/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.lucene.codecs.compressing;
import java.io.IOException;
import java.util.Arrays;
import org.apache.lucene.store.DataInput;
import org.apache.lucene.store.DataOutput;
import org.apache.lucene.util.packed.PackedInts;
/**
* LZ4 compression and decompression routines.
*
* http://code.google.com/p/lz4/
* http://fastcompression.blogspot.fr/p/lz4.html
*/
final class LZ4 {
private LZ4() {}
static final int MEMORY_USAGE = 14;
static final int MIN_MATCH = 4; // minimum length of a match
static final int MAX_DISTANCE = 1 << 16; // maximum distance of a reference
static final int LAST_LITERALS = 5; // the last 5 bytes must be encoded as literals
static final int HASH_LOG_HC = 15; // log size of the dictionary for compressHC
static final int HASH_TABLE_SIZE_HC = 1 << HASH_LOG_HC;
static final int OPTIMAL_ML = 0x0F + 4 - 1; // match length that doesn't require an additional byte
private static int hash(int i, int hashBits) {
return (i * -1640531535) >>> (32 - hashBits);
}
private static int hashHC(int i) {
return hash(i, HASH_LOG_HC);
}
private static int readInt(byte[] buf, int i) {
return ((buf[i] & 0xFF) << 24) | ((buf[i+1] & 0xFF) << 16) | ((buf[i+2] & 0xFF) << 8) | (buf[i+3] & 0xFF);
}
private static boolean readIntEquals(byte[] buf, int i, int j) {
return readInt(buf, i) == readInt(buf, j);
}
private static int commonBytes(byte[] b, int o1, int o2, int limit) {
assert o1 < o2;
int count = 0;
while (o2 < limit && b[o1++] == b[o2++]) {
++count;
}
return count;
}
private static int commonBytesBackward(byte[] b, int o1, int o2, int l1, int l2) {
int count = 0;
while (o1 > l1 && o2 > l2 && b[--o1] == b[--o2]) {
++count;
}
return count;
}
/**
* Decompress at least <code>decompressedLen</code> bytes into
* <code>dest[dOff:]</code>. Please note that <code>dest</code> must be large
* enough to be able to hold <b>all</b> decompressed data (meaning that you
* need to know the total decompressed length).
*/
public static int decompress(DataInput compressed, int decompressedLen, byte[] dest, int dOff) throws IOException {
final int destEnd = dest.length;
do {
// literals
final int token = compressed.readByte() & 0xFF;
int literalLen = token >>> 4;
if (literalLen != 0) {
if (literalLen == 0x0F) {
byte len;
while ((len = compressed.readByte()) == (byte) 0xFF) {
literalLen += 0xFF;
}
literalLen += len & 0xFF;
}
compressed.readBytes(dest, dOff, literalLen);
dOff += literalLen;
}
if (dOff >= decompressedLen) {
break;
}
// matchs
final int matchDec = (compressed.readByte() & 0xFF) | ((compressed.readByte() & 0xFF) << 8);
assert matchDec > 0;
int matchLen = token & 0x0F;
if (matchLen == 0x0F) {
int len;
while ((len = compressed.readByte()) == (byte) 0xFF) {
matchLen += 0xFF;
}
matchLen += len & 0xFF;
}
matchLen += MIN_MATCH;
// copying a multiple of 8 bytes can make decompression from 5% to 10% faster
final int fastLen = (matchLen + 7) & 0xFFFFFFF8;
if (matchDec < matchLen || dOff + fastLen > destEnd) {
// overlap -> naive incremental copy
for (int ref = dOff - matchDec, end = dOff + matchLen; dOff < end; ++ref, ++dOff) {
dest[dOff] = dest[ref];
}
} else {
// no overlap -> arraycopy
System.arraycopy(dest, dOff - matchDec, dest, dOff, fastLen);
dOff += matchLen;
}
} while (dOff < decompressedLen);
return dOff;
}
private static void encodeLen(int l, DataOutput out) throws IOException {
while (l >= 0xFF) {
out.writeByte((byte) 0xFF);
l -= 0xFF;
}
out.writeByte((byte) l);
}
