Struct roaring_bloom_filter::BloomFilter

pub struct BloomFilter { /* private fields */ }

A variant of bloom filter

Every hash function has it’s own slice instead of sharing the whole bitmap. This introduce the possibility of concurrency of manipulating multiply hash function at the same time.

Implementations

impl BloomFilter

pub fn from_scratch(
    slices_length: u32,
    slice_size: u64,
    target_false_positive_rate: f64
) -> BloomFilter

Create an empty bloom filter from scratch.

Generally, user should not use this initializer directly. Promise the limitation on yourself:

• 0 < k <= u32::MAX
• 0 < m <= u32::MAX
• 0 < f < 1

pub fn new(max_size: u64, target_false_positive: f64) -> BloomFilter

Create an empty bloom filter with max element’s size and false positive rate. The crate would calculate the best buckets length and bucket size.

pub fn add<T>(&mut self, value: &T) -> bool where
    T: Hash, 

Add new element into the bloom filter. Return true when any key are inserted in a slice.

pub fn contains<T>(&mut self, value: &T) -> bool where
    T: Hash, 

Check if the bloom filter contains the specific key. Return true when all key are present in all slices, which may contains false positive situation.

pub fn target_false_positive_rate(&self) -> f64

Get target false positive rate.

pub fn current_false_positive_rate(&self) -> f64

Get current false positive rate.

pub fn is_empty(&self) -> bool

If this bloom filter is empty.

pub fn is_full(&self) -> bool

If this bloom filter is full.

pub fn size(&self) -> u64

Get the number of inserted elements in this bloom filter.

pub fn len(&self) -> u64

Get the number of inserted bits in all slices.

Trait Implementations

impl Display for BloomFilter

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.