lookprime 0.1.0

Ultra-fast odd-only prime lookup with memory-mapped cache

lookprime

Ultra-fast odd-only prime lookup using a byte mask, with memory-mapped cache.

Install

pip install lookprime

CLI

Build (writes to cache by default):

lookprime build --limit 100000000

Benchmark (uses cached mask if present; otherwise builds & caches):

lookprime benchmark --limit 100000000 --duration 1.0 --iterations 25

Benchmark using a specific mask file:

lookprime benchmark --mask /path/to/lookprime_mask_100000000.lpm

Python Usage

from lookprime import open_mask, sieve_to_cache, count_primes_for_duration

path = sieve_to_cache(100_000_000)         # build or reuse cache
mask = open_mask(path)                     # mmap open

hits, max_n = count_primes_for_duration(mask, duration=1.0, start_n=1, chunk=32768)
print(hits, max_n)

mask.close()