private static void encodeLiterals(byte[] bytes, int token, int anchor, int literalLen, DataOutput out) throws IOException {
out.writeByte((byte) token);
// encode literal length
if (literalLen >= 0x0F) {
encodeLen(literalLen - 0x0F, out);
}
// encode literals
out.writeBytes(bytes, anchor, literalLen);
}
private static void encodeLastLiterals(byte[] bytes, int anchor, int literalLen, DataOutput out) throws IOException {
final int token = Math.min(literalLen, 0x0F) << 4;
encodeLiterals(bytes, token, anchor, literalLen, out);
}
private static void encodeSequence(byte[] bytes, int anchor, int matchRef, int matchOff, int matchLen, DataOutput out) throws IOException {
final int literalLen = matchOff - anchor;
assert matchLen >= 4;
// encode token
final int token = (Math.min(literalLen, 0x0F) << 4) | Math.min(matchLen - 4, 0x0F);
encodeLiterals(bytes, token, anchor, literalLen, out);
// encode match dec
final int matchDec = matchOff - matchRef;
assert matchDec > 0 && matchDec < 1 << 16;
out.writeByte((byte) matchDec);
out.writeByte((byte) (matchDec >>> 8));
// encode match len
if (matchLen >= MIN_MATCH + 0x0F) {
encodeLen(matchLen - 0x0F - MIN_MATCH, out);
}
}
static final class HashTable {
private int hashLog;
private PackedInts.Mutable hashTable;
void reset(int len) {
final int bitsPerOffset = PackedInts.bitsRequired(len - LAST_LITERALS);
final int bitsPerOffsetLog = 32 - Integer.numberOfLeadingZeros(bitsPerOffset - 1);
hashLog = MEMORY_USAGE + 3 - bitsPerOffsetLog;
if (hashTable == null || hashTable.size() < 1 << hashLog || hashTable.getBitsPerValue() < bitsPerOffset) {
hashTable = PackedInts.getMutable(1 << hashLog, bitsPerOffset, PackedInts.DEFAULT);
} else {
hashTable.clear();
}
}
}
/**
* Compress <code>bytes[off:off+len]</code> into <code>out</code> using
* at most 16KB of memory. <code>ht</code> shouldn't be shared across threads
* but can safely be reused.
*/
public static void compress(byte[] bytes, int off, int len, DataOutput out, HashTable ht) throws IOException {
final int base = off;
final int end = off + len;
int anchor = off++;
if (len > LAST_LITERALS + MIN_MATCH) {
final int limit = end - LAST_LITERALS;
final int matchLimit = limit - MIN_MATCH;
ht.reset(len);
final int hashLog = ht.hashLog;
final PackedInts.Mutable hashTable = ht.hashTable;
main:
while (off <= limit) {
// find a match
int ref;
while (true) {
if (off >= matchLimit) {
break main;
}
final int v = readInt(bytes, off);
final int h = hash(v, hashLog);
ref = base + (int) hashTable.get(h);
assert PackedInts.bitsRequired(off - base) <= hashTable.getBitsPerValue();
hashTable.set(h, off - base);
if (off - ref < MAX_DISTANCE && readInt(bytes, ref) == v) {
break;
}
++off;
}
// compute match length
final int matchLen = MIN_MATCH + commonBytes(bytes, ref + MIN_MATCH, off + MIN_MATCH, limit);
encodeSequence(bytes, anchor, ref, off, matchLen, out);
off += matchLen;
anchor = off;
}
}
// last literals
final int literalLen = end - anchor;
assert literalLen >= LAST_LITERALS || literalLen == len;
encodeLastLiterals(bytes, anchor, end - anchor, out);
}
private static class Match {
int start, ref, len;
void fix(int correction) {
start += correction;
ref += correction;
len -= correction;
}
int end() {
return start + len;
}
}
private static void copyTo(Match m1, Match m2) {
m2.len = m1.len;
m2.start = m1.start;
m2.ref = m1.ref;
}
static final class HCHashTable {
static final int MAX_ATTEMPTS = 256;
static final int MASK = MAX_DISTANCE - 1;
int nextToUpdate;
