Adding upstream version 1.9.upstream/1.9

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 112 insertions, 133 deletions

diff --git a/decoder.h b/decoder.h
index 922426c..e3a4f1c 100644
--- a/decoder.h
+++ b/decoder.h
@@ -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
@@ -56,7 +56,7 @@ Rd_member_position( const struct Range_decoder * const rdec )
   { return rdec->partial_member_pos + rdec->pos; }
 
 static inline void Rd_reset_member_position( struct Range_decoder * const rdec )
-  { rdec->partial_member_pos = -rdec->pos; }
+  { rdec->partial_member_pos = 0; rdec->partial_member_pos -= rdec->pos; }
 
 static inline uint8_t Rd_get_byte( struct Range_decoder * const rdec )
   {
@@ -68,23 +68,22 @@ static inline uint8_t Rd_get_byte( struct Range_decoder * const rdec )
 static inline int Rd_read_data( struct Range_decoder * const rdec,
                                 uint8_t * const outbuf, const int size )
   {
-  int rest = size;
-  while( rest > 0 && !Rd_finished( rdec ) )
+  int sz = 0;
+  while( sz < size && !Rd_finished( rdec ) )
     {
-    const int rd = min( rest, rdec->stream_pos - rdec->pos );
-    memcpy( outbuf + size - rest, rdec->buffer + rdec->pos, rd );
+    const int rd = min( size - sz, rdec->stream_pos - rdec->pos );
+    memcpy( outbuf + sz, rdec->buffer + rdec->pos, rd );
     rdec->pos += rd;
-    rest -= rd;
+    sz += rd;
     }
-  return size - rest;
+  return sz;
   }
 
 static inline void Rd_load( struct Range_decoder * const rdec )
   {
   int i;
   rdec->code = 0;
-  for( i = 0; i < 5; ++i )
-    rdec->code = (rdec->code << 8) | Rd_get_byte( rdec );
+  for( i = 0; i < 5; ++i ) rdec->code = (rdec->code << 8) | Rd_get_byte( rdec );
   rdec->range = 0xFFFFFFFFU;
   rdec->code &= rdec->range;	/* make sure that first byte is discarded */
   }
@@ -92,34 +91,30 @@ static inline void Rd_load( struct Range_decoder * const rdec )
 static inline void Rd_normalize( struct Range_decoder * const rdec )
   {
   if( rdec->range <= 0x00FFFFFFU )
-    {
-    rdec->range <<= 8;
-    rdec->code = (rdec->code << 8) | Rd_get_byte( rdec );
-    }
+    { rdec->range <<= 8; rdec->code = (rdec->code << 8) | Rd_get_byte( rdec ); }
   }
 
-static inline int Rd_decode( struct Range_decoder * const rdec,
-                             const int num_bits )
+static inline unsigned Rd_decode( struct Range_decoder * const rdec,
+                                  const int num_bits )
   {
-  int symbol = 0;
+  unsigned symbol = 0;
   int i;
   for( i = num_bits; i > 0; --i )
     {
-    uint32_t mask;
+    bool bit;
     Rd_normalize( rdec );
     rdec->range >>= 1;
 /*    symbol <<= 1; */
 /*    if( rdec->code >= rdec->range ) { rdec->code -= rdec->range; symbol |= 1; } */
-    mask = 0U - (rdec->code < rdec->range);
-    rdec->code -= rdec->range;
-    rdec->code += rdec->range & mask;
-    symbol = (symbol << 1) + (mask + 1);
+    bit = ( rdec->code >= rdec->range );
+    symbol = ( symbol << 1 ) + bit;
+    rdec->code -= rdec->range & ( 0U - bit );
     }
   return symbol;
   }
 
-static inline int Rd_decode_bit( struct Range_decoder * const rdec,
-                                 Bit_model * const probability )
+static inline unsigned Rd_decode_bit( struct Range_decoder * const rdec,
+                                      Bit_model * const probability )
   {
   uint32_t bound;
   Rd_normalize( rdec );
@@ -139,20 +134,20 @@ static inline int Rd_decode_bit( struct Range_decoder * const rdec,
     }
   }
 
