mirror of
https://github.com/cwinfo/matterbridge.git
synced 2025-07-10 17:16:31 +00:00
Wim
414
vendor/github.com/klauspost/compress/s2/encode_go.go
generated
vendored
414
vendor/github.com/klauspost/compress/s2/encode_go.go
generated
vendored
@ -4,14 +4,18 @@
|
||||
package s2
|
||||
|
||||
import (
|
||||
"bytes"
|
||||
"math/bits"
|
||||
)
|
||||
|
||||
const hasAmd64Asm = false
|
||||
|
||||
// encodeBlock encodes a non-empty src to a guaranteed-large-enough dst. It
|
||||
// assumes that the varint-encoded length of the decompressed bytes has already
|
||||
// been written.
|
||||
//
|
||||
// It also assumes that:
|
||||
//
|
||||
// len(dst) >= MaxEncodedLen(len(src))
|
||||
func encodeBlock(dst, src []byte) (d int) {
|
||||
if len(src) < minNonLiteralBlockSize {
|
||||
@ -25,6 +29,7 @@ func encodeBlock(dst, src []byte) (d int) {
|
||||
// been written.
|
||||
//
|
||||
// It also assumes that:
|
||||
//
|
||||
// len(dst) >= MaxEncodedLen(len(src))
|
||||
func encodeBlockBetter(dst, src []byte) (d int) {
|
||||
return encodeBlockBetterGo(dst, src)
|
||||
@ -35,6 +40,7 @@ func encodeBlockBetter(dst, src []byte) (d int) {
|
||||
// been written.
|
||||
//
|
||||
// It also assumes that:
|
||||
//
|
||||
// len(dst) >= MaxEncodedLen(len(src))
|
||||
func encodeBlockBetterSnappy(dst, src []byte) (d int) {
|
||||
return encodeBlockBetterSnappyGo(dst, src)
|
||||
@ -45,6 +51,7 @@ func encodeBlockBetterSnappy(dst, src []byte) (d int) {
|
||||
// been written.
|
||||
//
|
||||
// It also assumes that:
|
||||
//
|
||||
// len(dst) >= MaxEncodedLen(len(src))
|
||||
func encodeBlockSnappy(dst, src []byte) (d int) {
|
||||
if len(src) < minNonLiteralBlockSize {
|
||||
@ -56,6 +63,7 @@ func encodeBlockSnappy(dst, src []byte) (d int) {
|
||||
// emitLiteral writes a literal chunk and returns the number of bytes written.
|
||||
//
|
||||
// It assumes that:
|
||||
//
|
||||
// dst is long enough to hold the encoded bytes
|
||||
// 0 <= len(lit) && len(lit) <= math.MaxUint32
|
||||
func emitLiteral(dst, lit []byte) int {
|
||||
@ -146,6 +154,7 @@ func emitRepeat(dst []byte, offset, length int) int {
|
||||
// emitCopy writes a copy chunk and returns the number of bytes written.
|
||||
//
|
||||
// It assumes that:
|
||||
//
|
||||
// dst is long enough to hold the encoded bytes
|
||||
// 1 <= offset && offset <= math.MaxUint32
|
||||
// 4 <= length && length <= 1 << 24
|
||||
@ -214,6 +223,7 @@ func emitCopy(dst []byte, offset, length int) int {
|
||||
// emitCopyNoRepeat writes a copy chunk and returns the number of bytes written.
|
||||
//
|
||||
// It assumes that:
|
||||
//
|
||||
// dst is long enough to hold the encoded bytes
|
||||
// 1 <= offset && offset <= math.MaxUint32
|
||||
// 4 <= length && length <= 1 << 24
|
||||
@ -273,8 +283,8 @@ func emitCopyNoRepeat(dst []byte, offset, length int) int {
|
||||
// matchLen returns how many bytes match in a and b
|
||||
//
|
||||
// It assumes that:
|
||||
// len(a) <= len(b)
|
||||
//
|
||||
// len(a) <= len(b)
|
||||
func matchLen(a []byte, b []byte) int {
|
||||
b = b[:len(a)]
|
||||
var checked int
|
||||
@ -305,3 +315,405 @@ func matchLen(a []byte, b []byte) int {
|
||||
}
|
||||
return len(a) + checked
|
||||
}
|
||||
|
||||
func calcBlockSize(src []byte) (d int) {
|
||||
// Initialize the hash table.