private int base;
private final int[] hashTable;
private final short[] chainTable;
HCHashTable() {
hashTable = new int[HASH_TABLE_SIZE_HC];
chainTable = new short[MAX_DISTANCE];
}
private void reset(int base) {
this.base = base;
nextToUpdate = base;
Arrays.fill(hashTable, -1);
Arrays.fill(chainTable, (short) 0);
}
private int hashPointer(byte[] bytes, int off) {
final int v = readInt(bytes, off);
final int h = hashHC(v);
return hashTable[h];
}
private int next(int off) {
return off - (chainTable[off & MASK] & 0xFFFF);
}
private void addHash(byte[] bytes, int off) {
final int v = readInt(bytes, off);
final int h = hashHC(v);
int delta = off - hashTable[h];
assert delta > 0 : delta;
if (delta >= MAX_DISTANCE) {
delta = MAX_DISTANCE - 1;
}
chainTable[off & MASK] = (short) delta;
hashTable[h] = off;
}
void insert(int off, byte[] bytes) {
for (; nextToUpdate < off; ++nextToUpdate) {
addHash(bytes, nextToUpdate);
}
}
boolean insertAndFindBestMatch(byte[] buf, int off, int matchLimit, Match match) {
match.start = off;
match.len = 0;
int delta = 0;
int repl = 0;
insert(off, buf);
int ref = hashPointer(buf, off);
if (ref >= off - 4 && ref <= off && ref >= base) { // potential repetition
if (readIntEquals(buf, ref, off)) { // confirmed
delta = off - ref;
repl = match.len = MIN_MATCH + commonBytes(buf, ref + MIN_MATCH, off + MIN_MATCH, matchLimit);
match.ref = ref;
}
ref = next(ref);
}
for (int i = 0; i < MAX_ATTEMPTS; ++i) {
if (ref < Math.max(base, off - MAX_DISTANCE + 1) || ref > off) {
break;
}
if (buf[ref + match.len] == buf[off + match.len] && readIntEquals(buf, ref, off)) {
final int matchLen = MIN_MATCH + commonBytes(buf, ref + MIN_MATCH, off + MIN_MATCH, matchLimit);
if (matchLen > match.len) {
match.ref = ref;
match.len = matchLen;
}
}
ref = next(ref);
}
if (repl != 0) {
int ptr = off;
final int end = off + repl - (MIN_MATCH - 1);
while (ptr < end - delta) {
chainTable[ptr & MASK] = (short) delta; // pre load
++ptr;
}
do {
chainTable[ptr & MASK] = (short) delta;
hashTable[hashHC(readInt(buf, ptr))] = ptr;
++ptr;
} while (ptr < end);
nextToUpdate = end;
}
return match.len != 0;
}
boolean insertAndFindWiderMatch(byte[] buf, int off, int startLimit, int matchLimit, int minLen, Match match) {
match.len = minLen;
insert(off, buf);
final int delta = off - startLimit;
int ref = hashPointer(buf, off);
for (int i = 0; i < MAX_ATTEMPTS; ++i) {
if (ref < Math.max(base, off - MAX_DISTANCE + 1) || ref > off) {
break;
}
if (buf[ref - delta + match.len] == buf[startLimit + match.len]
&& readIntEquals(buf, ref, off)) {
final int matchLenForward = MIN_MATCH + commonBytes(buf, ref + MIN_MATCH, off + MIN_MATCH, matchLimit);
final int matchLenBackward = commonBytesBackward(buf, ref, off, base, startLimit);
final int matchLen = matchLenBackward + matchLenForward;
if (matchLen > match.len) {
match.len = matchLen;
match.ref = ref - matchLenBackward;
match.start = off - matchLenBackward;
}
}
ref = next(ref);
}
return match.len > minLen;
}
}
/**
* Compress <code>bytes[off:off+len]</code> into <code>out</code>. Compared to
* {@link LZ4#compress(byte[], int, int, DataOutput, HashTable)}, this method
* is slower and uses more memory (~ 256KB per thread) but should provide
* better compression ratios (especially on large inputs) because it chooses
* the best match among up to 256 candidates and then performs trade-offs to
* fix overlapping matches. <code>ht</code> shouldn't be shared across threads
* but can safely be reused.
