GH-115060: Speed up pathlib.Path.glob() by not scanning literal parts

[barneygale]

Don't bother calling os.scandir() to scan for literal pattern segments, like foo in foo/*.py. Instead, append the segment(s) as-is and call through to the next selector with exists=False, which signals that the path might not exist. Subsequent selectors will call os.scandir() or os.lstat() to filter out missing paths as needed.

Timings:

$ ./python -m timeit -s "from pathlib import Path; p = Path.cwd()""list(p.glob('Lib'))"
5000 loops, best of 5: 69.4 usec per loop
20000 loops, best of 5: 13.6 usec per loop
# --> 5.1x faster
$ ./python -m timeit -s "from pathlib import Path; p = Path.cwd()""list(p.glob('Lib/'))"
5000 loops, best of 5: 73.3 usec per loop
20000 loops, best of 5: 14.2 usec per loop
# --> 5.16x faster
$ ./python -m timeit -s "from pathlib import Path; p = Path.cwd()""list(p.glob('Lib/*'))"
1000 loops, best of 5: 362 usec per loop
1000 loops, best of 5: 301 usec per loop
# --> 1.2x faster
$ ./python -m timeit -s "from pathlib import Path; p = Path.cwd()""list(p.glob('Lib/*/__init__.py'))"
200 loops, best of 5: 1.18 msec per loop
1000 loops, best of 5: 273 usec per loop
# --> 4.32x faster
$ ./python -m timeit -s "from pathlib import Path; p = Path.cwd()""list(p.glob('Lib/**/__init__.py'))"
50 loops, best of 5: 9.46 msec per loop
50 loops, best of 5: 5.72 msec per loop
# --> 1.65x faster
$ ./python -m timeit -s "from pathlib import Path; p = Path.cwd()""list(p.glob('Lib/pathlib/__init__.py'))"
1000 loops, best of 5: 210 usec per loop
20000 loops, best of 5: 14.9 usec per loop
# --> 14.1x faster

• Issue: Speed up pathlib.Path.glob() by removing redundant regex matching #115060