2017-07-05 20:57:07 +00:00
|
|
|
#include <stdlib.h>
|
|
|
|
#include <string.h>
|
|
|
|
#include <stdint.h>
|
2017-07-06 22:23:15 +00:00
|
|
|
#include <stdio.h>
|
2017-07-05 20:57:07 +00:00
|
|
|
|
|
|
|
#include "ldm.h"
|
2017-07-10 20:08:19 +00:00
|
|
|
#include "util.h"
|
2017-07-05 20:57:07 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
// Insert every (HASH_ONLY_EVERY + 1) into the hash table.
|
|
|
|
#define HASH_ONLY_EVERY 0
|
2017-07-08 00:09:28 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
#define LDM_MEMORY_USAGE 20
|
2017-07-06 22:23:15 +00:00
|
|
|
#define LDM_HASHLOG (LDM_MEMORY_USAGE-2)
|
|
|
|
#define LDM_HASHTABLESIZE (1 << (LDM_MEMORY_USAGE))
|
|
|
|
#define LDM_HASHTABLESIZE_U32 ((LDM_HASHTABLESIZE) >> 2)
|
|
|
|
|
2017-07-10 22:48:47 +00:00
|
|
|
#define LDM_OFFSET_SIZE 4
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
#define WINDOW_SIZE (1 << 20)
|
2017-07-12 01:13:26 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
//These should be multiples of four.
|
2017-07-12 23:31:31 +00:00
|
|
|
#define LDM_HASH_LENGTH 4
|
|
|
|
#define MINMATCH 4
|
2017-07-06 22:23:15 +00:00
|
|
|
|
|
|
|
#define ML_BITS 4
|
|
|
|
#define ML_MASK ((1U<<ML_BITS)-1)
|
|
|
|
#define RUN_BITS (8-ML_BITS)
|
|
|
|
#define RUN_MASK ((1U<<RUN_BITS)-1)
|
|
|
|
|
2017-07-12 01:13:26 +00:00
|
|
|
#define COMPUTE_STATS
|
2017-07-13 17:38:19 +00:00
|
|
|
#define CHECKSUM_CHAR_OFFSET 0
|
2017-07-12 01:13:26 +00:00
|
|
|
//#define RUN_CHECKS
|
2017-07-06 23:47:08 +00:00
|
|
|
//#define LDM_DEBUG
|
|
|
|
|
2017-07-05 20:57:07 +00:00
|
|
|
typedef uint8_t BYTE;
|
|
|
|
typedef uint16_t U16;
|
|
|
|
typedef uint32_t U32;
|
|
|
|
typedef int32_t S32;
|
|
|
|
typedef uint64_t U64;
|
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
typedef uint32_t offset_t;
|
|
|
|
typedef uint32_t hash_t;
|
2017-07-11 05:27:43 +00:00
|
|
|
typedef signed char schar;
|
2017-07-10 19:38:27 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
typedef struct hashEntry {
|
2017-07-10 19:38:27 +00:00
|
|
|
offset_t offset;
|
2017-07-12 22:11:06 +00:00
|
|
|
} hashEntry;
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
typedef struct LDM_compressStats {
|
2017-07-10 22:48:47 +00:00
|
|
|
U32 numMatches;
|
|
|
|
U32 totalMatchLength;
|
|
|
|
U32 totalLiteralLength;
|
|
|
|
U64 totalOffset;
|
2017-07-12 01:13:26 +00:00
|
|
|
|
|
|
|
U32 numCollisions;
|
|
|
|
U32 numHashInserts;
|
2017-07-10 19:38:27 +00:00
|
|
|
} LDM_compressStats;
|
2017-07-08 00:09:28 +00:00
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
typedef struct LDM_CCtx {
|
|
|
|
size_t isize; /* Input size */
|
|
|
|
size_t maxOSize; /* Maximum output size */
|
|
|
|
|
|
|
|
const BYTE *ibase; /* Base of input */
|
|
|
|
const BYTE *ip; /* Current input position */
|
|
|
|
const BYTE *iend; /* End of input */
|
|
|
|
|
|
|
|
// Maximum input position such that hashing at the position does not exceed
|
|
|
|
// end of input.
|
|
|
|
const BYTE *ihashLimit;
|
|
|
|
|
|
|
|
// Maximum input position such that finding a match of at least the minimum
|
|
|
|
// match length does not exceed end of input.
|
|
|
|
const BYTE *imatchLimit;
|
|
|
|
|
|
|
|
const BYTE *obase; /* Base of output */
|
|
|
|
BYTE *op; /* Output */
|
|
|
|
|
|
|
|
const BYTE *anchor; /* Anchor to start of current (match) block */
|
|
|
|
|
|
|
|
LDM_compressStats stats; /* Compression statistics */
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
hashEntry hashTable[LDM_HASHTABLESIZE_U32];
|
2017-07-10 19:38:27 +00:00
|
|
|
|
|
|
|
const BYTE *lastPosHashed; /* Last position hashed */
|
|
|
|
hash_t lastHash; /* Hash corresponding to lastPosHashed */
|
2017-07-12 22:11:06 +00:00
|
|
|
U32 lastSum;
|
|
|
|
|
|
|
|
const BYTE *nextIp; // TODO: this is redundant (ip + step)
|
2017-07-12 16:47:00 +00:00
|
|
|
const BYTE *nextPosHashed;
|
|
|
|
hash_t nextHash; /* Hash corresponding to nextPosHashed */
|
2017-07-12 01:13:26 +00:00
|
|
|
U32 nextSum;
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
unsigned step; // ip step, should be 1.
