perf: rewrite `not_m_separated_for_all_subsets()` in Rust #11

	name: benchmark

	on:
	pull_request:

	permissions: read-all

	jobs:
	benchmark:
	runs-on: ubuntu-latest

	steps:
	- uses: actions/checkout@v6

	- uses: r-lib/actions/setup-r@v2
	with:
	r-version: release
	use-public-rspm: true

	- name: Install dependencies
	run: Rscript -e 'install.packages(c("remotes", "bench", "caugi", "optparse", "tibble"))'

	- name: Install PR branch package
	run: Rscript -e 'remotes::install_local(".", upgrade="never")'

	- name: Run benchmark on PR branch
	run: \|
	Rscript .github/workflows/benchmark.R

	- uses: actions/checkout@v6
	with:
	ref: main
	path: main_branch

	- name: Run benchmark on main branch
	run: \|
	Rscript .github/workflows/benchmark.R --output main_time.rds

	- name: Compare results
	run: \|
	Rscript -e '
	pr <- readRDS("pr_time.rds")
	main <- readRDS("main_time.rds")

	slowdown_median <- (pr$median - main$median) / main$median * 100

	# Make a table
	result_table <- tibble::tibble(
	Benchmark = names(pr$expression),
	PR_median = pr$median,
	PR_total_time = pr$total_time,
	Main_median = main$median,
	Main_total_time = main$total_time,
	Slowdown_percent_median = as.numeric(slowdown_median)
	)

	cat("\n=== Benchmark Comparison ===\n")
	print(result_table, row.names = FALSE, width = Inf)

	if (any(slowdown_median >= 20)) {
	cat("\nERROR: PR branch is >=20% (median time) slower than main for at least one benchmark entry\n")
	stop("Benchmark check failed")
	} else {
	cat("\nBenchmark check passed: PR is not more than 20% (median time) slower than main\n")
	}
	'

Provide feedback