AI review of 2026-11-05_14-57-37-8ef59912f47a-results.txt
PR:          #8997 https://github.com/cppcheck-opensource/cppcheck/pull/8997
Tested:      8ef59912f47af48927c51e733a00b2f00e5b41d6
Merge base:  a7b9e1cc22dd4d6f345dfec61d6172bb3253bad2
Reviewed:    2026-11-06 01:52:46 UTC
Model:       claude-opus-5-5 (effort high)
Results:     2 reviewed of 2 in the report
Verdicts:    2 regression
Tokens:      2428 input, 1731 cache read, 1731 cache write, 1823 output

The verdicts are written by AI and can be wrong.

---- 1 / 2 ----
Verdict:     REGRESSION (high confidence)
Package:     https://ftp.debian.org/debian/pool/main/n/nettle4/nettle4_4.0.orig.tar.gz
Result:      your nettle-4.0/testsuite/testutils.h:46:19: warning: Member variable 'tstring::next' has no initializer. [uninitMemberVarNoCtor]
Explanation: `struct tstring` is a plain C-style struct (inside `extern "C"`) with no constructor and no default member initializers. The members are `next`, `length` and a flexible array member `data[]`. `uninitMemberVarNoCtor` is meant for classes where some members have initializers and others do not. The new hack makes `isInitialized` return true for arrays declared with `[]` or `[0]`. That apparently makes the flexible array count as an initialized member, so the check now flags the ordinary members of a POD struct. Nothing in this struct is initialized in-class, so this new warning is a false positive.

---- 2 / 2 ----
Verdict:     REGRESSION (high confidence)
Package:     https://ftp.debian.org/debian/pool/main/n/nettle4/nettle4_4.0.orig.tar.gz
Result:      your nettle-4.0/testsuite/testutils.h:47:10: warning: Member variable 'tstring::length' has no initializer. [uninitMemberVarNoCtor]
Explanation: `tstring` is a plain C struct inside `extern "C"`. It has no constructor and no in-class initializers: `next`, `length`, and the flexible array member `data[]`. cppcheck normally does not report `uninitMemberVarNoCtor` for aggregates where no member is initialized. The PR now treats the flexible array `data[]` as 'initialized'. That makes the struct look partly initialized, so the other members are wrongly flagged. Nothing in this C-style struct is meant to have a default initializer, so the new warning is a false positive.