|
2017-07-12 01:13:26 +00:00
|
|
|
|
|
|
|
// DEBUG
|
|
|
|
const BYTE *DEBUG_setNextHash;
|
2017-07-10 19:38:27 +00:00
|
|
|
} LDM_CCtx;
|
2017-07-05 20:57:07 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Outputs compression statistics.
|
|
|
|
*/
|
|
|
|
static void printCompressStats(const LDM_CCtx *cctx) {
|
|
|
|
const LDM_compressStats *stats = &(cctx->stats);
|
|
|
|
#ifdef COMPUTE_STATS
|
|
|
|
printf("=====================\n");
|
|
|
|
printf("Compression statistics\n");
|
|
|
|
printf("Total number of matches: %u\n", stats->numMatches);
|
|
|
|
printf("Average match length: %.1f\n", ((double)stats->totalMatchLength) /
|
|
|
|
(double)stats->numMatches);
|
|
|
|
printf("Average literal length: %.1f\n",
|
|
|
|
((double)stats->totalLiteralLength) / (double)stats->numMatches);
|
|
|
|
printf("Average offset length: %.1f\n",
|
|
|
|
((double)stats->totalOffset) / (double)stats->numMatches);
|
|
|
|
printf("Num collisions, num hash inserts, %% collisions: %u, %u, %.3f\n",
|
|
|
|
stats->numCollisions, stats->numHashInserts,
|
|
|
|
stats->numHashInserts == 0 ?
|
|
|
|
1.0 : (100.0 * (double)stats->numCollisions) /
|
|
|
|
(double)stats->numHashInserts);
|
|
|
|
|
|
|
|
// Output occupancy of hash table.
|
|
|
|
{
|
|
|
|
U32 i = 0;
|
|
|
|
U32 ctr = 0;
|
|
|
|
for (; i < LDM_HASHTABLESIZE_U32; i++) {
|
|
|
|
if ((cctx->hashTable)[i].offset == 0) {
|
|
|
|
ctr++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
printf("Hash table size, empty slots, %% empty: %u %u %.3f\n",
|
|
|
|
LDM_HASHTABLESIZE_U32, ctr,
|
|
|
|
100.0 * (double)(ctr) / (double)LDM_HASHTABLESIZE_U32);
|
|
|
|
}
|
|
|
|
|
|
|
|
printf("=====================\n");
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Checks whether the MINMATCH bytes from p are the same as the MINMATCH
|
|
|
|
* bytes from match.
|
|
|
|
*
|
|
|
|
* This assumes MINMATCH is a multiple of four.
|
|
|
|
*
|
|
|
|
* Return 1 if valid, 0 otherwise.
|
|
|
|
*/
|
2017-07-12 01:13:26 +00:00
|
|
|
static int LDM_isValidMatch(const BYTE *p, const BYTE *match) {
|
2017-07-12 22:11:06 +00:00
|
|
|
/*
|
|
|
|
if (memcmp(p, match, MINMATCH) == 0) {
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
*/
|
|
|
|
|
|
|
|
//TODO: This seems to be faster for some reason?
|
2017-07-12 01:13:26 +00:00
|
|
|
U16 lengthLeft = MINMATCH;
|
|
|
|
const BYTE *curP = p;
|
|
|
|
const BYTE *curMatch = match;
|
|
|
|
|
|
|
|
for (; lengthLeft >= 8; lengthLeft -= 8) {
|
|
|
|
if (LDM_read64(curP) != LDM_read64(curMatch)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
curP += 8;
|
|
|
|
curMatch += 8;
|
|
|
|
}
|
|
|
|
if (lengthLeft > 0) {
|
2017-07-12 17:53:19 +00:00
|
|
|
return (LDM_read32(curP) == LDM_read32(curMatch));
|
2017-07-12 01:13:26 +00:00
|
|
|
}
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2017-07-11 05:27:43 +00:00
|
|
|
/**
|
2017-07-12 22:11:06 +00:00
|
|
|
* Convert a sum computed from getChecksum to a hash value in the range
|
2017-07-11 05:27:43 +00:00
|
|
|
* of the hash table.
|
|
|
|
*/
|
2017-07-12 22:11:06 +00:00
|
|
|
static hash_t checksumToHash(U32 sum) {
|
|
|
|
return ((sum * 2654435761U) >> ((32)-LDM_HASHLOG));
|
2017-07-11 05:27:43 +00:00
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Computes a checksum based on rsync's checksum.
