brotli/enc/hash_longest_match_inc.h
2016-06-13 16:28:19 +02:00

287 lines
10 KiB
C

/* NOLINT(build/header_guard) */
/* Copyright 2010 Google Inc. All Rights Reserved.
Distributed under MIT license.
See file LICENSE for detail or copy at https://opensource.org/licenses/MIT
*/
/* template parameters: FN, BUCKET_BITS, BLOCK_BITS,
NUM_LAST_DISTANCES_TO_CHECK */
/* A (forgetful) hash table to the data seen by the compressor, to
help create backward references to previous data.
This is a hash map of fixed size (BUCKET_SIZE) to a ring buffer of
fixed size (BLOCK_SIZE). The ring buffer contains the last BLOCK_SIZE
index positions of the given hash key in the compressed data. */
#define HashLongestMatch HASHER()
/* Number of hash buckets. */
#define BUCKET_SIZE (1 << BUCKET_BITS)
/* Only BLOCK_SIZE newest backward references are kept,
and the older are forgotten. */
#define BLOCK_SIZE (1u << BLOCK_BITS)
/* Mask for accessing entries in a block (in a ringbuffer manner). */
#define BLOCK_MASK ((1 << BLOCK_BITS) - 1)
#define HASH_MAP_SIZE (2 << BUCKET_BITS)
static BROTLI_INLINE size_t FN(HashTypeLength)(void) { return 4; }
static BROTLI_INLINE size_t FN(StoreLookahead)(void) { return 4; }
/* HashBytes is the function that chooses the bucket to place
the address in. The HashLongestMatch and HashLongestMatchQuickly
classes have separate, different implementations of hashing. */
static uint32_t FN(HashBytes)(const uint8_t *data) {
uint32_t h = BROTLI_UNALIGNED_LOAD32(data) * kHashMul32;
/* The higher bits contain more mixture from the multiplication,
so we take our results from there. */
return h >> (32 - BUCKET_BITS);
}
typedef struct HashLongestMatch {
/* Number of entries in a particular bucket. */
uint16_t num_[BUCKET_SIZE];
/* Buckets containing BLOCK_SIZE of backward references. */
uint32_t buckets_[BLOCK_SIZE << BUCKET_BITS];
/* True if num_ array needs to be initialized. */
int is_dirty_;
size_t num_dict_lookups_;
size_t num_dict_matches_;
} HashLongestMatch;
static void FN(Reset)(HashLongestMatch* self) {
self->is_dirty_ = 1;
self->num_dict_lookups_ = 0;
self->num_dict_matches_ = 0;
}
static void FN(InitEmpty)(HashLongestMatch* self) {
if (self->is_dirty_) {
memset(self->num_, 0, sizeof(self->num_));
self->is_dirty_ = 0;
}
}
static void FN(InitForData)(HashLongestMatch* self, const uint8_t* data,
size_t num) {
size_t i;
for (i = 0; i < num; ++i) {
const uint32_t key = FN(HashBytes)(&data[i]);
self->num_[key] = 0;
}
if (num != 0) {
self->is_dirty_ = 0;
}
}
static void FN(Init)(
MemoryManager* m, HashLongestMatch* self, const uint8_t* data, int lgwin,
size_t position, size_t bytes, int is_last) {
/* Choose which init method is faster.
Init() is about 100 times faster than InitForData(). */
const size_t kMaxBytesForPartialHashInit = HASH_MAP_SIZE >> 7;
BROTLI_UNUSED(m);
BROTLI_UNUSED(lgwin);
if (position == 0 && is_last && bytes <= kMaxBytesForPartialHashInit) {
FN(InitForData)(self, data, bytes);
} else {
FN(InitEmpty)(self);
}
}
/* Look at 4 bytes at &data[ix & mask].
Compute a hash from these, and store the value of ix at that position. */
static BROTLI_INLINE void FN(Store)(HashLongestMatch* self, const uint8_t *data,
const size_t mask, const size_t ix) {
const uint32_t key = FN(HashBytes)(&data[ix & mask]);
const size_t minor_ix = self->num_[key] & BLOCK_MASK;
self->buckets_[minor_ix + (key << BLOCK_BITS)] = (uint32_t)ix;
++self->num_[key];
}
static BROTLI_INLINE void FN(StoreRange)(HashLongestMatch* self,
const uint8_t *data, const size_t mask, const size_t ix_start,
const size_t ix_end) {
size_t i;
for (i = ix_start; i < ix_end; ++i) {
FN(Store)(self, data, mask, i);
}
}
static BROTLI_INLINE void FN(StitchToPreviousBlock)(HashLongestMatch* self,
size_t num_bytes, size_t position, const uint8_t* ringbuffer,
size_t ringbuffer_mask) {
if (num_bytes >= FN(HashTypeLength)() - 1 && position >= 3) {
/* Prepare the hashes for three last bytes of the last write.
These could not be calculated before, since they require knowledge
of both the previous and the current block. */
FN(Store)(self, ringbuffer, ringbuffer_mask, position - 3);
FN(Store)(self, ringbuffer, ringbuffer_mask, position - 2);
FN(Store)(self, ringbuffer, ringbuffer_mask, position - 1);
}
}
/* Find a longest backward match of &data[cur_ix] up to the length of
max_length and stores the position cur_ix in the hash table.
Does not look for matches longer than max_length.
Does not look for matches further away than max_backward.
Writes the best found match length into best_len_out.
Writes the index (&data[index]) offset from the start of the best match
into best_distance_out.
