zumun/woff2
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Updates to Brotli compression format, decoder and encoder

Browse Source 
This commit contains a batch of changes that were made to the Brotli
compression algorithm in the last month. Most important changes:

   * Fixes to the spec.
   * Change of code length code order.
   * Use a 2-level Huffman lookup table in the decoder.
   * Faster uncompressed meta-block decoding.
   * Optimized encoding of the Huffman code.
   * Detection of UTF-8 input encoding.
   * UTF-8 based literal cost modeling for improved
     backward reference selection.
This commit is contained in:
Zoltan Szabadka
2014-02-14 15:04:23 +01:00
parent dfc5a9f215
commit cbd5cb55f4
16 changed files with 912 additions and 635 deletions
Download Patch File Download Diff File Expand all files Collapse all files
brotli/dec/huffman.c
+108 -201
View File
@@ -12,11 +12,12 @@
See the License for the specific language governing permissions and
limitations under the License.
Utilities for building and looking up Huffman trees.
Utilities for building Huffman decoding tables.
*/
#include <assert.h>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include "./huffman.h"
#include "./safe_malloc.h"
@@ -25,231 +26,137 @@
extern "C" {
#endif
#define NON_EXISTENT_SYMBOL (-1)
#define MAX_ALLOWED_CODE_LENGTH 15
#define MAX_LENGTH 15
static void TreeNodeInit(HuffmanTreeNode* const node) {
node->children_ = -1; /* means: 'unassigned so far' */
}
static int NodeIsEmpty(const HuffmanTreeNode* const node) {
return (node->children_ < 0);
}
static int IsFull(const HuffmanTree* const tree) {
return (tree->num_nodes_ == tree->max_nodes_);
}
static void AssignChildren(HuffmanTree* const tree,
HuffmanTreeNode* const node) {
HuffmanTreeNode* const children = tree->root_ + tree->num_nodes_;
node->children_ = (int)(children - node);
assert(children - node == (int)(children - node));
tree->num_nodes_ += 2;
TreeNodeInit(children + 0);
TreeNodeInit(children + 1);
}
static int TreeInit(HuffmanTree* const tree, int num_leaves) {
assert(tree != NULL);
tree->root_ = NULL;
if (num_leaves == 0) return 0;
/* We allocate maximum possible nodes in the tree at once. */
/* Note that a Huffman tree is a full binary tree; and in a full binary */
/* tree with L leaves, the total number of nodes N = 2 * L - 1. */
tree->max_nodes_ = 2 * num_leaves - 1;
assert(tree->max_nodes_ < (1 << 16)); /* limit for the lut_jump_ table */
tree->root_ = (HuffmanTreeNode*)BrotliSafeMalloc((uint64_t)tree->max_nodes_,
sizeof(*tree->root_));
if (tree->root_ == NULL) return 0;
TreeNodeInit(tree->root_); /* Initialize root. */
tree->num_nodes_ = 1;
memset(tree->lut_bits_, 255, sizeof(tree->lut_bits_));
memset(tree->lut_jump_, 0, sizeof(tree->lut_jump_));
return 1;
}
void BrotliHuffmanTreeRelease(HuffmanTree* const tree) {
if (tree != NULL) {
if (tree->root_ != NULL) {
free(tree->root_);
}
tree->root_ = NULL;
tree->max_nodes_ = 0;
tree->num_nodes_ = 0;
/* Returns reverse(reverse(key, len) + 1, len), where reverse(key, len) is the
bit-wise reversal of the len least significant bits of key. */
static BROTLI_INLINE int GetNextKey(int key, int len) {
int step = 1 << (len - 1);
while (key & step) {
step >>= 1;
}
return (key & (step - 1)) + step;
}
/* Utility: converts Huffman code lengths to corresponding Huffman codes. */
/* 'huff_codes' should be pre-allocated. */
/* Returns false in case of error (memory allocation, invalid codes). */
static int HuffmanCodeLengthsToCodes(const uint8_t* const code_lengths,
int code_lengths_size,
int* const huff_codes) {
int symbol;
int code_len;
int code_length_hist[MAX_ALLOWED_CODE_LENGTH + 1] = { 0 };
int curr_code;
int next_codes[MAX_ALLOWED_CODE_LENGTH + 1] = { 0 };
int max_code_length = 0;
/* Stores code in table[0], table[step], table[2*step], ..., table[end] */
/* Assumes that end is an integer multiple of step */
static BROTLI_INLINE void ReplicateValue(HuffmanCode* table,
int step, int end,
