p0

How does this compare to the BPE crate [1]? Its main selling point is support for incrementally re-tokenising text, but it's also faster than tiktoken.

[1] https://crates.io/crates/bpe

Check out https://cs.github.com/about/syntax -- indeed, by default terms are searched in both content and paths. You can restrict to one or the other with `content:` or `path:`.

You could do this with `-path:docs/` (or `NOT path:docs/`).