*/
public static void compressHC(byte[] src, int srcOff, int srcLen, DataOutput out, HCHashTable ht) throws IOException {
final int srcEnd = srcOff + srcLen;
final int matchLimit = srcEnd - LAST_LITERALS;
final int mfLimit = matchLimit - MIN_MATCH;
int sOff = srcOff;
int anchor = sOff++;
ht.reset(srcOff);
final Match match0 = new Match();
final Match match1 = new Match();
final Match match2 = new Match();
final Match match3 = new Match();
main:
while (sOff <= mfLimit) {
if (!ht.insertAndFindBestMatch(src, sOff, matchLimit, match1)) {
++sOff;
continue;
}
// saved, in case we would skip too much
copyTo(match1, match0);
search2:
while (true) {
assert match1.start >= anchor;
if (match1.end() >= mfLimit
|| !ht.insertAndFindWiderMatch(src, match1.end() - 2, match1.start + 1, matchLimit, match1.len, match2)) {
// no better match
encodeSequence(src, anchor, match1.ref, match1.start, match1.len, out);
anchor = sOff = match1.end();
continue main;
}
if (match0.start < match1.start) {
if (match2.start < match1.start + match0.len) { // empirical
copyTo(match0, match1);
}
}
assert match2.start > match1.start;
if (match2.start - match1.start < 3) { // First Match too small : removed
copyTo(match2, match1);
continue search2;
}
search3:
while (true) {
if (match2.start - match1.start < OPTIMAL_ML) {
int newMatchLen = match1.len;
if (newMatchLen > OPTIMAL_ML) {
newMatchLen = OPTIMAL_ML;
}
if (match1.start + newMatchLen > match2.end() - MIN_MATCH) {
newMatchLen = match2.start - match1.start + match2.len - MIN_MATCH;
}
final int correction = newMatchLen - (match2.start - match1.start);
if (correction > 0) {
match2.fix(correction);
}
}
if (match2.start + match2.len >= mfLimit
|| !ht.insertAndFindWiderMatch(src, match2.end() - 3, match2.start, matchLimit, match2.len, match3)) {
// no better match -> 2 sequences to encode
if (match2.start < match1.end()) {
match1.len = match2.start - match1.start;
}
// encode seq 1
encodeSequence(src, anchor, match1.ref, match1.start, match1.len, out);
anchor = sOff = match1.end();
// encode seq 2
encodeSequence(src, anchor, match2.ref, match2.start, match2.len, out);
anchor = sOff = match2.end();
continue main;
}
if (match3.start < match1.end() + 3) { // Not enough space for match 2 : remove it
if (match3.start >= match1.end()) { // // can write Seq1 immediately ==> Seq2 is removed, so Seq3 becomes Seq1
if (match2.start < match1.end()) {
final int correction = match1.end() - match2.start;
match2.fix(correction);
if (match2.len < MIN_MATCH) {
copyTo(match3, match2);
}
}
encodeSequence(src, anchor, match1.ref, match1.start, match1.len, out);
anchor = sOff = match1.end();
copyTo(match3, match1);
copyTo(match2, match0);
continue search2;
}
copyTo(match3, match2);
continue search3;
}
// OK, now we have 3 ascending matches; let's write at least the first one
if (match2.start < match1.end()) {
if (match2.start - match1.start < 0x0F) {
if (match1.len > OPTIMAL_ML) {
match1.len = OPTIMAL_ML;
}
if (match1.end() > match2.end() - MIN_MATCH) {
match1.len = match2.end() - match1.start - MIN_MATCH;
}
final int correction = match1.end() - match2.start;
match2.fix(correction);
} else {
match1.len = match2.start - match1.start;
}
}
encodeSequence(src, anchor, match1.ref, match1.start, match1.len, out);
anchor = sOff = match1.end();
copyTo(match2, match1);
copyTo(match3, match2);
continue search3;
}
}
}
encodeLastLiterals(src, anchor, srcEnd - anchor, out);
}
}