|
|
|
|
*
|
|
|
|
* a(k,l) = \sum_{i = k}^l x_i (mod M)
|
|
|
|
* b(k,l) = \sum_{i = k}^l ((l - i + 1) * x_i) (mod M)
|
|
|
|
* checksum(k,l) = a(k,l) + 2^{16} * b(k,l)
|
|
|
|
*/
|
|
|
|
static U32 getChecksum(const char *data, U32 len) {
|
2017-07-11 05:27:43 +00:00
|
|
|
U32 i;
|
|
|
|
U32 s1, s2;
|
|
|
|
const schar *buf = (const schar *)data;
|
|
|
|
|
|
|
|
s1 = s2 = 0;
|
|
|
|
for (i = 0; i < (len - 4); i += 4) {
|
|
|
|
s2 += (4 * (s1 + buf[i])) + (3 * buf[i + 1]) +
|
2017-07-13 01:47:26 +00:00
|
|
|
(2 * buf[i + 2]) + (buf[i + 3]) +
|
|
|
|
(10 * CHECKSUM_CHAR_OFFSET);
|
|
|
|
s1 += buf[i] + buf[i + 1] + buf[i + 2] + buf[i + 3] +
|
|
|
|
+ (4 * CHECKSUM_CHAR_OFFSET);
|
|
|
|
|
2017-07-11 05:27:43 +00:00
|
|
|
}
|
|
|
|
for(; i < len; i++) {
|
2017-07-13 01:47:26 +00:00
|
|
|
s1 += buf[i] + CHECKSUM_CHAR_OFFSET;
|
2017-07-11 05:27:43 +00:00
|
|
|
s2 += s1;
|
|
|
|
}
|
|
|
|
return (s1 & 0xffff) + (s2 << 16);
|
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Update a checksum computed from getChecksum(data, len).
|
|
|
|
*
|
|
|
|
* The checksum can be updated along its ends as follows:
|
|
|
|
* a(k+1, l+1) = (a(k,l) - x_k + x_{l+1}) (mod M)
|
|
|
|
* b(k+1, l+1) = (b(k,l) - (l-k+1)*x_k + (a(k+1,l+1)) (mod M)
|
|
|
|
*
|
|
|
|
* Thus toRemove should correspond to data[0].
|
|
|
|
*/
|
|
|
|
static U32 updateChecksum(U32 sum, U32 len,
|
|
|
|
schar toRemove, schar toAdd) {
|
2017-07-12 01:13:26 +00:00
|
|
|
U32 s1 = (sum & 0xffff) - toRemove + toAdd;
|
2017-07-13 01:47:26 +00:00
|
|
|
U32 s2 = (sum >> 16) - ((toRemove + CHECKSUM_CHAR_OFFSET) * len) + s1;
|
2017-07-12 01:13:26 +00:00
|
|
|
|
|
|
|
return (s1 & 0xffff) + (s2 << 16);
|
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Update cctx->nextSum, cctx->nextHash, and cctx->nextPosHashed
|
|
|
|
* based on cctx->lastSum and cctx->lastPosHashed.
|
|
|
|
*
|
|
|
|
* This uses a rolling hash and requires that the last position hashed
|
|
|
|
* corresponds to cctx->nextIp - step.
|
|
|
|
*/
|
|
|
|
static void setNextHash(LDM_CCtx *cctx) {
|
2017-07-12 01:13:26 +00:00
|
|
|
#ifdef RUN_CHECKS
|
2017-07-12 16:47:00 +00:00
|
|
|
U32 check;
|
2017-07-12 01:13:26 +00:00
|
|
|
if ((cctx->nextIp - cctx->ibase != 1) &&
|
|
|
|
(cctx->nextIp - cctx->DEBUG_setNextHash != 1)) {
|
|
|
|
printf("CHECK debug fail: %zu %zu\n", cctx->nextIp - cctx->ibase,
|
|
|
|
cctx->DEBUG_setNextHash - cctx->ibase);
|
|
|
|
}
|
|
|
|
|
|
|
|
cctx->DEBUG_setNextHash = cctx->nextIp;
|
|
|
|
#endif
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
// cctx->nextSum = getChecksum((const char *)cctx->nextIp, LDM_HASH_LENGTH);
|
|
|
|
cctx->nextSum = updateChecksum(
|
2017-07-12 01:13:26 +00:00
|
|
|
cctx->lastSum, LDM_HASH_LENGTH,
|
|
|
|
(schar)((cctx->lastPosHashed)[0]),
|
|
|
|
(schar)((cctx->lastPosHashed)[LDM_HASH_LENGTH]));
|
2017-07-12 22:11:06 +00:00
|
|
|
cctx->nextPosHashed = cctx->nextIp;
|
|
|
|
cctx->nextHash = checksumToHash(cctx->nextSum);
|
2017-07-12 01:13:26 +00:00
|
|
|
|
|
|
|
#ifdef RUN_CHECKS
|
2017-07-12 22:11:06 +00:00
|
|
|
check = getChecksum((const char *)cctx->nextIp, LDM_HASH_LENGTH);
|
2017-07-12 16:47:00 +00:00
|
|
|
|
2017-07-12 01:13:26 +00:00
|
|
|
if (check != cctx->nextSum) {
|
|
|
|
printf("CHECK: setNextHash failed %u %u\n", check, cctx->nextSum);
|
|
|
|
}
|
|
|
|
|
|
|
|
if ((cctx->nextIp - cctx->lastPosHashed) != 1) {
|
|
|
|
printf("setNextHash: nextIp != lastPosHashed + 1. %zu %zu %zu\n",
|
|
|
|
cctx->nextIp - cctx->ibase, cctx->lastPosHashed - cctx->ibase,
|
|
|
|
cctx->ip - cctx->ibase);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
static void putHashOfCurrentPositionFromHash(
|
2017-07-12 01:13:26 +00:00
|
|
|
LDM_CCtx *cctx, hash_t hash, U32 sum) {
|
|
|
|
#ifdef COMPUTE_STATS
|
2017-07-12 22:11:06 +00:00
|
|
|
if (cctx->stats.numHashInserts < LDM_HASHTABLESIZE_U32) {
|
2017-07-12 01:13:26 +00:00
|
|
|
offset_t offset = (cctx->hashTable)[hash].offset;
|
|
|
|
cctx->stats.numHashInserts++;
|
2017-07-12 17:53:19 +00:00
|
|
|
if (offset != 0 && !LDM_isValidMatch(cctx->ip, offset + cctx->ibase)) {
|
2017-07-12 01:13:26 +00:00
|
|
|
cctx->stats.numCollisions++;
|
|
|
|
}
|
|
|
|
}
|
2017-07-12 17:53:19 +00:00
|
|
|
#endif
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
// Hash only every HASH_ONLY_EVERY times, based on cctx->ip.
