A pure python bloom filter (low storage requirement, probabilistic set datastructure) is provided. It is known to work on CPython 3.x, Pypy, and Jython.

Includes mmap, in-memory and disk-seek backends.

This project builds on drs-bloom-filter and bloom_filter_mod. Credits and links can be found in AUTHORS.md.

The user specifies the desired maximum number of elements and the desired maximum false positive probability, and the module calculates the rest.

from bloom_filter2 import BloomFilter

# instantiate BloomFilter with custom settings,
# max_elements is how many elements you expect the filter to hold.
# error_rate defines accuracy; You can use defaults with
# `BloomFilter()` without any arguments. Following example
# is same as defaults:
bloom = BloomFilter(max_elements=10000, error_rate=0.1)

# Test whether the bloom-filter has seen a key:
assert "test-key" not in bloom

# Mark the key as seen
bloom.add("test-key")

# Now check again
assert "test-key" in bloom

Bloom filter are pretty space efficient : only 200MB of memory usage for storing 100M elements with an error of 1%, compared to the 7GB required for set(range(10**8))

It still can be pretty useful to save/load to files with the mmap implementation, for example to avoid rebuilding the bloom filter. The mmap functionality also save some memory depending on system settings.

bloom = BloomFilter(max_elements=10**8, error_rate=0.01, filename=('/tmp/bloom.bin', -1))