-static inline int Rd_decode_tree( struct Range_decoder * const rdec,
-                                  Bit_model bm[], const int num_bits )
+static inline unsigned Rd_decode_tree3( struct Range_decoder * const rdec,
+                                        Bit_model bm[] )
   {
-  int symbol = 1;
-  int i;
-  for( i = num_bits; i > 0; --i )
-    symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
-  return symbol - (1 << num_bits);
+  unsigned symbol = 1;
+  symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
+  symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
+  symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
+  return symbol & 7;
   }
 
-static inline int Rd_decode_tree6( struct Range_decoder * const rdec,
-                                   Bit_model bm[] )
+static inline unsigned Rd_decode_tree6( struct Range_decoder * const rdec,
+                                        Bit_model bm[] )
   {
-  int symbol = 1;
+  unsigned symbol = 1;
   symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
   symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
   symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
@@ -162,69 +157,69 @@ static inline int Rd_decode_tree6( struct Range_decoder * const rdec,
   return symbol & 0x3F;
   }
 
-static inline int Rd_decode_tree_reversed( struct Range_decoder * const rdec,
-                                           Bit_model bm[], const int num_bits )
+static inline unsigned Rd_decode_tree8( struct Range_decoder * const rdec,
+                                        Bit_model bm[] )
   {
-  int model = 1;
-  int symbol = 0;
+  unsigned symbol = 1;
+  int i;
+  for( i = 0; i < 8; ++i )
+    symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
+  return symbol & 0xFF;
+  }
+
+static inline unsigned
+Rd_decode_tree_reversed( struct Range_decoder * const rdec,
+                         Bit_model bm[], const int num_bits )
+  {
+  unsigned model = 1;
+  unsigned symbol = 0;
   int i;
   for( i = 0; i < num_bits; ++i )
     {
-    const bool bit = Rd_decode_bit( rdec, &bm[model] );
-    model <<= 1;
-    if( bit ) { ++model; symbol |= (1 << i); }
+    const unsigned bit = Rd_decode_bit( rdec, &bm[model] );
+    model = ( model << 1 ) + bit;
+    symbol |= ( bit << i );
     }
   return symbol;
   }
 
-static inline int Rd_decode_tree_reversed4( struct Range_decoder * const rdec,
-                                            Bit_model bm[] )
+static inline unsigned
+Rd_decode_tree_reversed4( struct Range_decoder * const rdec, Bit_model bm[] )
   {
-  int model = 1;
-  int symbol = Rd_decode_bit( rdec, &bm[model] );
-  int bit;
-  model = (model << 1) + symbol;
-  bit = Rd_decode_bit( rdec, &bm[model] );
-  model = (model << 1) + bit; symbol |= (bit << 1);
+  unsigned symbol = Rd_decode_bit( rdec, &bm[1] );
+  unsigned model = 2 + symbol;
+  unsigned bit = Rd_decode_bit( rdec, &bm[model] );
+  model = ( model << 1 ) + bit; symbol |= ( bit << 1 );
   bit = Rd_decode_bit( rdec, &bm[model] );
-  model = (model << 1) + bit; symbol |= (bit << 2);
-  if( Rd_decode_bit( rdec, &bm[model] ) ) symbol |= 8;
+  model = ( model << 1 ) + bit; symbol |= ( bit << 2 );
+  symbol |= ( Rd_decode_bit( rdec, &bm[model] ) << 3 );
   return symbol;
   }
 
-static inline int Rd_decode_matched( struct Range_decoder * const rdec,
-                                     Bit_model bm[], int match_byte )
+static inline unsigned Rd_decode_matched( struct Range_decoder * const rdec,
+                                          Bit_model bm[], unsigned match_byte )
   {
-  Bit_model * const bm1 = bm + 0x100;
-  int symbol = 1;
-  while( symbol < 0x100 )
+  unsigned symbol = 1;
+  unsigned mask = 0x100;
+  while( true )
     {
-    int match_bit, bit;
-    match_byte <<= 1;
-    match_bit = match_byte & 0x100;
-    bit = Rd_decode_bit( rdec, &bm1[match_bit+symbol] );
-    symbol = ( symbol << 1 ) | bit;
-    if( match_bit != bit << 8 )
-      {
-      while( symbol < 0x100 )
-        symbol = ( symbol << 1 ) | Rd_decode_bit( rdec, &bm[symbol] );
-      break;
-      }
+    const unsigned match_bit = ( match_byte <<= 1 ) & mask;
+    const unsigned bit = Rd_decode_bit( rdec, &bm[symbol+match_bit+mask] );
+    symbol = ( symbol << 1 ) + bit;
+    if( symbol > 0xFF ) return symbol & 0xFF;
+    mask &= ~(match_bit ^ (bit << 8));	/* if( match_bit != bit ) mask = 0; */
     }
-  return symbol & 0xFF;
   }
 
