From eabdabb9822ed028cd8aeed94ded784b6364ba78 Mon Sep 17 00:00:00 2001 From: Umair Khan Date: Fri, 12 Aug 2016 17:49:15 +0530 Subject: [PATCH] avcodec: Implement masked lz decompression Signed-off-by: Umair Khan --- libavcodec/mlz.c | 173 +++++++++++++++++++++++++++++++++++++++++++++++ libavcodec/mlz.h | 70 +++++++++++++++++++ 2 files changed, 243 insertions(+) create mode 100644 libavcodec/mlz.c create mode 100644 libavcodec/mlz.h diff --git a/libavcodec/mlz.c b/libavcodec/mlz.c new file mode 100644 index 0000000000..04b684d4f5 --- /dev/null +++ b/libavcodec/mlz.c @@ -0,0 +1,173 @@ +/* + * Copyright (c) 2016 Umair Khan + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "mlz.h" + +av_cold void ff_mlz_init_dict(void* context, MLZ *mlz) { + mlz->dict = av_malloc_array(TABLE_SIZE, sizeof(*mlz->dict)); + + mlz->flush_code = FLUSH_CODE; + mlz->current_dic_index_max = DIC_INDEX_INIT; + mlz->dic_code_bit = CODE_BIT_INIT; + mlz->bump_code = (DIC_INDEX_INIT - 1); + mlz->next_code = FIRST_CODE; + mlz->freeze_flag = 0; + mlz->context = context; +} + +av_cold void ff_mlz_flush_dict(MLZ *mlz) { + MLZDict *dict = mlz->dict; + int i; + for ( i = 0; i < TABLE_SIZE; i++ ) { + dict[i].string_code = CODE_UNSET; + dict[i].parent_code = CODE_UNSET; + dict[i].match_len = 0; + } + mlz->current_dic_index_max = DIC_INDEX_INIT; + mlz->dic_code_bit = CODE_BIT_INIT; // DicCodeBitInit; + mlz->bump_code = mlz->current_dic_index_max - 1; + mlz->next_code = FIRST_CODE; + mlz->freeze_flag = 0; +} + +static void set_new_entry_dict(MLZDict* dict, int string_code, int parent_code, int char_code) { + dict[string_code].parent_code = parent_code; + dict[string_code].string_code = string_code; + dict[string_code].char_code = char_code; + if (parent_code < FIRST_CODE) { + dict[string_code].match_len = 2; + } else { + dict[string_code].match_len = (dict[parent_code].match_len) + 1; + } +} + +static int decode_string(MLZ* mlz, unsigned char *buff, int string_code, int *first_char_code, unsigned long bufsize) { + MLZDict* dict = mlz->dict; + unsigned long count, offset; + int current_code, parent_code, tmp_code; + + count = 0; + current_code = string_code; + *first_char_code = CODE_UNSET; + + while (count < bufsize) { + switch (current_code) { + case CODE_UNSET: + return count; + break; + default: + if (current_code < FIRST_CODE) { + *first_char_code = current_code; + buff[0] = current_code; + count++; + return count; + } else { + offset = dict[current_code].match_len - 1; + tmp_code = dict[current_code].char_code; + buff[offset] = tmp_code; + count++; + } + current_code = dict[current_code].parent_code; + if ((current_code < 0) || (current_code > (DIC_INDEX_MAX - 1))) { + av_log(mlz->context, AV_LOG_ERROR, "MLZ dic index error.\n"); + return count; + } + if (current_code > FIRST_CODE) { + parent_code = dict[current_code].parent_code; + offset = (dict[current_code].match_len) - 1; + if (parent_code < 0 || parent_code > DIC_INDEX_MAX-1) { + av_log(mlz->context, AV_LOG_ERROR, "MLZ dic index error.\n"); + return count; + } + if (( offset > (DIC_INDEX_MAX - 1))) { + av_log(mlz->context, AV_LOG_ERROR, "MLZ dic offset error.