aboutsummaryrefslogtreecommitdiff
path: root/src/bloom.h
blob: 9173b80d66d081e853a1ee02d7e698b2dd57e3e5 (plain)
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
// Copyright (c) 2012-2019 The Bitcoin Core developers
// Distributed under the MIT software license, see the accompanying
// file COPYING or http://www.opensource.org/licenses/mit-license.php.

#ifndef BITCOIN_BLOOM_H
#define BITCOIN_BLOOM_H

#include <serialize.h>

#include <vector>

class COutPoint;
class CTransaction;
class uint256;

//! 20,000 items with fp rate < 0.1% or 10,000 items and <0.0001%
static const unsigned int MAX_BLOOM_FILTER_SIZE = 36000; // bytes
static const unsigned int MAX_HASH_FUNCS = 50;

/**
 * First two bits of nFlags control how much IsRelevantAndUpdate actually updates
 * The remaining bits are reserved
 */
enum bloomflags
{
    BLOOM_UPDATE_NONE = 0,
    BLOOM_UPDATE_ALL = 1,
    // Only adds outpoints to the filter if the output is a pay-to-pubkey/pay-to-multisig script
    BLOOM_UPDATE_P2PUBKEY_ONLY = 2,
    BLOOM_UPDATE_MASK = 3,
};

/**
 * BloomFilter is a probabilistic filter which SPV clients provide
 * so that we can filter the transactions we send them.
 *
 * This allows for significantly more efficient transaction and block downloads.
 *
 * Because bloom filters are probabilistic, a SPV node can increase the false-
 * positive rate, making us send it transactions which aren't actually its,
 * allowing clients to trade more bandwidth for more privacy by obfuscating which
 * keys are controlled by them.
 */
class CBloomFilter
{
private:
    std::vector<unsigned char> vData;
    unsigned int nHashFuncs;
    unsigned int nTweak;
    unsigned char nFlags;

    unsigned int Hash(unsigned int nHashNum, const std::vector<unsigned char>& vDataToHash) const;

public:
    /**
     * Creates a new bloom filter which will provide the given fp rate when filled with the given number of elements
     * Note that if the given parameters will result in a filter outside the bounds of the protocol limits,
     * the filter created will be as close to the given parameters as possible within the protocol limits.
     * This will apply if nFPRate is very low or nElements is unreasonably high.
     * nTweak is a constant which is added to the seed value passed to the hash function
     * It should generally always be a random value (and is largely only exposed for unit testing)
     * nFlags should be one of the BLOOM_UPDATE_* enums (not _MASK)
     */
    CBloomFilter(const unsigned int nElements, const double nFPRate, const unsigned int nTweak, unsigned char nFlagsIn);
    CBloomFilter() : nHashFuncs(0), nTweak(0), nFlags(0) {}

    ADD_SERIALIZE_METHODS;

    template <typename Stream, typename Operation>
    inline void SerializationOp(Stream& s, Operation ser_action) {
        READWRITE(vData);
        READWRITE(nHashFuncs);
        READWRITE(nTweak);
        READWRITE(nFlags);
    }

    void insert(const std::vector<unsigned char>& vKey);
    void insert(const COutPoint& outpoint);
    void insert(const uint256& hash);

    bool contains(const std::vector<unsigned char>& vKey) const;
    bool contains(const COutPoint& outpoint) const;
    bool contains(const uint256& hash) const;

    //! True if the size is <= MAX_BLOOM_FILTER_SIZE and the number of hash functions is <= MAX_HASH_FUNCS
    //! (catch a filter which was just deserialized which was too big)
    bool IsWithinSizeConstraints() const;

    //! Also adds any outputs which match the filter to the filter (to match their spending txes)
    bool IsRelevantAndUpdate(const CTransaction& tx);
};

/**
 * RollingBloomFilter is a probabilistic "keep track of most recently inserted" set.
 * Construct it with the number of items to keep track of, and a false-positive
 * rate. Unlike CBloomFilter, by default nTweak is set to a cryptographically
 * secure random value for you. Similarly rather than clear() the method
 * reset() is provided, which also changes nTweak to decrease the impact of
 * false-positives.
 *
 * contains(item) will always return true if item was one of the last N to 1.5*N
 * insert()'ed ... but may also return true for items that were not inserted.
 *
 * It needs around 1.8 bytes per element per factor 0.1 of false positive rate.
 * (More accurately: 3/(log(256)*log(2)) * log(1/fpRate) * nElements bytes)
 */
class CRollingBloomFilter
{
public:
    CRollingBloomFilter(const unsigned int nElements, const double nFPRate);

    void insert(const std::vector<unsigned char>& vKey);
    void insert(const uint256& hash);
    bool contains(const std::vector<unsigned char>& vKey) const;
    bool contains(const uint256& hash) const;

    void reset();

private:
    int nEntriesPerGeneration;
    int nEntriesThisGeneration;
    int nGeneration;
    std::vector<uint64_t> data;
    unsigned int nTweak;
    int nHashFuncs;
};

#endif // BITCOIN_BLOOM_H