Result: done
Files: crates/nikaia-std/src/tools/check_takes.nika crates/nikaia-std/src/tools/lowered/check_takes.rs tests/compiler/tests/takes_an_optional_handed_over_twice_says_its_copy.nika tests/compiler/tests/takes_an_optional_handed_over_twice_says_its_copy.stderr
Commit: check_takes: NK2105’s help writes t?.clone() for an optional handed over twice
Body: The plain-call NK2105 (data_handed_over_is_used_again) wrote {name}.clone() whatever the type, which the compiler refuses on a T?. A shared helper a_copy_said(path, ty) now gives ?.clone() when the type’s text ends with ?, used by this site and the repeat-frame refusal that already did so inline. (a_copy_of was taken by threads.nika.) No existing .stderr changes: the three with this help line hand over non-optionals.
Fixes: #1056
Tests: new takes_an_optional_handed_over_twice_says_its_copy passes; its output on the unfixed binary differs on the help line (t.clone()), so it fails there; t?.clone() in the probe compiles and runs (prints 1, 1); tests/compiler –both-settings 456 passed, 0 failed; packages/yaml 14 passed; lower-std fixpoint reached; clippy and pre-push-check clean; fmt::every_nika_file not run (no file under packages/ changed)
Gate: green
Deviation: none
Found: none
Needs: none
Changelog: NK2105’s help for an optional handed over and read again suggests t?.clone(), the copy that compiles, instead of t.clone() (#1056).
Owner: none
Disk freed: 0 GB (cache 6.5G kept, 3.9G free on /; scratch binary copy removed)
Rs2nika: none
Pushed: local 9c7f0298e6024c2c10b83c614d7caa9a80b2e4f9