|
||||
const (
|
||||
tableBits = 13
|
||||
maxTableSize = 1 << tableBits
|
||||
)
|
||||
|
||||
var table [maxTableSize]uint32
|
||||
|
||||
// sLimit is when to stop looking for offset/length copies. The inputMargin
|
||||
// lets us use a fast path for emitLiteral in the main loop, while we are
|
||||
// looking for copies.
|
||||
sLimit := len(src) - inputMargin
|
||||
|
||||
// Bail if we can't compress to at least this.
|
||||
dstLimit := len(src) - len(src)>>5 - 5
|
||||
|
||||
// nextEmit is where in src the next emitLiteral should start from.
|
||||
nextEmit := 0
|
||||
|
||||
// The encoded form must start with a literal, as there are no previous
|
||||
// bytes to copy, so we start looking for hash matches at s == 1.
|
||||
s := 1
|
||||
cv := load64(src, s)
|
||||
|
||||
// We search for a repeat at -1, but don't output repeats when nextEmit == 0
|
||||
repeat := 1
|
||||
|
||||
for {
|
||||
candidate := 0
|
||||
for {
|
||||
// Next src position to check
|
||||
nextS := s + (s-nextEmit)>>6 + 4
|
||||
if nextS > sLimit {
|
||||
goto emitRemainder
|
||||
}
|
||||
hash0 := hash6(cv, tableBits)
|
||||
hash1 := hash6(cv>>8, tableBits)
|
||||
candidate = int(table[hash0])
|
||||
candidate2 := int(table[hash1])
|
||||
table[hash0] = uint32(s)
|
||||
table[hash1] = uint32(s + 1)
|
||||
hash2 := hash6(cv>>16, tableBits)
|
||||
|
||||
// Check repeat at offset checkRep.
|
||||
const checkRep = 1
|
||||
if uint32(cv>>(checkRep*8)) == load32(src, s-repeat+checkRep) {
|
||||
base := s + checkRep
|
||||
// Extend back
|
||||
for i := base - repeat; base > nextEmit && i > 0 && src[i-1] == src[base-1]; {
|
||||
i--
|
||||
base--
|
||||
}
|
||||
d += emitLiteralSize(src[nextEmit:base])
|
||||
|
||||
// Extend forward
|
||||
candidate := s - repeat + 4 + checkRep
|
||||
s += 4 + checkRep
|
||||
for s <= sLimit {
|
||||
if diff := load64(src, s) ^ load64(src, candidate); diff != 0 {
|
||||
s += bits.TrailingZeros64(diff) >> 3
|
||||
break
|
||||
}
|
||||
s += 8
|
||||
candidate += 8
|
||||
}
|
||||
|
||||
d += emitCopyNoRepeatSize(repeat, s-base)
|
||||
nextEmit = s
|
||||
if s >= sLimit {
|
||||
goto emitRemainder
|
||||
}
|
||||
|
||||
cv = load64(src, s)
|
||||
continue
|
||||
}
|
||||
|
||||
if uint32(cv) == load32(src, candidate) {
|
||||
break
|
||||
}
|
||||
candidate = int(table[hash2])
|
||||
if uint32(cv>>8) == load32(src, candidate2) {
|
||||
table[hash2] = uint32(s + 2)
|
||||
candidate = candidate2
|
||||
s++
|
||||
break
|
||||
}
|
||||
table[hash2] = uint32(s + 2)
|
||||
if uint32(cv>>16) == load32(src, candidate) {
|
||||
s += 2
|
||||
break
|
||||
}
|
||||
|
||||
cv = load64(src, nextS)
|
||||
s = nextS
|
||||
}
|
||||
|
||||
// Extend backwards
|
||||
for candidate > 0 && s > nextEmit && src[candidate-1] == src[s-1] {
|
||||
candidate--
|
||||
s--
|
||||
}
|
||||
|
||||
// Bail if we exceed the maximum size.