-static inline int Rd_decode_len( struct Range_decoder * const rdec,
-                                 struct Len_model * const lm,
-                                 const int pos_state )
+static inline unsigned Rd_decode_len( struct Range_decoder * const rdec,
+                                      struct Len_model * const lm,
+                                      const int pos_state )
   {
   if( Rd_decode_bit( rdec, &lm->choice1 ) == 0 )
-    return Rd_decode_tree( rdec, lm->bm_low[pos_state], len_low_bits );
+    return Rd_decode_tree3( rdec, lm->bm_low[pos_state] );
   if( Rd_decode_bit( rdec, &lm->choice2 ) == 0 )
-    return len_low_symbols +
-           Rd_decode_tree( rdec, lm->bm_mid[pos_state], len_mid_bits );
-  return len_low_symbols + len_mid_symbols +
-         Rd_decode_tree( rdec, lm->bm_high, len_high_bits );
+    return len_low_symbols + Rd_decode_tree3( rdec, lm->bm_mid[pos_state] );
+  return len_low_symbols + len_mid_symbols + Rd_decode_tree8( rdec, lm->bm_high );
   }
 
 
@@ -233,49 +228,37 @@ struct LZ_decoder
   unsigned long long partial_data_pos;
   struct Range_decoder * rdec;
   unsigned dictionary_size;
-  int buffer_size;
+  unsigned buffer_size;
   uint8_t * buffer;		/* output buffer */
-  int pos;			/* current pos in buffer */
-  int stream_pos;		/* first byte not yet written to file */
+  unsigned pos;			/* current pos in buffer */
+  unsigned stream_pos;		/* first byte not yet written to file */
   uint32_t crc;
   int outfd;			/* output file descriptor */
   bool pos_wrapped;
-
-  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];
-  Bit_model bm_align[dis_align_size];
-
-  struct Len_model match_len_model;
-  struct Len_model rep_len_model;
+  bool pos_wrapped_dic;
   };
 
 void LZd_flush_data( struct LZ_decoder * const d );
 
-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 );
 
 static inline uint8_t LZd_peek_prev( const struct LZ_decoder * const d )
   {
-  const int i = ( ( d->pos > 0 ) ? d->pos : d->buffer_size ) - 1;
-  return d->buffer[i];
+  if( d->pos > 0 ) return d->buffer[d->pos-1];
+  if( d->pos_wrapped ) return d->buffer[d->buffer_size-1];
+  return 0;			/* prev_byte of first byte */
   }
 
 static inline uint8_t LZd_peek( const struct LZ_decoder * const d,
-                                const int distance )
+                                const unsigned distance )
   {
   uint8_t b;
-  const int i = d->pos - distance - 1;
-  if( i >= 0 ) b = d->buffer[i];
-  else if( i + d->buffer_size >= d->pos )
-    b = d->buffer[i+d->buffer_size];
-  else if( seek_read( d->outfd, &b, 1, i - d->stream_pos ) != 1 )
+  if( d->pos > distance ) b = d->buffer[d->pos-distance-1];
+  else if( d->buffer_size > distance )
+    b = d->buffer[d->buffer_size+d->pos-distance-1];
+  else if( seek_read_back( d->outfd, &b, 1,
+                           distance + 1 + d->stream_pos - d->pos ) != 1 )
     { show_error( "Seek error", errno, false ); cleanup_and_fail( 1 ); }
   return b;
   }
@@ -287,19 +270,28 @@ static inline void LZd_put_byte( struct LZ_decoder * const d, const uint8_t b )
   }
 