Write the score of the best match into best_score_out.
Returns 1 when match is found, otherwise 0. */
static BROTLI_INLINE int FN(FindLongestMatch)(HashLongestMatch* self,
const uint8_t* BROTLI_RESTRICT data, const size_t ring_buffer_mask,
const int* BROTLI_RESTRICT distance_cache, const size_t cur_ix,
const size_t max_length, const size_t max_backward,
size_t* BROTLI_RESTRICT best_len_out,
size_t* BROTLI_RESTRICT best_len_code_out,
size_t* BROTLI_RESTRICT best_distance_out,
double* BROTLI_RESTRICT best_score_out) {
const size_t cur_ix_masked = cur_ix & ring_buffer_mask;
int is_match_found = 0;
/* Don't accept a short copy from far away. */
double best_score = *best_score_out;
size_t best_len = *best_len_out;
size_t i;
*best_len_code_out = 0;
*best_len_out = 0;
/* Try last distance first. */
for (i = 0; i < NUM_LAST_DISTANCES_TO_CHECK; ++i) {
const size_t idx = kDistanceCacheIndex[i];
const size_t backward =
(size_t)(distance_cache[idx] + kDistanceCacheOffset[i]);
size_t prev_ix = (size_t)(cur_ix - backward);
if (prev_ix >= cur_ix) {
continue;
}
if (PREDICT_FALSE(backward > max_backward)) {
continue;
}
prev_ix &= ring_buffer_mask;
if (cur_ix_masked + best_len > ring_buffer_mask ||
prev_ix + best_len > ring_buffer_mask ||
data[cur_ix_masked + best_len] != data[prev_ix + best_len]) {
continue;
}
{
const size_t len = FindMatchLengthWithLimit(&data[prev_ix],
&data[cur_ix_masked],
max_length);
if (len >= 3 || (len == 2 && i < 2)) {
/* Comparing for >= 2 does not change the semantics, but just saves for
a few unnecessary binary logarithms in backward reference score,
since we are not interested in such short matches. */
double score = BackwardReferenceScoreUsingLastDistance(len, i);
if (best_score < score) {
best_score = score;
best_len = len;
*best_len_out = best_len;
*best_len_code_out = best_len;
*best_distance_out = backward;
*best_score_out = best_score;
is_match_found = 1;
}
}
}
}
{
const uint32_t key = FN(HashBytes)(&data[cur_ix_masked]);
const uint32_t * BROTLI_RESTRICT const bucket =
&self->buckets_[key << BLOCK_BITS];
const size_t down =
(self->num_[key] > BLOCK_SIZE) ? (self->num_[key] - BLOCK_SIZE) : 0u;
for (i = self->num_[key]; i > down;) {
size_t prev_ix = bucket[--i & BLOCK_MASK];
const size_t backward = cur_ix - prev_ix;
if (PREDICT_FALSE(backward == 0 || backward > max_backward)) {
break;
}
prev_ix &= ring_buffer_mask;
if (cur_ix_masked + best_len > ring_buffer_mask ||
prev_ix + best_len > ring_buffer_mask ||
data[cur_ix_masked + best_len] != data[prev_ix + best_len]) {
continue;
}
{
const size_t len = FindMatchLengthWithLimit(&data[prev_ix],
&data[cur_ix_masked],
max_length);
if (len >= 4) {
/* Comparing for >= 3 does not change the semantics, but just saves
for a few unnecessary binary logarithms in backward reference
score, since we are not interested in such short matches. */
double score = BackwardReferenceScore(len, backward);
if (best_score < score) {
best_score = score;
best_len = len;
*best_len_out = best_len;
*best_len_code_out = best_len;
*best_distance_out = backward;
*best_score_out = best_score;
is_match_found = 1;
}
}
}
}
self->buckets_[(key << BLOCK_BITS) + (self->num_[key] & BLOCK_MASK)] =
(uint32_t)cur_ix;
++self->num_[key];
}
if (!is_match_found &&
self->num_dict_matches_ >= (self->num_dict_lookups_ >> 7)) {
size_t dict_key = Hash14(&data[cur_ix_masked]) << 1;
int k;
for (k = 0; k < 2; ++k, ++dict_key) {
const uint16_t v = kStaticDictionaryHash[dict_key];
++self->num_dict_lookups_;
if (v > 0) {
const size_t len = v & 31;
const size_t dist = v >> 5;
const size_t offset =
kBrotliDictionaryOffsetsByLength[len] + len * dist;
if (len <= max_length) {
const size_t matchlen =
FindMatchLengthWithLimit(&data[cur_ix_masked],
&kBrotliDictionary[offset], len);
if (matchlen + kCutoffTransformsCount > len && matchlen > 0) {
const size_t transform_id = kCutoffTransforms[len - matchlen];
const size_t transform_step =
(size_t)1 << kBrotliDictionarySizeBitsByLength[len];
const size_t word_id = dist + transform_id * transform_step;
const size_t backward = max_backward + word_id + 1;
double score = BackwardReferenceScore(matchlen, backward);
if (best_score < score) {
++self->num_dict_matches_;
best_score = score;
best_len = matchlen;
*best_len_out = best_len;
*best_len_code_out = len;
*best_distance_out = backward;
*best_score_out = best_score;
is_match_found = 1;
}
}
}
}
}
}
return is_match_found;
}
#undef HASH_MAP_SIZE
#undef BLOCK_MASK
#undef BLOCK_SIZE
#undef BUCKET_SIZE
#undef HashLongestMatch