of `SchemaValidator` and `SchemaSerializer` #5607

Job	Run time
build on ${{ matrix.os }} (${{ matrix.target }} - ${{ matrix.interpreter \|\| 'all' }}${{ (((matrix.os == 'linux') && format(' - {0}', (((matrix.manylinux == 'auto') && 'manylinux') \|\| matrix.manylinux))) \|\| '') }})	0s
rust benchmarks	3m 7s
build sdist	17s
build pgo-optimized on ${{ matrix.os }} / ${{ matrix.interpreter }}	0s
build-wasm-emscripten	2m 50s
coverage	2m 4s
lint	56s
test-debug 3.13	1m 6s
test-debug pypy3.10	1m 24s
test MSRV	2m 31s
test on ubuntu	2m 31s
test on macos	2m 47s
test-pydantic-integration	1m 22s
test on windows	4m 55s
test 3.9	2m 10s
test 3.10	2m 4s
test 3.11	2m 7s
test 3.12	2m 5s
test 3.13	2m 7s
test pypy3.9	2m 23s
test pypy3.10	2m 27s
test build on ${{ matrix.target }}-${{ matrix.distro }}	0s
test build on ${{ matrix.os }}	0s
inspect-pypi-assets	0s
check	2s
release	0s
	41m 15s

Provide feedback