Tagging subscribers to this area: @eerhardt, @dotnet/area-system-text-regularexpressions
See info in area-owners.md if you want to be subscribed.

Any perf numbers?

Any perf numbers?

With this:

[Params(RegexOptions.None, RegexOptions.Compiled)]
public RegexOptions Options { get; set; }

[GlobalSetup]
public void Setup() => _regexes = File.ReadAllLines(@"words.txt").Select(word => new Regex(word, Options)).ToArray();

private readonly string _haystack = File.ReadAllText(@"i386.txt");
private Regex[] _regexes;

[Benchmark]
public int SearchAll()
{
    int sum = 0;
    foreach (Regex r in _regexes)
    {
        sum += Count(r, _haystack);
    }
    return sum;
}

private static int Count(Regex r, string input)
{
    int count = 0;
    Match m = r.Match(input);
    while (m.Success)
    {
        count++;
        m = m.NextMatch();
    }
    return count;
}

where those two words.txt and i386.txt files are the ones from https://github.com/cloudflare/sliceslice-rs/tree/a27b76c8527d44d5b3534c84b878d8289eacb7ff/data, I get these numbers:

Now obviously there are cases where this would do worse, in particular cases where the IndexOf overhead doesn't pay off (like a string of all the same character), but on the whole this seems a better approach. The resulting code is simpler and any improvements we make to IndexOf will implicitly benefit here. Plus we can subsequently change how this is all structured to avoid even creating the Boyer-Moore tables if they're not going to be used, and decrease the ctor costs a bit.