1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167

use crate::bloom::BloomFilter;

/// Builder for Bloom Filters.
#[derive(Clone)]
#[derive(Debug)]
pub struct FilterBuilder {
    pub expected_elements: u64,
    pub false_positive_probability: f64,
    pub size: u64,
    pub hashes: u32,
    pub(crate) done: bool,
}

#[cfg(target_pointer_width = "32")]
pub(crate) const SUFFIX: u64 = 0b0001_1111;
#[cfg(target_pointer_width = "64")]
pub(crate) const SUFFIX: usize = 0b0011_1111;
#[cfg(target_pointer_width = "32")]
pub(crate) const MASK: u64 = 0b11111111_11111111_11111111_11111111_11111111_11111111_11111111_11100000;
#[cfg(target_pointer_width = "64")]
pub(crate) const MASK: u64 = 0b11111111_11111111_11111111_11111111_11111111_11111111_11111111_11000000;

/// Calculates the optimal size `m` of the bloom filter in bits given `n` (expected
/// number of elements in bloom filter) and `p` (tolerable false positive rate).
#[inline]
fn optimal_m(n: u64, p: f64) -> u64 {
    let fact = -(n as f64) * p.ln();
    let div = 2f64.ln().powi(2);
    let m: f64 = fact / div;
    let mut m = m.ceil() as u64;
    if (m & SUFFIX as u64) != 0 {
        m = (m & MASK) + SUFFIX as u64 + 1;
    };
    m
}

/// Calculates the optimal `hashes` (number of hash function) given `n` (expected number of
/// elements in bloom filter) and `m` (size of bloom filter in bits).
#[inline]
fn optimal_k(n: u64, m: u64) -> u32 {
    let k: f64 = (m as f64 * 2f64.ln()) / n as f64;
    k.ceil() as u32
}

/// Calculates the amount of elements a Bloom filter for which the given configuration of size `m`
/// and hashes `k` is optimal.
#[inline]
fn optimal_n(k: u32, m: u64) -> u64 {
    let n = (2f64.ln() * m as f64) / k as f64;
    n.ceil() as u64
}


/// Calculates the best-case (uniform hash function) false positive probability.
/// `k` number of hashes.
/// `m` The size of the bloom filter in bits.
/// `n` number of elements inserted in the filter.
#[inline]
fn optimal_p(k: u32, m: u64, n: u64) -> f64 {
    let nk = -(k as f64);
    (1.0 - (nk * n as f64 / m as f64).exp()).powi(k as i32)
}

impl FilterBuilder {
    /// Constructs a new Bloom Filter Builder by specifying the expected size of the filter and the
    /// tolerable false positive probability. The size of the BLoom filter in in bits and the
    /// optimal number of hash functions will be inferred from this.
    ///
    /// # Examples
    ///
    /// ```rust
    /// use fastbloom_rs::FilterBuilder;
    /// let mut builder = FilterBuilder::new(100_000_000, 0.01);
    /// let bloom = builder.build_bloom_filter();
    ///
    /// ```
    pub fn new(expected_elements: u64, false_positive_probability: f64) -> Self {
        FilterBuilder {
            expected_elements,
            false_positive_probability,
            size: 0,
            hashes: 0,
            done: false,
        }
    }

    /// Constructs a new Bloom Filter Builder by specifying the size of the bloom filter in bits
    /// and the number of hashes. The expected size of the filter and the tolerable false positive
    /// probability will be inferred from this.
    pub fn from_size_and_hashes(size: u64, hashes: u32) -> Self {
        let n = optimal_n(hashes, size);
        let p = optimal_p(hashes, size, n);
        FilterBuilder {
            expected_elements: n,
            false_positive_probability: p,
            size,
            hashes,
            done: true,
        }
    }

    /// set the expected size of the filter.
    fn expected_elements(&mut self, expected_elements: u64) {
        assert!(expected_elements > 0, "expected_elements must larger than 0!");
        self.expected_elements = expected_elements;
    }

    /// set the tolerable false positive probability.
    fn false_positive_probability(&mut self, false_positive_probability: f64) {
        assert!(false_positive_probability < 1.0 && false_positive_probability > 0.0,
                "false_positive_probability must between (0.0, 1.0)!");
        self.false_positive_probability = false_positive_probability;
    }

    /// set  the size of the bloom filter in bits.
    fn size(&mut self, size: u64) {
        assert_eq!(size & SUFFIX as u64, 0);
        self.size = size;
    }


    /// Checks if all necessary parameters were set and tries to infer optimal parameters (e.g.
    /// size and hashes from given expected_elements (`n`) and falsePositiveProbability (`p`)).
    /// This is done automatically.
    pub(crate) fn complete(&mut self) {
        if !self.done {
            if self.size == 0 {
                self.size = optimal_m(self.expected_elements, self.false_positive_probability);
                self.hashes = optimal_k(self.expected_elements, self.size);
            }
            self.done = true;
        }
    }

    /// Constructs a Bloom filter using the specified parameters and computing missing parameters
    /// if possible (e.g. the optimal Bloom filter bit size).
    pub fn build_bloom_filter(&mut self) -> BloomFilter {
        self.complete();
        BloomFilter::new(self.clone())
    }

    /// Checks whether a configuration is compatible to another configuration based on the size of
    /// the Bloom filter and its hash functions.
    pub(crate) fn is_compatible_to(&self, other: &FilterBuilder) -> bool {
        self.size == other.size && self.hashes == other.hashes
    }
}

#[test]
fn optimal_test() {
    let m = optimal_m(100_000_000, 0.01);
    let k = optimal_k(100_000_000, m);
    let n = optimal_n(k, m);
    let p = optimal_p(k, m, n);
    println!("{m} {k} {n} {p}");
    assert_eq!(m, 958505856);
    assert_eq!(k, 7)
}

#[test]
fn builder_test() {
    let mut bloom = FilterBuilder::new(100_000_000, 0.01)
        .build_bloom_filter();
    bloom.add(b"helloworld");
    assert_eq!(bloom.contains(b"helloworld"), true);
    assert_eq!(bloom.contains(b"helloworld!"), false);
}