Nice changes, fwiw I reread the docs and I think explanations of Stride(and renaming it from BucketSize afaik) in documentation is a big improvement. As for clang-tidy <https://github.com/boostorg/bloom/pull/32/commits/8b467fc87039a0e02984cee00338e081c675df23>I would assume that just using force inline macro would be cleaner, but not sure. On Tue, Jun 24, 2025 at 11:46 PM Joaquin M López Muñoz via Boost < boost@lists.boost.org> wrote:

Hi,

For those interested, I've just merged a branch incorporating most of the feedback received during the review of Boost.Bloom. Arnauld Becheler has once again been tremendously helpful with the process. More details here:

https://github.com/boostorg/bloom/pull/32

https://github.com/boostorg/bloom/issues?q=is%3Aissue%20state%3Aclosed%20aut...

Revised documentation here:

https://32.bloom.prtest3.cppalliance.org/libs/bloom/doc/html/index.html

Boost.Bloom will be released with Boost 1.89. Thanks to all the people who have contributed to its improvement!

Joaquin M Lopez Munoz

PS. List of changes:

* removed superfluous inline (Alexander Grund) * made hasher equivalence a precondition for &=/|= (Andrzej Krzemienski) * documented exception safety guarantees (Andrzej Krzemienski) * mentioned Bloom filters are called so after Burton H Bloom (Dmitry Arkhipov) * added warning about OOM for very small FPR (Ivan Matek) * stressed config chart x axis is capacity/num elements rather than plain capacity (Ivan Matek) * s/[SIMD] is available/is enabled at compile time (Ivan Matek) * shut down clang-tidy warnings (Ivan Matek) * used "set union" for more clarity (Andrzej Krzemienski) * stressed early on that boost::bloom::filter is _not_ a container (Claudio DeSouza) * added bulk operations to roadmap (Dmitry Arkhipov) * added try_insert to roadmap (Konstantin Savvidy) * added estimated_size to roadmap (Konstantin Savvidy) * added alternative filters to roadmap (Konstantin Savvidy) * used <cstdint> instead of <boost/cstdint.hpp> (Rubén Pérez) * mentioned endianness when serializing filters (Rubén Pérez) * corrected sloppiness about optimum k determination (Tomer Vromen) * added run-time specification of k to roadmap (Tomer Vromen) * added test/CMakeLists.txt (Rubén Pérez) * added CMake-based testing to GHA (Rubén Pérez) (#8) * added <boost/bloom.hpp> (Rubén Pérez) * added Codecov reporting (Rubén Pérez) (#9) * moved from boost::unordered::hash_is_avalanching to ContainerHash's boost::hash_is_avalanching (Ivan Matek/Peter Dimov) * added syntax highlighting to code snippets (Rubén Pérez) * avoided C-style casts in examples (Rubén Pérez) * added acknowledgements section (Peter Turcan) * added Getting Started section (Peter Turcan) * fixed example Jamfile and added example building to CI (Rubén Pérez) (#10) * added diagram about overlapping vs. non-overlapping subarrays (Rubén Pérez/Ivan Matek/Vinnie Falco) * made first code snippet self-contained (Rubén Pérez/Peter Turcan) * added more comments to genome.cpp (Rubén Pérez) * added support for arrays as blocks (Tomer Vromen) (#24) * removed emplace (Seth Heeren/Peter Dimov) (#25) * required the allocator to be of unsigned char (Seth Heeren/Peter Dimov) (#26) * added compile-time validation of Block types (Rubén Pérez) (#27) * added value type to displayed filter names in tables (Tomer Vromen) (#28) * used -march=native rather than -mavx2 (Ivan Matek) * adopted hash strategy with fastrange plus a separate MCG (Kostas Savvidis/Peter Dimov) (#30) * made doc improvements (Peter Turcan/Arnaud Becheler) (#29) * several maintenance commits

_______________________________________________ Boost mailing list -- boost@lists.boost.org To unsubscribe send an email to boost-leave@lists.boost.org https://lists.boost.org/mailman3/lists/boost.lists.boost.org/ Archived at: https://lists.boost.org/archives/list/boost@lists.boost.org/message/JA5YSHG5...