\n"); + return count; + } + } + break; + } + } + return count; +} + +static int input_code(GetBitContext* gb, int len) { + int tmp_code = 0; + int i; + for (i = 0; i < len; ++i) { + tmp_code |= get_bits1(gb) << i; + } + return tmp_code; +} + +int ff_mlz_decompression(MLZ* mlz, GetBitContext* gb, int size, unsigned char *buff) { + MLZDict *dict = mlz->dict; + unsigned long output_chars; + int string_code, last_string_code, char_code; + + string_code = 0; + char_code = -1; + last_string_code = -1; + output_chars = 0; + + while (output_chars < size) { + string_code = input_code(gb, mlz->dic_code_bit); + switch (string_code) { + case FLUSH_CODE: + case MAX_CODE: + ff_mlz_flush_dict(mlz); + char_code = -1; + last_string_code = -1; + break; + case FREEZE_CODE: + mlz->freeze_flag = 1; + break; + default: + if (string_code > mlz->current_dic_index_max) { + av_log(mlz->context, AV_LOG_ERROR, "String code %d exceeds maximum value of %d.\n", string_code, mlz->current_dic_index_max); + return output_chars; + } + if (string_code == (int) mlz->bump_code) { + ++mlz->dic_code_bit; + mlz->current_dic_index_max *= 2; + mlz->bump_code = mlz->current_dic_index_max - 1; + } else { + if (string_code >= mlz->next_code) { + output_chars += decode_string(mlz, &buff[output_chars], last_string_code, &char_code, size - output_chars); + output_chars += decode_string(mlz, &buff[output_chars], char_code, &char_code, size - output_chars); + set_new_entry_dict(dict, mlz->next_code, last_string_code, char_code); + mlz->next_code++; + } else { + output_chars += decode_string(mlz, &buff[output_chars], string_code, &char_code, size - output_chars); + if (output_chars <= size && !mlz->freeze_flag) { + if (last_string_code != -1) { + set_new_entry_dict(dict, mlz->next_code, last_string_code, char_code); + mlz->next_code++; + } + } else { + break; + } + } + last_string_code = string_code; + } + break; + } + } + return output_chars; +} diff --git a/libavcodec/mlz.h b/libavcodec/mlz.h new file mode 100644 index 0000000000..158715e7ec --- /dev/null +++ b/libavcodec/mlz.h @@ -0,0 +1,70 @@ +/* + * Copyright (c) 2016 Umair Khan + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#ifndef AVCODEC_MLZ_H +#define AVCODEC_MLZ_H + +#include "get_bits.h" + +#define CODE_UNSET -1 +#define CODE_BIT_INIT 9 +#define DIC_INDEX_INIT 512 // 2^9 +#define DIC_INDEX_MAX 32768l // 2^15 +#define FLUSH_CODE 256 +#define FREEZE_CODE 257 +#define FIRST_CODE 258 +#define MAX_CODE 32767l +#define TABLE_SIZE 35023l // TABLE_SIZE must be a prime number + +/** Dictionary structure for mlz decompression + */ +typedef struct MLZDict { + int string_code; + int parent_code; + int char_code; + int match_len; +} MLZDict; + +/** MLZ data strucure + */ +typedef struct MLZ { + int dic_code_bit; + int current_dic_index_max; + unsigned int bump_code; + unsigned int flush_code; + int next_code; + int freeze_flag; + MLZDict* dict; + void* context; +} MLZ; + +/** Initialize the dictionary + */ +void ff_mlz_init_dict(void* context, MLZ *mlz); + +/** Flush the dictionary + */ +void ff_mlz_flush_dict(MLZ *dict); + +/** Run mlz decompression on the next size bits and the output will be stored in buff + */ +int ff_mlz_decompression(MLZ* mlz, GetBitContext* gb, int size, unsigned char *buff); + +#endif /*AVCODEC_MLZ_H*/