|
|
|
|
// Note: this works only when cctx->step is 1.
|
|
|
|
if (((cctx->ip - cctx->ibase) & HASH_ONLY_EVERY) == HASH_ONLY_EVERY) {
|
|
|
|
(cctx->hashTable)[hash] = (hashEntry){ (offset_t)(cctx->ip - cctx->ibase) };
|
2017-07-12 17:53:19 +00:00
|
|
|
}
|
|
|
|
|
2017-07-12 01:13:26 +00:00
|
|
|
cctx->lastPosHashed = cctx->ip;
|
|
|
|
cctx->lastHash = hash;
|
|
|
|
cctx->lastSum = sum;
|
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Copy over the cctx->lastHash, cctx->lastSum, and cctx->lastPosHashed
|
|
|
|
* fields from the "next" fields.
|
|
|
|
*
|
|
|
|
* This requires that cctx->ip == cctx->nextPosHashed.
|
|
|
|
*/
|
2017-07-12 16:47:00 +00:00
|
|
|
static void LDM_updateLastHashFromNextHash(LDM_CCtx *cctx) {
|
|
|
|
#ifdef RUN_CHECKS
|
|
|
|
if (cctx->ip != cctx->nextPosHashed) {
|
2017-07-12 22:11:06 +00:00
|
|
|
printf("CHECK failed: updateLastHashFromNextHash %zu\n",
|
|
|
|
cctx->ip - cctx->ibase);
|
2017-07-12 16:47:00 +00:00
|
|
|
}
|
|
|
|
#endif
|
2017-07-12 22:11:06 +00:00
|
|
|
putHashOfCurrentPositionFromHash(cctx, cctx->nextHash, cctx->nextSum);
|
2017-07-12 16:47:00 +00:00
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Insert hash of the current position into the hash table.
|
|
|
|
*/
|
2017-07-12 01:13:26 +00:00
|
|
|
static void LDM_putHashOfCurrentPosition(LDM_CCtx *cctx) {
|
2017-07-12 22:11:06 +00:00
|
|
|
U32 sum = getChecksum((const char *)cctx->ip, LDM_HASH_LENGTH);
|
|
|
|
hash_t hash = checksumToHash(sum);
|
2017-07-12 01:13:26 +00:00
|
|
|
|
|
|
|
#ifdef RUN_CHECKS
|
2017-07-12 22:11:06 +00:00
|
|
|
if (cctx->nextPosHashed != cctx->ip && (cctx->ip != cctx->ibase)) {
|
|
|
|
printf("CHECK failed: putHashOfCurrentPosition %zu\n",
|
|
|
|
cctx->ip - cctx->ibase);
|
2017-07-12 01:13:26 +00:00
|
|
|
}
|
|
|
|
#endif
|
2017-07-10 19:38:27 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
putHashOfCurrentPositionFromHash(cctx, hash, sum);
|
2017-07-10 22:48:47 +00:00
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Returns the position of the entry at hashTable[hash].
|
|
|
|
*/
|
|
|
|
static const BYTE *getPositionOnHash(LDM_CCtx *cctx, hash_t hash) {
|
|
|
|
return cctx->hashTable[hash].offset + cctx->ibase;
|
2017-07-06 22:23:15 +00:00
|
|
|
}
|
2017-07-05 20:57:07 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Counts the number of bytes that match from pIn and pMatch,
|
|
|
|
* up to pInLimit.
|
|
|
|
*
|
|
|
|
* TODO: make more efficient.