|
||||
if d+(s-nextEmit) > dstLimit {
|
||||
return 0
|
||||
}
|
||||
|
||||
// A 4-byte match has been found. We'll later see if more than 4 bytes
|
||||
// match. But, prior to the match, src[nextEmit:s] are unmatched. Emit
|
||||
// them as literal bytes.
|
||||
|
||||
d += emitLiteralSize(src[nextEmit:s])
|
||||
|
||||
// Call emitCopy, and then see if another emitCopy could be our next
|
||||
// move. Repeat until we find no match for the input immediately after
|
||||
// what was consumed by the last emitCopy call.
|
||||
//
|
||||
// If we exit this loop normally then we need to call emitLiteral next,
|
||||
// though we don't yet know how big the literal will be. We handle that
|
||||
// by proceeding to the next iteration of the main loop. We also can
|
||||
// exit this loop via goto if we get close to exhausting the input.
|
||||
for {
|
||||
// Invariant: we have a 4-byte match at s, and no need to emit any
|
||||
// literal bytes prior to s.
|
||||
base := s
|
||||
repeat = base - candidate
|
||||
|
||||
// Extend the 4-byte match as long as possible.
|
||||
s += 4
|
||||
candidate += 4
|
||||
for s <= len(src)-8 {
|
||||
if diff := load64(src, s) ^ load64(src, candidate); diff != 0 {
|
||||
s += bits.TrailingZeros64(diff) >> 3
|
||||
break
|
||||
}
|
||||
s += 8
|
||||
candidate += 8
|
||||
}
|
||||
|
||||
d += emitCopyNoRepeatSize(repeat, s-base)
|
||||
if false {
|
||||
// Validate match.
|
||||
a := src[base:s]
|
||||
b := src[base-repeat : base-repeat+(s-base)]
|
||||
if !bytes.Equal(a, b) {
|
||||
panic("mismatch")
|
||||
}
|
||||
}
|
||||
|
||||
nextEmit = s
|
||||
if s >= sLimit {
|
||||
goto emitRemainder
|
||||
}
|
||||
|
||||
if d > dstLimit {
|
||||
// Do we have space for more, if not bail.
|
||||
return 0
|
||||
}
|
||||
// Check for an immediate match, otherwise start search at s+1
|
||||
x := load64(src, s-2)
|
||||
m2Hash := hash6(x, tableBits)
|
||||
currHash := hash6(x>>16, tableBits)
|
||||
candidate = int(table[currHash])
|
||||
table[m2Hash] = uint32(s - 2)
|
||||
table[currHash] = uint32(s)
|
||||
if uint32(x>>16) != load32(src, candidate) {
|
||||
cv = load64(src, s+1)
|
||||
s++
|
||||
break
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
emitRemainder:
|
||||
if nextEmit < len(src) {
|
||||
// Bail if we exceed the maximum size.
|
||||
if d+len(src)-nextEmit > dstLimit {
|
||||
return 0
|
||||
}
|
||||
d += emitLiteralSize(src[nextEmit:])
|
||||
}
|
||||
return d
|
||||
}
|
||||
|
||||
func calcBlockSizeSmall(src []byte) (d int) {
|
||||
// Initialize the hash table.
|
||||
const (
|
||||
tableBits = 9
|
||||
maxTableSize = 1 << tableBits
|
||||
)
|
||||
|
||||
var table [maxTableSize]uint32
|
||||
|
||||
// sLimit is when to stop looking for offset/length copies. The inputMargin
|
||||
// lets us use a fast path for emitLiteral in the main loop, while we are
|
||||
// looking for copies.