HuffmanCode code) {
do {
end -= step;
table[end] = code;
} while (end > 0);
}
assert(code_lengths != NULL);
assert(code_lengths_size > 0);
assert(huff_codes != NULL);
/* Calculate max code length. */
for (symbol = 0; symbol < code_lengths_size; ++symbol) {
if (code_lengths[symbol] > max_code_length) {
max_code_length = code_lengths[symbol];
}
/* Returns the table width of the next 2nd level table. count is the histogram
of bit lengths for the remaining symbols, len is the code length of the next
processed symbol */
static BROTLI_INLINE int NextTableBitSize(const int* const count,
int len, int root_bits) {
int left = 1 << (len - root_bits);
while (len < MAX_LENGTH) {
left -= count[len];
if (left <= 0) break;
++len;
left <<= 1;
}
if (max_code_length > MAX_ALLOWED_CODE_LENGTH) return 0;
return len - root_bits;
}
/* Calculate code length histogram. */
for (symbol = 0; symbol < code_lengths_size; ++symbol) {
++code_length_hist[code_lengths[symbol]];
}
code_length_hist[0] = 0;
int BrotliBuildHuffmanTable(HuffmanCode* root_table,
int root_bits,
const uint8_t* const code_lengths,
int code_lengths_size) {
HuffmanCode code; /* current table entry */
HuffmanCode* table; /* next available space in table */
int len; /* current code length */
int symbol; /* symbol index in original or sorted table */
int key; /* reversed prefix code */
int step; /* step size to replicate values in current table */
int low; /* low bits for current root entry */
int mask; /* mask for low bits */
int table_bits; /* key length of current table */
int table_size; /* size of current table */
int total_size; /* sum of root table size and 2nd level table sizes */
int* sorted; /* symbols sorted by code length */
int count[MAX_LENGTH + 1] = { 0 }; /* number of codes of each length */
int offset[MAX_LENGTH + 1]; /* offsets in sorted table for each length */
/* Calculate the initial values of 'next_codes' for each code length. */
/* next_codes[code_len] denotes the code to be assigned to the next symbol */
/* of code length 'code_len'. */
curr_code = 0;
next_codes[0] = -1; /* Unused, as code length = 0 implies */
/* code doesn't exist. */
for (code_len = 1; code_len <= max_code_length; ++code_len) {
curr_code = (curr_code + code_length_hist[code_len - 1]) << 1;
next_codes[code_len] = curr_code;
sorted = (int*)malloc((size_t)code_lengths_size * sizeof(*sorted));
if (sorted == NULL) {
return 0;
}
/* Get symbols. */
for (symbol = 0; symbol < code_lengths_size; ++symbol) {
if (code_lengths[symbol] > 0) {
huff_codes[symbol] = next_codes[code_lengths[symbol]]++;
} else {
huff_codes[symbol] = NON_EXISTENT_SYMBOL;
}
/* build histogram of code lengths */
for (symbol = 0; symbol < code_lengths_size; symbol++) {
count[code_lengths[symbol]]++;
}
return 1;
}
static const uint8_t kReverse7[128] = {
0, 64, 32, 96, 16, 80, 48, 112, 8, 72, 40, 104, 24, 88, 56, 120,
4, 68, 36, 100, 20, 84, 52, 116, 12, 76, 44, 108, 28, 92, 60, 124,
2, 66, 34, 98, 18, 82, 50, 114, 10, 74, 42, 106, 26, 90, 58, 122,
6, 70, 38, 102, 22, 86, 54, 118, 14, 78, 46, 110, 30, 94, 62, 126,
1, 65, 33, 97, 17, 81, 49, 113, 9, 73, 41, 105, 25, 89, 57, 121,
5, 69, 37, 101, 21, 85, 53, 117, 13, 77, 45, 109, 29, 93, 61, 125,
3, 67, 35, 99, 19, 83, 51, 115, 11, 75, 43, 107, 27, 91, 59, 123,
7, 71, 39, 103, 23, 87, 55, 119, 15, 79, 47, 111, 31, 95, 63, 127
};
static int ReverseBitsShort(int bits, int num_bits) {
return kReverse7[bits] >> (7 - num_bits);
}
static int TreeAddSymbol(HuffmanTree* const tree,
int symbol, int code, int code_length) {
int step = HUFF_LUT_BITS;
int base_code;
HuffmanTreeNode* node = tree->root_;
const HuffmanTreeNode* const max_node = tree->root_ + tree->max_nodes_;
assert(symbol == (int16_t)symbol);
if (code_length <= HUFF_LUT_BITS) {
int i = 1 << (HUFF_LUT_BITS - code_length);
base_code = ReverseBitsShort(code, code_length);
do {
int idx;
--i;
idx = base_code | (i << code_length);
tree->lut_symbol_[idx] = (int16_t)symbol;
tree->lut_bits_[idx] = (uint8_t)code_length;