|
|
|
|
*/
|
|
|
|
static unsigned countMatchLength(const BYTE *pIn, const BYTE *pMatch,
|
|
|
|
const BYTE *pInLimit) {
|
2017-07-06 22:23:15 +00:00
|
|
|
const BYTE * const pStart = pIn;
|
|
|
|
while (pIn < pInLimit - 1) {
|
2017-07-12 01:13:26 +00:00
|
|
|
BYTE const diff = LDM_readByte(pMatch) ^ LDM_readByte(pIn);
|
2017-07-06 22:23:15 +00:00
|
|
|
if (!diff) {
|
|
|
|
pIn++;
|
|
|
|
pMatch++;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
return (unsigned)(pIn - pStart);
|
|
|
|
}
|
|
|
|
return (unsigned)(pIn - pStart);
|
|
|
|
}
|
|
|
|
|
2017-07-13 01:47:26 +00:00
|
|
|
void LDM_readHeader(const void *src, U64 *compressSize,
|
|
|
|
U64 *decompressSize) {
|
|
|
|
const U64 *ip = (const U64 *)src;
|
2017-07-10 13:50:49 +00:00
|
|
|
*compressSize = *ip++;
|
|
|
|
*decompressSize = *ip;
|
2017-07-07 19:44:29 +00:00
|
|
|
}
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Initialize a compression context.
|
|
|
|
*/
|
|
|
|
static void initializeCCtx(LDM_CCtx *cctx,
|
|
|
|
const void *src, size_t srcSize,
|
|
|
|
void *dst, size_t maxDstSize) {
|
2017-07-10 19:38:27 +00:00
|
|
|
cctx->isize = srcSize;
|
|
|
|
cctx->maxOSize = maxDstSize;
|
2017-07-08 00:09:28 +00:00
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
cctx->ibase = (const BYTE *)src;
|
|
|
|
cctx->ip = cctx->ibase;
|
|
|
|
cctx->iend = cctx->ibase + srcSize;
|
2017-07-08 00:09:28 +00:00
|
|
|
|
2017-07-12 01:13:26 +00:00
|
|
|
cctx->ihashLimit = cctx->iend - LDM_HASH_LENGTH;
|
2017-07-10 19:38:27 +00:00
|
|
|
cctx->imatchLimit = cctx->iend - MINMATCH;
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
cctx->obase = (BYTE *)dst;
|
2017-07-10 20:08:19 +00:00
|
|
|
cctx->op = (BYTE *)dst;
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
cctx->anchor = cctx->ibase;
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
memset(&(cctx->stats), 0, sizeof(cctx->stats));
|
|
|
|
memset(cctx->hashTable, 0, sizeof(cctx->hashTable));
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
cctx->lastPosHashed = NULL;
|
2017-07-10 22:48:47 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
cctx->step = 1; // Fixed to be 1 for now. Changing may break things.
|
2017-07-12 01:13:26 +00:00
|
|
|
cctx->nextIp = cctx->ip + cctx->step;
|
2017-07-12 16:47:00 +00:00
|
|
|
cctx->nextPosHashed = 0;
|
2017-07-12 01:13:26 +00:00
|
|
|
|
|
|
|
cctx->DEBUG_setNextHash = 0;
|
2017-07-10 22:48:47 +00:00
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/**
|
|
|
|
* Finds the "best" match.
|
|
|
|
*
|
|
|
|
* Returns 0 if successful and 1 otherwise (i.e. no match can be found
|
|
|
|
* in the remaining input that is long enough).
|
|
|
|
*
|
|
|
|
*/
|
2017-07-12 01:13:26 +00:00
|
|
|
static int LDM_findBestMatch(LDM_CCtx *cctx, const BYTE **match) {
|
|
|
|
cctx->nextIp = cctx->ip + cctx->step;
|
|
|
|
|
|
|
|
do {
|
|
|
|
hash_t h;
|
|
|
|
U32 sum;
|
2017-07-12 22:11:06 +00:00
|
|
|
setNextHash(cctx);
|
2017-07-12 01:13:26 +00:00
|
|
|
h = cctx->nextHash;
|
|
|
|
sum = cctx->nextSum;
|
|
|
|
cctx->ip = cctx->nextIp;
|
|
|
|
cctx->nextIp += cctx->step;
|
|
|
|
|
|
|
|
if (cctx->ip > cctx->imatchLimit) {
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
*match = getPositionOnHash(cctx, h);
|
|
|
|
putHashOfCurrentPositionFromHash(cctx, h, sum);
|
2017-07-12 01:13:26 +00:00
|
|
|
|
2017-07-10 22:48:47 +00:00
|
|
|
} while (cctx->ip - *match > WINDOW_SIZE ||
|
2017-07-12 01:13:26 +00:00
|
|
|
!LDM_isValidMatch(cctx->ip, *match));
|
2017-07-12 22:11:06 +00:00
|
|
|
setNextHash(cctx);
|
2017-07-10 22:48:47 +00:00
|
|
|
return 0;
|
2017-07-10 19:38:27 +00:00
|
|
|
}
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-12 01:13:26 +00:00
|
|
|
/**
|
|
|
|
* Write current block (literals, literal length, match offset,
|
|
|
|
* match length).
|
|
|
|
*
|
2017-07-12 22:11:06 +00:00
|
|
|
* Update input pointer, inserting hashes into hash table along the way.