|
||||
sLimit := len(src) - inputMargin
|
||||
|
||||
// Bail if we can't compress to at least this.
|
||||
dstLimit := len(src) - len(src)>>5 - 5
|
||||
|
||||
// nextEmit is where in src the next emitLiteral should start from.
|
||||
nextEmit := 0
|
||||
|
||||
// The encoded form must start with a literal, as there are no previous
|
||||
// bytes to copy, so we start looking for hash matches at s == 1.
|
||||
s := 1
|
||||
cv := load64(src, s)
|
||||
|
||||
// We search for a repeat at -1, but don't output repeats when nextEmit == 0
|
||||
repeat := 1
|
||||
|
||||
for {
|
||||
candidate := 0
|
||||
for {
|
||||
// Next src position to check
|
||||
nextS := s + (s-nextEmit)>>6 + 4
|
||||
if nextS > sLimit {
|
||||
goto emitRemainder
|
||||
}
|
||||
hash0 := hash6(cv, tableBits)
|
||||
hash1 := hash6(cv>>8, tableBits)
|
||||
candidate = int(table[hash0])
|
||||
candidate2 := int(table[hash1])
|
||||
table[hash0] = uint32(s)
|
||||
table[hash1] = uint32(s + 1)
|
||||
hash2 := hash6(cv>>16, tableBits)
|
||||
|
||||
// Check repeat at offset checkRep.
|
||||
const checkRep = 1
|
||||
if uint32(cv>>(checkRep*8)) == load32(src, s-repeat+checkRep) {
|
||||
base := s + checkRep
|
||||
// Extend back
|
||||
for i := base - repeat; base > nextEmit && i > 0 && src[i-1] == src[base-1]; {
|
||||
i--
|
||||
base--
|
||||
}
|
||||
d += emitLiteralSize(src[nextEmit:base])
|
||||
|
||||
// Extend forward
|
||||
candidate := s - repeat + 4 + checkRep
|
||||
s += 4 + checkRep
|
||||
for s <= sLimit {
|
||||
if diff := load64(src, s) ^ load64(src, candidate); diff != 0 {
|
||||
s += bits.TrailingZeros64(diff) >> 3
|
||||
break
|
||||
}
|
||||
s += 8
|
||||
candidate += 8
|
||||
}
|
||||
|
||||
d += emitCopyNoRepeatSize(repeat, s-base)
|
||||
nextEmit = s
|
||||
if s >= sLimit {
|
||||
goto emitRemainder
|
||||
}
|
||||
|
||||
cv = load64(src, s)
|
||||
continue
|
||||
}
|
||||
|
||||
if uint32(cv) == load32(src, candidate) {
|
||||
break
|
||||
}
|
||||
candidate = int(table[hash2])
|
||||
if uint32(cv>>8) == load32(src, candidate2) {
|
||||
table[hash2] = uint32(s + 2)
|
||||
candidate = candidate2
|
||||
s++
|
||||
break
|
||||
}
|
||||
table[hash2] = uint32(s + 2)
|
||||
if uint32(cv>>16) == load32(src, candidate) {
|
||||
s += 2
|
||||
break
|
||||
}
|
||||
|
||||
cv = load64(src, nextS)
|
||||
s = nextS
|
||||
}
|
||||
|
||||
// Extend backwards
|
||||
for candidate > 0 && s > nextEmit && src[candidate-1] == src[s-1] {
|
||||
candidate--
|
||||
s--
|
||||
}
|
||||
|
||||
// Bail if we exceed the maximum size.
|
||||
if d+(s-nextEmit) > dstLimit {
|
||||
return 0
|
||||
}
|
||||
|
||||
// A 4-byte match has been found. We'll later see if more than 4 bytes
|
||||
// match. But, prior to the match, src[nextEmit:s] are unmatched. Emit
|
||||
// them as literal bytes.
|
||||
|
||||
d += emitLiteralSize(src[nextEmit:s])
|
||||
|
||||
// Call emitCopy, and then see if another emitCopy could be our next
|
||||
// move. Repeat until we find no match for the input immediately after
|
||||
// what was consumed by the last emitCopy call.
