gh-139156: Use PyBytesWriter in PyUnicode_AsUnicodeEscapeString()

[vstinner]

Replace PyBytes_FromStringAndSize() and _PyBytes_Resize() with the PyBytesWriter API.

• Issue: Use PyBytesWriter in Unicode codecs #139156

[vstinner]

Benchmark:

import pyperf
runner = pyperf.Runner()
sizes = (3, 100, 1000)
for size in sizes:
    runner.timeit(f'{size:,} ASCII chars',
        setup=f's="x"*{size}',
        stmt='s.encode("unicode_escape")')
for size in sizes:
    runner.timeit(f'{size:,} UCS-1 chars',
        setup=f's=chr(0xe9) * {size}',
        stmt='s.encode("unicode_escape")')
for size in sizes:
    runner.timeit(f'{size:,} UCS-2 chars',
        setup=f's=chr(0x20ac) * {size}',
        stmt='s.encode("unicode_escape")')
for size in sizes:
    runner.timeit(f'{size:,} UCS-4 chars',
        setup=f's=chr(0x10ffff) * {size}',
        stmt='s.encode("unicode_escape")')

Results:

Benchmark	ref	pep782
3 ASCII chars	528 ns	515 ns: 1.03x faster
100 ASCII chars	639 ns	663 ns: 1.04x slower
1,000 ASCII chars	1.67 us	1.70 us: 1.02x slower
3 UCS-1 chars	503 ns	515 ns: 1.02x slower
100 UCS-1 chars	758 ns	773 ns: 1.02x slower
3 UCS-2 chars	502 ns	520 ns: 1.04x slower
100 UCS-2 chars	926 ns	939 ns: 1.01x slower
1,000 UCS-2 chars	4.64 us	4.66 us: 1.00x slower
3 UCS-4 chars	508 ns	524 ns: 1.03x slower
100 UCS-4 chars	1.03 us	1.04 us: 1.02x slower
1,000 UCS-4 chars	5.08 us	5.11 us: 1.01x slower
Geometric mean	(ref)	1.01x slower

Benchmark hidden because not significant (1): 1,000 UCS-1 chars