|
2017-07-12 01:13:26 +00:00
|
|
|
*/
|
2017-07-12 22:11:06 +00:00
|
|
|
static void outputBlock(LDM_CCtx *cctx,
|
|
|
|
unsigned const literalLength,
|
|
|
|
unsigned const offset,
|
|
|
|
unsigned const matchLength) {
|
2017-07-12 01:13:26 +00:00
|
|
|
BYTE *token = cctx->op++;
|
|
|
|
|
|
|
|
/* Encode the literal length. */
|
|
|
|
if (literalLength >= RUN_MASK) {
|
|
|
|
int len = (int)literalLength - RUN_MASK;
|
|
|
|
*token = (RUN_MASK << ML_BITS);
|
|
|
|
for (; len >= 255; len -= 255) {
|
|
|
|
*(cctx->op)++ = 255;
|
|
|
|
}
|
|
|
|
*(cctx->op)++ = (BYTE)len;
|
|
|
|
} else {
|
|
|
|
*token = (BYTE)(literalLength << ML_BITS);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Encode the literals. */
|
|
|
|
memcpy(cctx->op, cctx->anchor, literalLength);
|
|
|
|
cctx->op += literalLength;
|
|
|
|
|
|
|
|
/* Encode the offset. */
|
|
|
|
LDM_write32(cctx->op, offset);
|
|
|
|
cctx->op += LDM_OFFSET_SIZE;
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/* Encode the match length. */
|
2017-07-12 01:13:26 +00:00
|
|
|
if (matchLength >= ML_MASK) {
|
|
|
|
unsigned matchLengthRemaining = matchLength;
|
|
|
|
*token += ML_MASK;
|
|
|
|
matchLengthRemaining -= ML_MASK;
|
|
|
|
LDM_write32(cctx->op, 0xFFFFFFFF);
|
|
|
|
while (matchLengthRemaining >= 4*0xFF) {
|
|
|
|
cctx->op += 4;
|
|
|
|
LDM_write32(cctx->op, 0xffffffff);
|
|
|
|
matchLengthRemaining -= 4*0xFF;
|
|
|
|
}
|
|
|
|
cctx->op += matchLengthRemaining / 255;
|
|
|
|
*(cctx->op)++ = (BYTE)(matchLengthRemaining % 255);
|
|
|
|
} else {
|
|
|
|
*token += (BYTE)(matchLength);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2017-07-13 17:38:19 +00:00
|
|
|
// TODO: maxDstSize is unused. This function may seg fault when writing
|
|
|
|
// beyond the size of dst, as it does not check maxDstSize. Writing to
|
|
|
|
// a buffer and performing checks is a possible solution.
|
|
|
|
//
|
2017-07-12 22:11:06 +00:00
|
|
|
// This is based upon lz4.
|
2017-07-10 19:38:27 +00:00
|
|
|
size_t LDM_compress(const void *src, size_t srcSize,
|
|
|
|
void *dst, size_t maxDstSize) {
|
|
|
|
LDM_CCtx cctx;
|
2017-07-12 22:11:06 +00:00
|
|
|
initializeCCtx(&cctx, src, srcSize, dst, maxDstSize);
|
2017-07-10 19:38:27 +00:00
|
|
|
|
|
|
|
/* Hash the first position and put it into the hash table. */
|
|
|
|
LDM_putHashOfCurrentPosition(&cctx);
|
|
|
|
|
|
|
|
// TODO: loop condition is not accurate.
|
|
|
|
while (1) {
|
2017-07-06 22:23:15 +00:00
|
|
|
const BYTE *match;
|
|
|
|
|
2017-07-10 22:48:47 +00:00
|
|
|
/**
|
|
|
|
* Find a match.
|
|
|
|
* If no more matches can be found (i.e. the length of the remaining input
|
|
|
|
* is less than the minimum match length), then stop searching for matches
|
|
|
|
* and encode the final literals.
|
|
|
|
*/
|
|
|
|
if (LDM_findBestMatch(&cctx, &match) != 0) {
|
|
|
|
goto _last_literals;
|
2017-07-06 22:23:15 +00:00
|
|
|
}
|
2017-07-12 22:11:06 +00:00
|
|
|
#ifdef COMPUTE_STATS
|
2017-07-10 22:48:47 +00:00
|
|
|
cctx.stats.numMatches++;
|
2017-07-12 22:11:06 +00:00
|
|
|
#endif
|
2017-07-10 22:48:47 +00:00
|
|
|
|
|
|
|
/**
|
2017-07-11 05:27:43 +00:00
|
|
|
* Catch up: look back to extend the match backwards from the found match.
|
2017-07-10 22:48:47 +00:00
|
|
|
*/
|
|
|
|
while (cctx.ip > cctx.anchor && match > cctx.ibase &&
|
|
|
|
cctx.ip[-1] == match[-1]) {
|
2017-07-10 19:38:27 +00:00
|
|
|
cctx.ip--;
|
2017-07-07 21:14:01 +00:00
|
|
|
match--;
|
|
|
|
}
|
|
|
|
|
2017-07-10 22:48:47 +00:00
|
|
|
/**
|
|
|
|
* Write current block (literals, literal length, match offset, match
|
|
|
|
* length) and update pointers and hashes.
