Combine rstrip calls in trailing_whitespace check

[correctmost]

This provides a small speed-up on large codebases (~90ms).

Stats

Before

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
  1609756    0.432    0.000    0.432    0.000 {method 'rstrip' of 'str' objects}

Command	Mean [s]	Min [s]	Max [s]	Relative
pycodestyle .	18.178 ± 0.198	17.951	18.666	1.00

After

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
  1137564    0.311    0.000    0.311    0.000 {method 'rstrip' of 'str' objects}

Command	Mean [s]	Min [s]	Max [s]	Relative
pycodestyle .	18.090 ± 0.155	17.784	18.264	1.00

Set-up

I profiled pycodestyle with the yt-dlp codebase because it is similar in composition to a private codebase I have.

git clone https://github.com/yt-dlp/yt-dlp.git
cd yt-dlp

git checkout ef36d517f9b05785d61abca7691d9ab7d63cc75c

# Callgraph command
python -m cProfile -o stats $(which pycodestyle)

# Benchmarking command
hyperfine --ignore-failure --warmup 2 --runs 10 --export-markdown=baseline.md 'pycodestyle .'

setup.cfg

[pycodestyle]
ignore = W504
max-line-length = 100

[asottile]