 static inline void LZd_copy_block( struct LZ_decoder * const d,
-                                   const int distance, int len )
+                                   const unsigned distance, unsigned len )
   {
-  int i = d->pos - distance - 1;
-  bool fast;
-  if( i < 0 )
-    { i += d->buffer_size;
-      fast = ( len <= d->buffer_size - i && len <= i - d->pos ); }
+  unsigned lpos = d->pos, i = lpos - distance - 1;
+  bool fast, fast2;
+  if( lpos > distance )
+    {
+    fast = ( len < d->buffer_size - lpos );
+    fast2 = ( fast && len <= lpos - i );
+    }
   else
-    fast = ( len < d->buffer_size - d->pos && len <= d->pos - i );
-  if( fast )					/* no wrap, no overlap */
     {
-    memcpy( d->buffer + d->pos, d->buffer + i, len );
+    i += d->buffer_size;
+    fast = ( len < d->buffer_size - i );	/* (i == pos) may happen */
+    fast2 = ( fast && len <= i - lpos );
+    }
+  if( fast )					/* no wrap */
+    {
     d->pos += len;
+    if( fast2 )					/* no wrap, no overlap */
+      memcpy( d->buffer + lpos, d->buffer + i, len );
+    else
+      for( ; len > 0; --len ) d->buffer[lpos++] = d->buffer[i++];
     }
   else for( ; len > 0; --len )
     {
@@ -310,16 +302,16 @@ static inline void LZd_copy_block( struct LZ_decoder * const d,
   }
 
 static inline void LZd_copy_block2( struct LZ_decoder * const d,
-                                    const int distance, int len )
+                                    const unsigned distance, unsigned len )
   {
-  if( distance < d->buffer_size )		/* block is in buffer */
+  if( d->buffer_size > distance )		/* block is in buffer */
     { LZd_copy_block( d, distance, len ); return; }
   if( len < d->buffer_size - d->pos )		/* no wrap */
     {
-    const int offset = d->pos - d->stream_pos - distance - 1;
-    if( len <= -offset )			/* block is in file */
+    const unsigned offset = distance + 1 + d->stream_pos - d->pos;
+    if( len <= offset )				/* block is in file */
       {
-      if( seek_read( d->outfd, d->buffer + d->pos, len, offset ) != len )
+      if( seek_read_back( d->outfd, d->buffer + d->pos, len, offset ) != len )
         { show_error( "Seek error", errno, false ); cleanup_and_fail( 1 ); }
       d->pos += len;
       return;
@@ -331,8 +323,8 @@ static inline void LZd_copy_block2( struct LZ_decoder * const d,
 
 static inline bool LZd_init( struct LZ_decoder * const d,
                              struct Range_decoder * const rde,
-                             const int buffer_size,
-                             const int dict_size, const int ofd )
+                             const unsigned buffer_size,
+                             const unsigned dict_size, const int ofd )
   {
   d->partial_data_pos = 0;
   d->rdec = rde;
@@ -345,20 +337,7 @@ static inline bool LZd_init( struct LZ_decoder * const d,
   d->crc = 0xFFFFFFFFU;
   d->outfd = ofd;
   d->pos_wrapped = false;
-
-  Bm_array_init( d->bm_literal[0], (1 << literal_context_bits) * 0x300 );
-  Bm_array_init( d->bm_match[0], states * pos_states );
-  Bm_array_init( d->bm_rep, states );
-  Bm_array_init( d->bm_rep0, states );
-  Bm_array_init( d->bm_rep1, states );
-  Bm_array_init( d->bm_rep2, states );
-  Bm_array_init( d->bm_len[0], states * pos_states );
-  Bm_array_init( d->bm_dis_slot[0], len_states * (1 << dis_slot_bits) );
-  Bm_array_init( d->bm_dis, modeled_distances - end_dis_model );
-  Bm_array_init( d->bm_align, dis_align_size );
-  Lm_init( &d->match_len_model );
-  Lm_init( &d->rep_len_model );
-  d->buffer[d->buffer_size-1] = 0;		/* prev_byte of first byte */
+  d->pos_wrapped_dic = false;
   return true;
   }