|
|
|
|
*/
|
2017-07-12 22:11:06 +00:00
|
|
|
{
|
|
|
|
unsigned const literalLength = (unsigned)(cctx.ip - cctx.anchor);
|
|
|
|
unsigned const offset = cctx.ip - match;
|
|
|
|
unsigned const matchLength = countMatchLength(
|
|
|
|
cctx.ip + MINMATCH, match + MINMATCH, cctx.ihashLimit);
|
|
|
|
|
|
|
|
#ifdef COMPUTE_STATS
|
|
|
|
cctx.stats.totalLiteralLength += literalLength;
|
|
|
|
cctx.stats.totalOffset += offset;
|
|
|
|
cctx.stats.totalMatchLength += matchLength + MINMATCH;
|
|
|
|
#endif
|
|
|
|
outputBlock(&cctx, literalLength, offset, matchLength);
|
|
|
|
|
|
|
|
// Move ip to end of block, inserting hashes at each position.
|
|
|
|
cctx.nextIp = cctx.ip + cctx.step;
|
|
|
|
while (cctx.ip < cctx.anchor + MINMATCH + matchLength + literalLength) {
|
|
|
|
if (cctx.ip > cctx.lastPosHashed) {
|
|
|
|
// TODO: Simplify.
|
|
|
|
LDM_updateLastHashFromNextHash(&cctx);
|
|
|
|
setNextHash(&cctx);
|
|
|
|
}
|
|
|
|
cctx.ip++;
|
|
|
|
cctx.nextIp++;
|
|
|
|
}
|
|
|
|
}
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-10 22:48:47 +00:00
|
|
|
// Set start of next block to current input pointer.
|
2017-07-10 19:38:27 +00:00
|
|
|
cctx.anchor = cctx.ip;
|
2017-07-12 16:47:00 +00:00
|
|
|
LDM_updateLastHashFromNextHash(&cctx);
|
2017-07-06 22:23:15 +00:00
|
|
|
}
|
2017-07-07 21:14:01 +00:00
|
|
|
_last_literals:
|
2017-07-10 22:48:47 +00:00
|
|
|
/* Encode the last literals (no more matches). */
|
2017-07-06 22:23:15 +00:00
|
|
|
{
|
2017-07-10 19:38:27 +00:00
|
|
|
size_t const lastRun = (size_t)(cctx.iend - cctx.anchor);
|
2017-07-06 22:23:15 +00:00
|
|
|
if (lastRun >= RUN_MASK) {
|
|
|
|
size_t accumulator = lastRun - RUN_MASK;
|
2017-07-10 19:38:27 +00:00
|
|
|
*(cctx.op)++ = RUN_MASK << ML_BITS;
|
2017-07-06 22:23:15 +00:00
|
|
|
for(; accumulator >= 255; accumulator -= 255) {
|
2017-07-10 19:38:27 +00:00
|
|
|
*(cctx.op)++ = 255;
|
2017-07-06 22:23:15 +00:00
|
|
|
}
|
2017-07-10 19:38:27 +00:00
|
|
|
*(cctx.op)++ = (BYTE)accumulator;
|
2017-07-06 22:23:15 +00:00
|
|
|
} else {
|
2017-07-10 19:38:27 +00:00
|
|
|
*(cctx.op)++ = (BYTE)(lastRun << ML_BITS);
|
2017-07-06 22:23:15 +00:00
|
|
|
}
|
2017-07-10 19:38:27 +00:00
|
|
|
memcpy(cctx.op, cctx.anchor, lastRun);
|
|
|
|
cctx.op += lastRun;
|
2017-07-06 22:23:15 +00:00
|
|
|
}
|
2017-07-12 17:53:19 +00:00
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
#ifdef COMPUTE_STATS
|
|
|
|
printCompressStats(&cctx);
|
|
|
|
#endif
|
|
|
|
|
2017-07-10 20:08:19 +00:00
|
|
|
return (cctx.op - (const BYTE *)cctx.obase);
|
2017-07-06 22:23:15 +00:00
|
|
|
}
|
|
|
|
|
2017-07-10 13:50:49 +00:00
|
|
|
typedef struct LDM_DCtx {
|
2017-07-10 14:38:09 +00:00
|
|
|
size_t compressSize;
|
|
|
|
size_t maxDecompressSize;
|
2017-07-10 19:38:27 +00:00
|
|
|
|
|
|
|
const BYTE *ibase; /* Base of input */
|
|
|
|
const BYTE *ip; /* Current input position */
|
|
|
|
const BYTE *iend; /* End of source */
|
|
|
|
|
|
|
|
const BYTE *obase; /* Base of output */
|
|
|
|
BYTE *op; /* Current output position */
|
|
|
|
const BYTE *oend; /* End of output */
|
2017-07-10 13:50:49 +00:00
|
|
|
} LDM_DCtx;
|
|
|
|
|
2017-07-10 14:38:09 +00:00
|
|
|
static void LDM_initializeDCtx(LDM_DCtx *dctx,
|
|
|
|
const void *src, size_t compressSize,
|
|
|
|
void *dst, size_t maxDecompressSize) {
|
2017-07-10 19:38:27 +00:00
|
|
|
dctx->compressSize = compressSize;
|
|
|
|
dctx->maxDecompressSize = maxDecompressSize;
|
|
|
|
|
2017-07-10 14:38:09 +00:00
|
|
|
dctx->ibase = src;
|
|
|
|
