80 findings, not the 14 the first run showed -- clippy stops at the first failing compilation unit, so `--keep-going` is what makes the list complete. 60 were machine-applicable (`cargo clippy --fix`). The rest by hand: * five descending `sort_by` -> `sort_by_key(Reverse(..))` * `chunks_exact(4)` on both sides of four zips, so the compared items stay `[u8; 4]` rather than one array against one slice * three `type` aliases for the census maps and the captured-quad tuple * `&PathBuf` -> `&Path` in two disc tests * two range loops; one of them keeps `#[allow(needless_range_loop)]` with the reason -- the index is into a map's value, which changes each iteration * the module doc list in `invert_capture` re-indented to markdown's rules * `blit`'s eight arguments get `#[allow(too_many_arguments)]`, not a struct One dead `let off = b.len();` in a `ratc` test is dropped rather than renamed. The sibling test at :162 is the one that asserts an offset; if this one was meant to as well, that is a test change and not a lint fix. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2.7 KiB
2.7 KiB