fix: reject negative proton isotope indexes

[carrerasdarren-cell]

Summary

• prevent a negative proton isotope delta from indexing before
  nNumRemovedProtonsIsotopic
• add a GetStructFromINCHI() regression for public OSS-Fuzz issue
  488729177

Root cause

The single-proton cleanup path checked only that iso_atw_diff was no larger
than NUM_H_ISOTOPES. Because the field is signed, the malformed input
InChI=1/T/q+1/i12-112T produced iso_atw_diff == -112, passed that upper
bound, and accessed the three-element array at index -113.

The public testcase reproduces on current dev as an ASan two-byte stack
out-of-bounds read in mark_alt_bonds_and_taut_groups(). Requiring a positive
delta before subtracting one makes the API reject the malformed string with
its normal inchi_Ret_ERROR result.

Verification

• exact OSS-Fuzz testcase: ASan/UBSan failure before the patch, clean rejection
  after the patch
• focused API regression passes under ASan
• all 17 C++ unit-test executables pass under ASan
• CLI suite: 15 passed, 3 skipped, 10 xfailed
• both CLI and libinchi targets build successfully
• git diff --check passes

Type of change

• feat - new feature (minor bump)
• fix - bug fix (patch bump)
• perf - performance improvement (patch bump)
• BREAKING CHANGE - incompatible change
• refactor / docs / chore / build / ci / test (no release)

Checklist

• PR title follows Conventional Commits
• Builds cleanly for both targets (./INCHI-1-TEST/build_with_cmake.sh all)
• Compiles without new warnings on GCC/Clang/MSVC where applicable
• Code follows .editorconfig (4-space indent, LF, no trailing whitespace)
• Unit tests pass (ctest in the full_build test dir)
• CLI tests pass (pytest INCHI-1-TEST/tests/test_executable)
• New behavior is covered by tests
• Public API / behavior changes are documented (not applicable)

Related issues

Refs #209 (GOF #488729177 only; the issue also tracks two unrelated leaks).

[fbaensch-beilstein]

Thanks for the fix