Module: Omnizip::Algorithms::Zstandard::XXHash64

Defined in:
lib/omnizip/algorithms/zstandard/xxhash.rb

Overview

XXHash64 (https://github.com/Cyan4973/xxHash), single-shot.

Zstandard frame checksums are the low 32 bits of XXHash64 of the decoded content with seed 0 (the zstd C reference truncates XXH64_digest to a U32; RFC 8878's "xxh64" wording agrees).

Constant Summary collapse

MASK64 = 
0xFFFFFFFFFFFFFFFF
MASK32 = 
0xFFFFFFFF
PRIME64_1 = 
0x9E3779B185EBCA87
PRIME64_2 = 
0xC2B2AE3D27D4EB4F
PRIME64_3 = 
0x165667B19E3779F9
PRIME64_4 = 
0x85EBCA77C2B2AE63
PRIME64_5 = 
0x27D4EB2F165667C5

Class Method Summary collapse

Class Method Details

.digest(data) ⇒ Integer

Digest data with seed 0.

Parameters:

  • data (String) —

    binary string

Returns:

  • (Integer) —

    64-bit digest



47
48
49
 
# File 'lib/omnizip/algorithms/zstandard/xxhash.rb', line 47

def digest(data)
  digest_seeded(data, 0)
end

.digest_seeded(data, seed) ⇒ Object 



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
 
# File 'lib/omnizip/algorithms/zstandard/xxhash.rb', line 51

def digest_seeded(data, seed)
  len = data.bytesize
  hash = 0
  tail_start = 0

  if len >= 32
    v1 = (seed + PRIME64_1 + PRIME64_2) & MASK64
    v2 = (seed + PRIME64_2) & MASK64
    v3 = seed & MASK64
    v4 = (seed - PRIME64_1) & MASK64

    i = 0
    while i + 32 <= len
      v1 = round64(v1, read_u64(data, i))
      v2 = round64(v2, read_u64(data, i + 8))
      v3 = round64(v3, read_u64(data, i + 16))
      v4 = round64(v4, read_u64(data, i + 24))
      i += 32
    end

    hash = rotl64(v1, 1)
    hash = (hash + rotl64(v2, 7)) & MASK64
    hash = (hash + rotl64(v3, 12)) & MASK64
    hash = (hash + rotl64(v4, 18)) & MASK64

    hash = merge_round64(hash, v1)
    hash = merge_round64(hash, v2)
    hash = merge_round64(hash, v3)
    hash = merge_round64(hash, v4)

    tail_start = len - (len % 32)
  else
    hash = (seed + PRIME64_5) & MASK64
  end

  hash = (hash + len) & MASK64

  i = tail_start
  while i + 8 <= len
    k1 = round64(0, read_u64(data, i))
    hash ^= k1
    hash = ((rotl64(hash, 27) * PRIME64_1) & MASK64) + PRIME64_4
    hash &= MASK64
    i += 8
  end
  if i + 4 <= len
    hash ^= (read_u32(data, i) * PRIME64_1) & MASK64
    hash = ((rotl64(hash, 23) * PRIME64_2) & MASK64) + PRIME64_3
    hash &= MASK64
    i += 4
  end
  while i < len
    hash ^= (data.getbyte(i) * PRIME64_5) & MASK64
    hash = (rotl64(hash, 11) * PRIME64_1) & MASK64
    i += 1
  end

  hash ^= hash >> 33
  hash = (hash * PRIME64_2) & MASK64
  hash ^= hash >> 29
  hash = (hash * PRIME64_3) & MASK64
  hash ^= hash >> 32
  hash
end

.frame_checksum(data) ⇒ Integer

Zstandard frame checksum: low 32 bits of XXHash64.

Returns:

  • (Integer) —

    32-bit checksum



119
120
121
 
# File 'lib/omnizip/algorithms/zstandard/xxhash.rb', line 119

def frame_checksum(data)
  digest(data) & MASK32
end

.merge_round64(acc, value) ⇒ Object 



151
152
153
154
155
 
# File 'lib/omnizip/algorithms/zstandard/xxhash.rb', line 151

def merge_round64(acc, value)
  value = round64(0, value)
  acc ^= value
  (((acc * PRIME64_1) & MASK64) + PRIME64_4) & MASK64
end

.read_u32(data, offset) ⇒ Object 



134
135
136
137
138
139
 
# File 'lib/omnizip/algorithms/zstandard/xxhash.rb', line 134

def read_u32(data, offset)
  data.getbyte(offset) |
    (data.getbyte(offset + 1) << 8) |
    (data.getbyte(offset + 2) << 16) |
    (data.getbyte(offset + 3) << 24)
end

.read_u64(data, offset) ⇒ Object 



123
124
125
126
127
128
129
130
131
132
 
# File 'lib/omnizip/algorithms/zstandard/xxhash.rb', line 123

def read_u64(data, offset)
  data.getbyte(offset) |
    (data.getbyte(offset + 1) << 8) |
    (data.getbyte(offset + 2) << 16) |
    (data.getbyte(offset + 3) << 24) |
    (data.getbyte(offset + 4) << 32) |
    (data.getbyte(offset + 5) << 40) |
    (data.getbyte(offset + 6) << 48) |
    (data.getbyte(offset + 7) << 56)
end

.rotl64(value, count) ⇒ Object 



141
142
143
 
# File 'lib/omnizip/algorithms/zstandard/xxhash.rb', line 141

def rotl64(value, count)
  ((value << count) | (value >> (64 - count))) & MASK64
end

.round64(acc, input) ⇒ Object 



145
146
147
148
149
 
# File 'lib/omnizip/algorithms/zstandard/xxhash.rb', line 145

def round64(acc, input)
  acc = ((acc + (input * PRIME64_2)) & MASK64) & MASK64
  acc = rotl64(acc, 31)
  (acc * PRIME64_1) & MASK64
end