diff options
Diffstat (limited to 'decoder.c')
| -rw-r--r-- | decoder.c | 108 |
1 files changed, 65 insertions, 43 deletions
@@ -1,5 +1,5 @@ /* Lunzip - Decompressor for the lzip format - Copyright (C) 2010-2016 Antonio Diaz Diaz. + Copyright (C) 2010-2017 Antonio Diaz Diaz. This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by @@ -53,7 +53,7 @@ void Pp_show_msg( struct Pretty_print * const pp, const char * const msg ) /* Returns the number of bytes really read. If (returned value < size) and (errno == 0), means EOF was reached. */ -static int readblock( const int fd, uint8_t * const buf, const int size ) +int readblock( const int fd, uint8_t * const buf, const int size ) { int sz = 0; errno = 0; @@ -87,10 +87,10 @@ static int writeblock( const int fd, const uint8_t * const buf, const int size ) } -int seek_read( const int fd, uint8_t * const buf, const int size, - const int offset ) +unsigned seek_read_back( const int fd, uint8_t * const buf, const int size, + const int offset ) { - if( lseek( fd, offset, SEEK_END ) >= 0 ) + if( lseek( fd, -offset, SEEK_END ) >= 0 ) return readblock( fd, buf, size ); return 0; } @@ -121,8 +121,9 @@ void LZd_flush_data( struct LZ_decoder * const d ) writeblock( d->outfd, d->buffer + d->stream_pos, size ) != size ) { show_error( "Write error", errno, false ); cleanup_and_fail( 1 ); } if( d->pos >= d->buffer_size ) - { d->partial_data_pos += d->pos; d->pos = 0; - if( d->partial_data_pos >= d->dictionary_size ) d->pos_wrapped = true; } + { d->partial_data_pos += d->pos; d->pos = 0; d->pos_wrapped = true; + if( d->partial_data_pos >= d->dictionary_size ) + d->pos_wrapped_dic = true; } d->stream_pos = d->pos; } } @@ -185,7 +186,7 @@ static bool LZd_verify_trailer( struct LZ_decoder * const d, ( 8.0 * member_size ) / data_size, 100.0 * ( 1.0 - ( (double)member_size / data_size ) ) ); if( !error && verbosity >= 4 ) - fprintf( stderr, "data CRC %08X, data size %9llu, member size %8llu. ", + fprintf( stderr, "CRC %08X, decompressed %9llu, compressed %8llu. ", LZd_crc( d ), data_size, member_size ); return !error; } @@ -198,49 +199,77 @@ int LZd_decode_member( struct LZ_decoder * const d, { struct Range_decoder * const rdec = d->rdec; void (* const copy_block) - ( struct LZ_decoder * const d, const int distance, int len ) = - ( (unsigned)d->buffer_size >= d->dictionary_size ) ? + ( struct LZ_decoder * const d, const unsigned distance, unsigned len ) = + ( d->buffer_size >= d->dictionary_size ) ? &LZd_copy_block : &LZd_copy_block2; + Bit_model bm_literal[1<<literal_context_bits][0x300]; + Bit_model bm_match[states][pos_states]; + Bit_model bm_rep[states]; + Bit_model bm_rep0[states]; + Bit_model bm_rep1[states]; + Bit_model bm_rep2[states]; + Bit_model bm_len[states][pos_states]; + Bit_model bm_dis_slot[len_states][1<<dis_slot_bits]; + Bit_model bm_dis[modeled_distances-end_dis_model+1]; + Bit_model bm_align[dis_align_size]; + struct Len_model match_len_model; + struct Len_model rep_len_model; unsigned rep0 = 0; /* rep[0-3] latest four distances */ unsigned rep1 = 0; /* used for efficient coding of */ unsigned rep2 = 0; /* repeated distances */ unsigned rep3 = 0; State state = 0; + Bm_array_init( bm_literal[0], (1 << literal_context_bits) * 0x300 ); + Bm_array_init( bm_match[0], states * pos_states ); + Bm_array_init( bm_rep, states ); + Bm_array_init( bm_rep0, states ); + Bm_array_init( bm_rep1, states ); + Bm_array_init( bm_rep2, states ); + Bm_array_init( bm_len[0], states * pos_states ); + Bm_array_init( bm_dis_slot[0], len_states * (1 << dis_slot_bits) ); + Bm_array_init( bm_dis, modeled_distances - end_dis_model + 1 ); + Bm_array_init( bm_align, dis_align_size ); + Lm_init( &match_len_model ); + Lm_init( &rep_len_model ); + Rd_load( rdec ); while( !Rd_finished( rdec ) ) { const int pos_state = LZd_data_position( d ) & pos_state_mask; - if( Rd_decode_bit( rdec, &d->bm_match[state][pos_state] ) == 0 ) /* 1st bit */ + if( Rd_decode_bit( rdec, &bm_match[state][pos_state] ) == 0 ) /* 1st bit */ { - const uint8_t prev_byte = LZd_peek_prev( d ); + Bit_model * const bm = bm_literal[get_lit_state(LZd_peek_prev( d ))]; if( St_is_char( state ) ) { state -= ( state < 4 ) ? state : 3; - LZd_put_byte( d, Rd_decode_tree( rdec, - d->bm_literal[get_lit_state(prev_byte)], 8 ) ); + LZd_put_byte( d, Rd_decode_tree8( rdec, bm ) ); } else { state -= ( state < 10 ) ? 