111 lines
3.1 KiB
Go
111 lines
3.1 KiB
Go
// Copyright 2015, Joe Tsai. All rights reserved.
|
|
// Use of this source code is governed by a BSD-style
|
|
// license that can be found in the LICENSE.md file.
|
|
|
|
package bzip2
|
|
|
|
import "github.com/dsnet/compress/bzip2/internal/sais"
|
|
|
|
// The Burrows-Wheeler Transform implementation used here is based on the
|
|
// Suffix Array by Induced Sorting (SA-IS) methodology by Nong, Zhang, and Chan.
|
|
// This implementation uses the sais algorithm originally written by Yuta Mori.
|
|
//
|
|
// The SA-IS algorithm runs in O(n) and outputs a Suffix Array. There is a
|
|
// mathematical relationship between Suffix Arrays and the Burrows-Wheeler
|
|
// Transform, such that a SA can be converted to a BWT in O(n) time.
|
|
//
|
|
// References:
|
|
// http://www.hpl.hp.com/techreports/Compaq-DEC/SRC-RR-124.pdf
|
|
// https://github.com/cscott/compressjs/blob/master/lib/BWT.js
|
|
// https://www.quora.com/How-can-I-optimize-burrows-wheeler-transform-and-inverse-transform-to-work-in-O-n-time-O-n-space
|
|
type burrowsWheelerTransform struct {
|
|
buf []byte
|
|
sa []int
|
|
perm []uint32
|
|
}
|
|
|
|
func (bwt *burrowsWheelerTransform) Encode(buf []byte) (ptr int) {
|
|
if len(buf) == 0 {
|
|
return -1
|
|
}
|
|
|
|
// TODO(dsnet): Find a way to avoid the duplicate input string method.
|
|
// We only need to do this because suffix arrays (by definition) only
|
|
// operate non-wrapped suffixes of a string. On the other hand,
|
|
// the BWT specifically used in bzip2 operate on a strings that wrap-around
|
|
// when being sorted.
|
|
|
|
// Step 1: Concatenate the input string to itself so that we can use the
|
|
// suffix array algorithm for bzip2's variant of BWT.
|
|
n := len(buf)
|
|
bwt.buf = append(append(bwt.buf[:0], buf...), buf...)
|
|
if cap(bwt.sa) < 2*n {
|
|
bwt.sa = make([]int, 2*n)
|
|
}
|
|
t := bwt.buf[:2*n]
|
|
sa := bwt.sa[:2*n]
|
|
|
|
// Step 2: Compute the suffix array (SA). The input string, t, will not be
|
|
// modified, while the results will be written to the output, sa.
|
|
sais.ComputeSA(t, sa)
|
|
|
|
// Step 3: Convert the SA to a BWT. Since ComputeSA does not mutate the
|
|
// input, we have two copies of the input; in buf and buf2. Thus, we write
|
|
// the transformation to buf, while using buf2.
|
|
var j int
|
|
buf2 := t[n:]
|
|
for _, i := range sa {
|
|
if i < n {
|
|
if i == 0 {
|
|
ptr = j
|
|
i = n
|
|
}
|
|
buf[j] = buf2[i-1]
|
|
j++
|
|
}
|
|
}
|
|
return ptr
|
|
}
|
|
|
|
func (bwt *burrowsWheelerTransform) Decode(buf []byte, ptr int) {
|
|
if len(buf) == 0 {
|
|
return
|
|
}
|
|
|
|
// Step 1: Compute cumm, where cumm[ch] reports the total number of
|
|
// characters that precede the character ch in the alphabet.
|
|
var cumm [256]int
|
|
for _, v := range buf {
|
|
cumm[v]++
|
|
}
|
|
var sum int
|
|
for i, v := range cumm {
|
|
cumm[i] = sum
|
|
sum += v
|
|
}
|
|
|
|
// Step 2: Compute perm, where perm[ptr] contains a pointer to the next
|
|
// byte in buf and the next pointer in perm itself.
|
|
if cap(bwt.perm) < len(buf) {
|
|
bwt.perm = make([]uint32, len(buf))
|
|
}
|
|
perm := bwt.perm[:len(buf)]
|
|
for i, b := range buf {
|
|
perm[cumm[b]] = uint32(i)
|
|
cumm[b]++
|
|
}
|
|
|
|
// Step 3: Follow each pointer in perm to the next byte, starting with the
|
|
// origin pointer.
|
|
if cap(bwt.buf) < len(buf) {
|
|
bwt.buf = make([]byte, len(buf))
|
|
}
|
|
buf2 := bwt.buf[:len(buf)]
|
|
i := perm[ptr]
|
|
for j := range buf2 {
|
|
buf2[j] = buf[i]
|
|
i = perm[i]
|
|
}
|
|
copy(buf, buf2)
|
|
}
|