dctx->ip = (const BYTE *)src;
|
2017-07-10 19:38:27 +00:00
|
|
|
dctx->iend = dctx->ip + dctx->compressSize;
|
2017-07-10 14:38:09 +00:00
|
|
|
dctx->op = dst;
|
2017-07-10 19:38:27 +00:00
|
|
|
dctx->oend = dctx->op + dctx->maxDecompressSize;
|
2017-07-10 14:38:09 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
size_t LDM_decompress(const void *src, size_t compressSize,
|
|
|
|
void *dst, size_t maxDecompressSize) {
|
|
|
|
LDM_DCtx dctx;
|
|
|
|
LDM_initializeDCtx(&dctx, src, compressSize, dst, maxDecompressSize);
|
|
|
|
|
|
|
|
while (dctx.ip < dctx.iend) {
|
2017-07-10 20:08:19 +00:00
|
|
|
BYTE *cpy;
|
|
|
|
const BYTE *match;
|
2017-07-10 22:48:47 +00:00
|
|
|
size_t length, offset;
|
2017-07-10 20:08:19 +00:00
|
|
|
|
2017-07-10 22:48:47 +00:00
|
|
|
/* Get the literal length. */
|
2017-07-10 14:38:09 +00:00
|
|
|
unsigned const token = *(dctx.ip)++;
|
2017-07-10 19:38:27 +00:00
|
|
|
if ((length = (token >> ML_BITS)) == RUN_MASK) {
|
2017-07-06 22:23:15 +00:00
|
|
|
unsigned s;
|
|
|
|
do {
|
2017-07-10 14:38:09 +00:00
|
|
|
s = *(dctx.ip)++;
|
2017-07-06 22:23:15 +00:00
|
|
|
length += s;
|
|
|
|
} while (s == 255);
|
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/* Copy the literals. */
|
2017-07-10 14:38:09 +00:00
|
|
|
cpy = dctx.op + length;
|
|
|
|
memcpy(dctx.op, dctx.ip, length);
|
|
|
|
dctx.ip += length;
|
|
|
|
dctx.op = cpy;
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-10 19:38:27 +00:00
|
|
|
//TODO : dynamic offset size
|
2017-07-10 14:38:09 +00:00
|
|
|
offset = LDM_read32(dctx.ip);
|
2017-07-10 22:48:47 +00:00
|
|
|
dctx.ip += LDM_OFFSET_SIZE;
|
2017-07-10 14:38:09 +00:00
|
|
|
match = dctx.op - offset;
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-10 22:48:47 +00:00
|
|
|
/* Get the match length. */
|
2017-07-06 22:23:15 +00:00
|
|
|
length = token & ML_MASK;
|
|
|
|
if (length == ML_MASK) {
|
|
|
|
unsigned s;
|
|
|
|
do {
|
2017-07-10 14:38:09 +00:00
|
|
|
s = *(dctx.ip)++;
|
2017-07-06 22:23:15 +00:00
|
|
|
length += s;
|
|
|
|
} while (s == 255);
|
|
|
|
}
|
|
|
|
length += MINMATCH;
|
2017-07-10 22:48:47 +00:00
|
|
|
|
|
|
|
/* Copy match. */
|
2017-07-10 14:38:09 +00:00
|
|
|
cpy = dctx.op + length;
|
2017-07-06 22:23:15 +00:00
|
|
|
|
2017-07-12 01:13:26 +00:00
|
|
|
// Inefficient for now.
|
2017-07-10 14:38:09 +00:00
|
|
|
while (match < cpy - offset && dctx.op < dctx.oend) {
|
|
|
|
*(dctx.op)++ = *match++;
|
2017-07-06 23:47:08 +00:00
|
|
|
}
|
2017-07-06 22:23:15 +00:00
|
|
|
}
|
2017-07-10 14:38:09 +00:00
|
|
|
return dctx.op - (BYTE *)dst;
|
2017-07-05 20:57:07 +00:00
|
|
|
}
|
|
|
|
|
2017-07-12 22:11:06 +00:00
|
|
|
/*
|
2017-07-12 01:13:26 +00:00
|
|
|
void LDM_test(const void *src, size_t srcSize,
|
|
|
|
void *dst, size_t maxDstSize) {
|
|
|
|
const BYTE *ip = (const BYTE *)src + 1125;
|
2017-07-12 22:11:06 +00:00
|
|
|
U32 sum = getChecksum((const char *)ip, LDM_HASH_LENGTH);
|
2017-07-12 01:13:26 +00:00
|
|
|
U32 sum2;
|
|
|
|
++ip;
|
|
|
|
for (; ip < (const BYTE *)src + 1125 + 100; ip++) {
|
2017-07-12 22:11:06 +00:00
|
|
|
sum2 = updateChecksum(sum, LDM_HASH_LENGTH,
|
2017-07-12 01:13:26 +00:00
|
|
|
ip[-1], ip[LDM_HASH_LENGTH - 1]);
|
2017-07-12 22:11:06 +00:00
|
|
|
sum = getChecksum((const char *)ip, LDM_HASH_LENGTH);
|
2017-07-12 01:13:26 +00:00
|
|
|
printf("TEST HASH: %zu %u %u\n", ip - (const BYTE *)src, sum, sum2);
|
|
|
|
}
|
|
|
|
}
|
2017-07-12 22:11:06 +00:00
|
|
|
*/
|
2017-07-12 01:13:26 +00:00
|
|
|
|
2017-07-05 20:57:07 +00:00
|
|
|
|