3 : 6; - LZd_put_byte( d, Rd_decode_matched( rdec, - d->bm_literal[get_lit_state(prev_byte)], - LZd_peek( d, rep0 ) ) ); + LZd_put_byte( d, Rd_decode_matched( rdec, bm, LZd_peek( d, rep0 ) ) ); } } else /* match or repeated match */ { int len; - if( Rd_decode_bit( rdec, &d->bm_rep[state] ) != 0 ) /* 2nd bit */ + if( Rd_decode_bit( rdec, &bm_rep[state] ) != 0 ) /* 2nd bit */ { - if( Rd_decode_bit( rdec, &d->bm_rep0[state] ) != 0 ) /* 3rd bit */ + if( Rd_decode_bit( rdec, &bm_rep0[state] ) == 0 ) /* 3rd bit */ + { + if( Rd_decode_bit( rdec, &bm_len[state][pos_state] ) == 0 ) /* 4th bit */ + { state = St_set_short_rep( state ); + LZd_put_byte( d, LZd_peek( d, rep0 ) ); continue; } + } + else { unsigned distance; - if( Rd_decode_bit( rdec, &d->bm_rep1[state] ) == 0 ) /* 4th bit */ + if( Rd_decode_bit( rdec, &bm_rep1[state] ) == 0 ) /* 4th bit */ distance = rep1; else { - if( Rd_decode_bit( rdec, &d->bm_rep2[state] ) == 0 ) /* 5th bit */ + if( Rd_decode_bit( rdec, &bm_rep2[state] ) == 0 ) /* 5th bit */ distance = rep2; else { distance = rep3; rep3 = rep2; } @@ -249,36 +278,29 @@ int LZd_decode_member( struct LZ_decoder * const d, rep1 = rep0; rep0 = distance; } - else - { - if( Rd_decode_bit( rdec, &d->bm_len[state][pos_state] ) == 0 ) /* 4th bit */ - { state = St_set_short_rep( state ); - LZd_put_byte( d, LZd_peek( d, rep0 ) ); continue; } - } state = St_set_rep( state ); - len = min_match_len + Rd_decode_len( rdec, &d->rep_len_model, pos_state ); + len = min_match_len + Rd_decode_len( rdec, &rep_len_model, pos_state ); } else /* match */ { - int dis_slot; - const unsigned rep0_saved = rep0; - len = min_match_len + Rd_decode_len( rdec, &d->match_len_model, pos_state ); - dis_slot = Rd_decode_tree6( rdec, d->bm_dis_slot[get_len_state(len)] ); - if( dis_slot < start_dis_model ) rep0 = dis_slot; - else + unsigned distance; + len = min_match_len + Rd_decode_len( rdec, &match_len_model, pos_state ); + distance = Rd_decode_tree6( rdec, bm_dis_slot[get_len_state(len)] ); + if( distance >= start_dis_model ) { + const unsigned dis_slot = distance; const int direct_bits = ( dis_slot >> 1 ) - 1; - rep0 = ( 2 | ( dis_slot & 1 ) ) << direct_bits; + distance = ( 2 | ( dis_slot & 1 ) ) << direct_bits; if( dis_slot < end_dis_model ) - rep0 += Rd_decode_tree_reversed( rdec, - d->bm_dis + rep0 - dis_slot - 1, direct_bits ); + distance += Rd_decode_tree_reversed( rdec, + bm_dis + ( distance - dis_slot ), direct_bits ); else { - rep0 += Rd_decode( rdec, direct_bits - dis_align_bits ) << dis_align_bits; - rep0 += Rd_decode_tree_reversed4( rdec, d->bm_align ); - if( rep0 == 0xFFFFFFFFU ) /* marker found */ + distance += + Rd_decode( rdec, direct_bits - dis_align_bits ) << dis_align_bits; + distance += Rd_decode_tree_reversed4( rdec, bm_align ); + if( distance == 0xFFFFFFFFU ) /* marker found */ { - rep0 = rep0_saved; Rd_normalize( rdec ); LZd_flush_data( d ); if( len == min_match_len ) /* End Of Stream marker */ @@ -298,10 +320,10 @@ int LZd_decode_member( struct LZ_decoder * const d, } } } - rep3 = rep2; rep2 = rep1; rep1 = rep0_saved; + rep3 = rep2; rep2 = rep1; rep1 = rep0; rep0 = distance; state = St_set_match( state ); if( rep0 >= d->dictionary_size || - ( rep0 >= LZd_data_position( d ) && !d->pos_wrapped ) ) + ( rep0 >= LZd_data_position( d ) && !d->pos_wrapped_dic ) ) { LZd_flush_data( d ); return 1; } } copy_block( d, rep0, len ); |