Hello All, *About DataSketches-cpp* This is the core C++ component of the DataSketches library. It includes all of the sketch algorithms as a header-only C++11 library, and it is the basis for the DataSketches Python package.
*Summary* This minor release focuses on correctness and cross-language compatibility with DataSketches-java: - Serialized images of several sketches (Theta, Tuple, HLL, Frequent Items) now match Java's encoding of edge cases such as empty and single-item sketches. - Deserialization of CPC, Count-Min, Theta, Tuple, HLL, Bloom filter and VarOpt sketches now validates header fields and buffer sizes, so malformed or truncated input throws an exception. - New features: an optional trim flag on update_theta_sketch::compact(), count_min_sketch::serialize_to(), a const_iterator over t-digest centroids, and reset() on the frequent items sketch. - Bug fixes for t-digest, REQ, VarOpt, Quantiles and HLL union, plus build fixes for GCC 15. There are no intentional breaking API changes. For details, please see the release notes: https://github.com/apache/datasketches-cpp/releases/tag/5.3.0 *Links* - DataSketches Download Site: https://datasketches.apache.org/docs/Community/Downloads.html - C++ Code Documentation: https://apache.github.io/datasketches-cpp/5.3.0/ - Releases: https://github.com/apache/datasketches-cpp/releases - Developer build and test documentation: https://github.com/apache/datasketches-cpp/blob/5.3.0/README.md - DataSketches Website: https://datasketches.apache.org Thank you to the PMC members and the 16 community contributors to this release! Lee. [email protected]
