Result: done
Files: crates/nikaia-std/src/tools/keeps.nika crates/nikaia-std/src/tools/lowered/keeps.rs tests/compiler/tests/coalesce_on_a_lent_tuple_part.nika tests/compiler/tests/coalesce_on_a_lent_tuple_part.stdout
Commit: keeps: a part moved out by ?? keeps its parameter, a tuple part as a field (#891)
Body: The left side of ?? is classified with keep_part instead of keep_named, so a tuple part (p.0, p.1.0) keeps its parameter as a struct field does: the parameter is taken, and the lowering’s unwrap_or_else no longer moves out of a view. A ?? passed at a position the callee’s result points into (shown(p ?? "none")) is walked side by side instead, so its parameter stays lent; before, fn whole(p: String?) -> String { shown(p ?? "none") } was refused with NK2302. Steps 1-3 in one commit.
Fixes: #891
Pushed: 3ba85829773378489473bd053285c67f3991844e
Tests: new coalesce_on_a_lent_tuple_part failed on the unfixed compiler (NK2302 on shown(p ?? ..); without that form rustc’s “cannot move out of p.0/p.1.0/s.name”), passes now; nikaia test –both-settings tests/compiler 176 passed; lower-std changed only lowered/keeps.rs (the fix itself), fixpoint
Gate: green
Deviation: none
Found: none
Needs: none
Changelog: ?? on a part of a lent tuple parameter (p.0 ?? "..") no longer fails in rustc
Owner: none
Disk freed: 0