} while (i > 0);
} else {
base_code = ReverseBitsShort((code >> (code_length - HUFF_LUT_BITS)),
HUFF_LUT_BITS);
/* generate offsets into sorted symbol table by code length */
offset[1] = 0;
for (len = 1; len < MAX_LENGTH; len++) {
offset[len + 1] = offset[len] + count[len];
}
while (code_length-- > 0) {
if (node >= max_node) {
return 0;
}
if (NodeIsEmpty(node)) {
if (IsFull(tree)) return 0; /* error: too many symbols. */
AssignChildren(tree, node);
} else if (!HuffmanTreeNodeIsNotLeaf(node)) {
return 0; /* leaf is already occupied. */
}
node += node->children_ + ((code >> code_length) & 1);
if (--step == 0) {
tree->lut_jump_[base_code] = (int16_t)(node - tree->root_);
}
}
if (NodeIsEmpty(node)) {
node->children_ = 0; /* turn newly created node into a leaf. */
} else if (HuffmanTreeNodeIsNotLeaf(node)) {
return 0; /* trying to assign a symbol to already used code. */
}
node->symbol_ = symbol; /* Add symbol in this node. */
return 1;
}
int BrotliHuffmanTreeBuildImplicit(HuffmanTree* const tree,
const uint8_t* const code_lengths,
int code_lengths_size) {
int symbol;
int num_symbols = 0;
int root_symbol = 0;
assert(tree != NULL);
assert(code_lengths != NULL);
/* Find out number of symbols and the root symbol. */
for (symbol = 0; symbol < code_lengths_size; ++symbol) {
if (code_lengths[symbol] > 0) {
/* Note: code length = 0 indicates non-existent symbol. */
++num_symbols;
root_symbol = symbol;
/* sort symbols by length, by symbol order within each length */
for (symbol = 0; symbol < code_lengths_size; symbol++) {
if (code_lengths[symbol] != 0) {
sorted[offset[code_lengths[symbol]]++] = symbol;
}
}
/* Initialize the tree. Will fail for num_symbols = 0 */
if (!TreeInit(tree, num_symbols)) return 0;
table = root_table;
table_bits = root_bits;
table_size = 1 << table_bits;
total_size = table_size;
/* Build tree. */
if (num_symbols == 1) { /* Trivial case. */
const int max_symbol = code_lengths_size;
if (root_symbol < 0 || root_symbol >= max_symbol) {
BrotliHuffmanTreeRelease(tree);
return 0;
/* special case code with only one value */
if (offset[MAX_LENGTH] == 1) {
code.bits = 0;
code.value = (uint16_t)sorted[0];
for (key = 0; key < total_size; ++key) {
table[key] = code;
}
return TreeAddSymbol(tree, root_symbol, 0, 0);
} else { /* Normal case. */
int ok = 0;
free(sorted);
return total_size;
}
/* Get Huffman codes from the code lengths. */
int* const codes =
(int*)BrotliSafeMalloc((uint64_t)code_lengths_size, sizeof(*codes));
if (codes == NULL) goto End;
if (!HuffmanCodeLengthsToCodes(code_lengths, code_lengths_size, codes)) {
goto End;
/* fill in root table */
key = 0;
symbol = 0;
for (len = 1, step = 2; len <= root_bits; ++len, step <<= 1) {
for (; count[len] > 0; --count[len]) {
code.bits = (uint8_t)(len);
code.value = (uint16_t)sorted[symbol++];
ReplicateValue(&table[key], step, table_size, code);
key = GetNextKey(key, len);
}
}
/* Add symbols one-by-one. */
for (symbol = 0; symbol < code_lengths_size; ++symbol) {
if (code_lengths[symbol] > 0) {
if (!TreeAddSymbol(tree, symbol, codes[symbol], code_lengths[symbol])) {
goto End;
}
/* fill in 2nd level tables and add pointers to root table */
mask = total_size - 1;
low = -1;
for (len = root_bits + 1, step = 2; len <= MAX_LENGTH; ++len, step <<= 1) {
for (; count[len] > 0; --count[len]) {
if ((key & mask) != low) {
table += table_size;
table_bits = NextTableBitSize(count, len, root_bits);
table_size = 1 << table_bits;
total_size += table_size;
low = key & mask;
root_table[low].bits = (uint8_t)(table_bits + root_bits);
root_table[low].value = (uint16_t)((table - root_table) - low);
}
code.bits = (uint8_t)(len - root_bits);
code.value = (uint16_t)sorted[symbol++];
ReplicateValue(&table[key >> root_bits], step, table_size, code);
key = GetNextKey(key, len);
}
ok = 1;
End:
free(codes);
ok = ok && IsFull(tree);
if (!ok) BrotliHuffmanTreeRelease(tree);
return ok;
}
free(sorted);
return total_size;
}
#if defined(__cplusplus) || defined(c_plusplus)