mirror/erigon-pulse
1
0
Fork 0
mirror of https://gitlab.com/pulsechaincom/erigon-pulse.git synced 2024-12-28 14:47:16 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

feat: implement eliasfano32.Min() (#682)

Browse Source 
feat: implement eliasfano32.Min()
feature:
1. implement eliasfano32.Min() by calculating lower/upper bits.

refactor:
1. fix miscalculate of `superQSize` Annotation

relevant Issue in erigon:
https://github.com/ledgerwatch/erigon/issues/5623
This commit is contained in:
J1ang 2022-10-14 14:46:24 +08:00 committed by GitHub
parent 980cf42504
commit 7e68f20838
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 42 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
recsplit/eliasfano32/elias_fano.go
View File

@ -39,7 +39,7 @@ const (
superQ uint64 = 1 << 14
superQMask uint64 = superQ - 1
qPerSuperQ uint64 = superQ / q // 64
superQSize uint64 = 1 + qPerSuperQ/2 // 1 + 64/4 = 17
superQSize uint64 = 1 + qPerSuperQ/2 // 1 + 64/2 = 33
)
// EliasFano can be used to encode one monotone sequence
@ -217,6 +217,10 @@ func (ef EliasFano) Max() uint64 {
return ef.maxOffset
}
func (ef EliasFano) Min() uint64 {
return ef.Get(0)
}
func (ef EliasFano) Count() uint64 {
return ef.count + 1
}
@ -297,7 +301,7 @@ func (ef *EliasFano) AppendBytes(buf []byte) []byte {
return buf
}
const maxDataSize = 0xFFFFFFFFFFFF
const maxDataSize = 0xFFFFFFFFFFFF //2^48
// Read inputs the state of golomb rice encoding from a reader s
func ReadEliasFano(r []byte) (*EliasFano, int) {
@ -314,6 +318,38 @@ func ReadEliasFano(r []byte) (*EliasFano, int) {
func Max(r []byte) uint64 { return binary.BigEndian.Uint64(r[8:16]) - 1 }
func Min(r []byte) uint64 {
count := binary.BigEndian.Uint64(r[:8])
u := binary.BigEndian.Uint64(r[8:16])
p := (*[maxDataSize / 8]uint64)(unsafe.Pointer(&r[16]))
var l uint64
if u/(count+1) == 0 {
l = 0
} else {
l = 63 ^ uint64(bits.LeadingZeros64(u/(count+1))) // pos of first non-zero bit
}
wordsLowerBits := int(((count+1)*l+63)/64 + 1)
wordsUpperBits := int((count + 1 + (u >> l) + 63) / 64)
lowerBits := p[:wordsLowerBits]
upperBits := p[wordsLowerBits : wordsLowerBits+wordsUpperBits]
jump := p[wordsLowerBits+wordsUpperBits:]
lower := lowerBits[0]
mask := uint64(0xffffffff)
j := jump[0] + jump[1]&mask
currWord := j / 64
window := upperBits[currWord] & (uint64(0xffffffffffffffff) << (j % 64))
if bitCount := bits.OnesCount64(window); bitCount <= 0 {
currWord++
window = upperBits[currWord]
}
sel := bitutil.Select64(window, 0)
lowerBitsMask := (uint64(1) << l) - 1
val := ((currWord*64+uint64(sel))<<l | (lower & lowerBitsMask))
return val
}
// DoubleEliasFano can be used to encode two monotone sequences
// it is called "double" because the lower bits array contains two sequences interleaved
type DoubleEliasFano struct {

3
recsplit/eliasfano32/elias_fano_fuzz_test.go
View File

@ -65,6 +65,9 @@ func FuzzSingleEliasFano(f *testing.F) {
if ef.Max() != Max(buf.Bytes()) {
t.Fatalf("max: got %d, expected %d", ef.Max(), Max(buf.Bytes()))
}
if ef.Min() != Min(buf.Bytes()) {
t.Fatalf("min: got %d, expected %d", ef.Min(), Min(buf.Bytes()))
}
})
}

1
recsplit/eliasfano32/elias_fano_test.go
View File

@ -55,6 +55,7 @@ func TestEliasFano(t *testing.T) {
buf := bytes.NewBuffer(nil)
ef.Write(buf)
assert.Equal(t, ef.Max(), Max(buf.Bytes()))
assert.Equal(t, ef.Min(), Min(buf.Bytes()))
}
func TestIterator(t *testing.T) {