larzstruct

Data structures the standard library leaves out. Pure Python, zero deps.

A probabilistic set (Bloom filter), a prefix tree (Trie), a union-find (disjoint set), and a fixed-size ring buffer — the ones you reach for often enough to want, without re-implementing them each time.

from larzstruct import BloomFilter, Trie, DisjointSet, RingBuffer

bf = BloomFilter(capacity=10000, error_rate=0.01)
bf.add("alice"); "alice" in bf              # True (never a false negative)

t = Trie(); t.insert("apple"); t.insert("app")
t.starts_with("ap"); t.keys("ap")            # ['app', 'apple']

ds = DisjointSet(); ds.union(1, 2); ds.connected(1, 2)   # True

rb = RingBuffer(3); rb.extend([1, 2, 3, 4]); rb.to_list()  # [2, 3, 4]

What's inside

• BloomFilter — space-efficient set membership; computes the optimal bit-array size and hash count from your capacity + target error_rate. No false negatives; false positives stay under the target. Great for "have I seen this?" at scale.
• Trie — insert, in, starts_with, and keys(prefix) for fast prefix queries and autocomplete.
• DisjointSet (union-find) — union/find/connected/groups with path compression and union by rank. Connectivity, clustering, cycle detection.
• RingBuffer — fixed capacity, appends past capacity overwrite the oldest; iterates oldest-first, with latest(n), is_full(), to_list().

Zero dependencies — pure standard library.

Install

pip install larzstruct

Tests

python -m unittest discover -s tests -v   # 17 tests

The Larz stack

One of 30+ pure-Python, zero-dependency libraries at github.com/larz-scripter.

License

MIT © larz-scripter