sidd-27 / bloomsday

What is bloomsday?

Bloomsday delivers a Rust crate for building blocked Bloom filters that handle high-throughput membership queries—like checking if a key might exist in a massive set—without storing the actual data. It implements the Parquet Split Block Bloom Filter format for cache-friendly lookups on modern CPUs, solving space-efficient deduplication in data pipelines or caches. Developers get a simple API to insert keys (strings, ints, anything Hash-able) via fast xxHash or raw u64 hashes, plus optional serde for serialization.

Why is it gaining traction?

It clocks 2.3x faster lookups than sbbf-rs in benchmarks for 1M items at 1% false positives, thanks to compiler auto-vectorization that kicks in with native CPU flags. Zero runtime deps beyond std and xxhash-rust keep it lean for high performance backend github setups, while staying 100% safe Rust—no unsafe code to debug. Rust high performance github fans dig the raw hash API for custom integrations like high performance mysql github or blocked filters in streams.

Who should use this?

Backend engineers at scale building high performance computing github services, like deduping logs in Spark jobs or filtering queries in Rust high performance github apps. Database devs needing quick existence checks without full indexes, or network pros approximating sets in high performance browser networking pdf github-inspired stacks. Ideal for blocked Bloom use in Bloomsday Ireland events tracking or Bloomsday festival 2026 attendee lists.

Verdict

With 12 stars and 1.0% credibility score, this v0.1.0 crate is early but promising—solid tests, benchmarks, and README make it worth a cargo bench trial for perf-critical paths. Use it if your workload matches the negatives/positives tested; otherwise, stick to battle-tested alternatives until more adoption hits. (198 words)