|
||||
//
|
||||
// If we exit this loop normally then we need to call emitLiteral next,
|
||||
// though we don't yet know how big the literal will be. We handle that
|
||||
// by proceeding to the next iteration of the main loop. We also can
|
||||
// exit this loop via goto if we get close to exhausting the input.
|
||||
for {
|
||||
// Invariant: we have a 4-byte match at s, and no need to emit any
|
||||
// literal bytes prior to s.
|
||||
base := s
|
||||
repeat = base - candidate
|
||||
|
||||
// Extend the 4-byte match as long as possible.
|
||||
s += 4
|
||||
candidate += 4
|
||||
for s <= len(src)-8 {
|
||||
if diff := load64(src, s) ^ load64(src, candidate); diff != 0 {
|
||||
s += bits.TrailingZeros64(diff) >> 3
|
||||
break
|
||||
}
|
||||
s += 8
|
||||
candidate += 8
|
||||
}
|
||||
|
||||
d += emitCopyNoRepeatSize(repeat, s-base)
|
||||
if false {
|
||||
// Validate match.
|
||||
a := src[base:s]
|
||||
b := src[base-repeat : base-repeat+(s-base)]
|
||||
if !bytes.Equal(a, b) {
|
||||
panic("mismatch")
|
||||
}
|
||||
}
|
||||
|
||||
nextEmit = s
|
||||
if s >= sLimit {
|
||||
goto emitRemainder
|
||||
}
|
||||
|
||||
if d > dstLimit {
|
||||
// Do we have space for more, if not bail.
|
||||
return 0
|
||||
}
|
||||
// Check for an immediate match, otherwise start search at s+1
|
||||
x := load64(src, s-2)
|
||||
m2Hash := hash6(x, tableBits)
|
||||
currHash := hash6(x>>16, tableBits)
|
||||
candidate = int(table[currHash])
|
||||
table[m2Hash] = uint32(s - 2)
|
||||
table[currHash] = uint32(s)
|
||||
if uint32(x>>16) != load32(src, candidate) {
|
||||
cv = load64(src, s+1)
|
||||
s++
|
||||
break
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
emitRemainder:
|
||||
if nextEmit < len(src) {
|
||||
// Bail if we exceed the maximum size.
|
||||
if d+len(src)-nextEmit > dstLimit {
|
||||
return 0
|
||||
}
|
||||
d += emitLiteralSize(src[nextEmit:])
|
||||
}
|
||||
return d
|
||||
}
|
||||
|
||||
// emitLiteral writes a literal chunk and returns the number of bytes written.
|
||||
//
|
||||
// It assumes that:
|
||||
//
|
||||
// dst is long enough to hold the encoded bytes
|
||||
// 0 <= len(lit) && len(lit) <= math.MaxUint32
|
||||
func emitLiteralSize(lit []byte) int {
|
||||
if len(lit) == 0 {
|
||||
return 0
|
||||
}
|
||||
switch {
|
||||
case len(lit) <= 60:
|
||||
return len(lit) + 1
|
||||
case len(lit) <= 1<<8:
|
||||
return len(lit) + 2
|
||||
case len(lit) <= 1<<16:
|
||||
return len(lit) + 3
|
||||
case len(lit) <= 1<<24:
|
||||
return len(lit) + 4
|
||||
default:
|
||||
return len(lit) + 5
|
||||
}
|
||||
}
|
||||
|
||||
func cvtLZ4BlockAsm(dst []byte, src []byte) (uncompressed int, dstUsed int) {
|
||||
panic("cvtLZ4BlockAsm should be unreachable")
|
||||
}
|
||||
|
||||
func cvtLZ4BlockSnappyAsm(dst []byte, src []byte) (uncompressed int, dstUsed int) {
|
||||
panic("cvtLZ4BlockSnappyAsm should be unreachable")
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user