/* trees.c -- output deflated data using Huffman coding
- * Copyright (C) 1995-2003 Jean-loup Gailly
- * For conditions of distribution and use, see copyright notice in zlib.h
+ * Copyright (C) 1995-2010 Jean-loup Gailly
+ * detect_data_type() function provided freely by Cosmin Truta, 2006
+ * For conditions of distribution and use, see copyright notice in zlib.h
*/
/*
int blcodes));
local void compress_block OF((deflate_state *s, ct_data *ltree,
ct_data *dtree));
-local void set_data_type OF((deflate_state *s));
+local int detect_data_type OF((deflate_state *s));
local unsigned bi_reverse OF((unsigned value, int length));
local void bi_windup OF((deflate_state *s));
local void bi_flush OF((deflate_state *s));
* unused bits in value.
*/
if (s->bi_valid > (int)Buf_size - length) {
- s->bi_buf |= (value << s->bi_valid);
+ s->bi_buf |= (ush)value << s->bi_valid;
put_short(s, s->bi_buf);
s->bi_buf = (ush)value >> (Buf_size - s->bi_valid);
s->bi_valid += length - Buf_size;
} else {
- s->bi_buf |= value << s->bi_valid;
+ s->bi_buf |= (ush)value << s->bi_valid;
s->bi_valid += length;
}
}
{ int len = length;\
if (s->bi_valid > (int)Buf_size - len) {\
int val = value;\
- s->bi_buf |= (val << s->bi_valid);\
+ s->bi_buf |= (ush)val << s->bi_valid;\
put_short(s, s->bi_buf);\
s->bi_buf = (ush)val >> (Buf_size - s->bi_valid);\
s->bi_valid += len - Buf_size;\
} else {\
- s->bi_buf |= (value) << s->bi_valid;\
+ s->bi_buf |= (ush)(value) << s->bi_valid;\
s->bi_valid += len;\
}\
}
if (static_init_done) return;
/* For some embedded targets, global variables are not initialized: */
+#ifdef NO_INIT_GLOBAL_POINTERS
static_l_desc.static_tree = static_ltree;
static_l_desc.extra_bits = extra_lbits;
static_d_desc.static_tree = static_dtree;
static_d_desc.extra_bits = extra_dbits;
static_bl_desc.extra_bits = extra_blbits;
+#endif
/* Initialize the mapping length (0..255) -> length code (0..28) */
length = 0;
Assert (header != NULL, "Can't open trees.h");
fprintf(header,
- "/* header created automatically with -DGEN_TREES_H */\n\n");
+ "/* header created automatically with -DGEN_TREES_H */\n\n");
fprintf(header, "local const ct_data static_ltree[L_CODES+2] = {\n");
for (i = 0; i < L_CODES+2; i++) {
- fprintf(header, "{{%3u},{%3u}}%s", static_ltree[i].Code,
- static_ltree[i].Len, SEPARATOR(i, L_CODES+1, 5));
+ fprintf(header, "{{%3u},{%3u}}%s", static_ltree[i].Code,
+ static_ltree[i].Len, SEPARATOR(i, L_CODES+1, 5));
}
fprintf(header, "local const ct_data static_dtree[D_CODES] = {\n");
for (i = 0; i < D_CODES; i++) {
- fprintf(header, "{{%2u},{%2u}}%s", static_dtree[i].Code,
- static_dtree[i].Len, SEPARATOR(i, D_CODES-1, 5));
+ fprintf(header, "{{%2u},{%2u}}%s", static_dtree[i].Code,
+ static_dtree[i].Len, SEPARATOR(i, D_CODES-1, 5));
}
- fprintf(header, "const uch _dist_code[DIST_CODE_LEN] = {\n");
+ fprintf(header, "const uch ZLIB_INTERNAL _dist_code[DIST_CODE_LEN] = {\n");
for (i = 0; i < DIST_CODE_LEN; i++) {
- fprintf(header, "%2u%s", _dist_code[i],
- SEPARATOR(i, DIST_CODE_LEN-1, 20));
+ fprintf(header, "%2u%s", _dist_code[i],
+ SEPARATOR(i, DIST_CODE_LEN-1, 20));
}
- fprintf(header, "const uch _length_code[MAX_MATCH-MIN_MATCH+1]= {\n");
+ fprintf(header,
+ "const uch ZLIB_INTERNAL _length_code[MAX_MATCH-MIN_MATCH+1]= {\n");
for (i = 0; i < MAX_MATCH-MIN_MATCH+1; i++) {
- fprintf(header, "%2u%s", _length_code[i],
- SEPARATOR(i, MAX_MATCH-MIN_MATCH, 20));
+ fprintf(header, "%2u%s", _length_code[i],
+ SEPARATOR(i, MAX_MATCH-MIN_MATCH, 20));
}
fprintf(header, "local const int base_length[LENGTH_CODES] = {\n");
for (i = 0; i < LENGTH_CODES; i++) {
- fprintf(header, "%1u%s", base_length[i],
- SEPARATOR(i, LENGTH_CODES-1, 20));
+ fprintf(header, "%1u%s", base_length[i],
+ SEPARATOR(i, LENGTH_CODES-1, 20));
}
fprintf(header, "local const int base_dist[D_CODES] = {\n");
for (i = 0; i < D_CODES; i++) {
- fprintf(header, "%5u%s", base_dist[i],
- SEPARATOR(i, D_CODES-1, 10));
+ fprintf(header, "%5u%s", base_dist[i],
+ SEPARATOR(i, D_CODES-1, 10));
}
fclose(header);
/* ===========================================================================
* Initialize the tree data structures for a new zlib stream.
*/
-void _tr_init(s)
+void ZLIB_INTERNAL _tr_init(s)
deflate_state *s;
{
tr_static_init();
while (n != 0) {
m = s->heap[--h];
if (m > max_code) continue;
- if (tree[m].Len != (unsigned) bits) {
+ if ((unsigned) tree[m].Len != (unsigned) bits) {
Trace((stderr,"code %d bits %d->%d\n", m, tree[m].Len, bits));
s->opt_len += ((long)bits - (long)tree[m].Len)
*(long)tree[m].Freq;
/* ===========================================================================
* Send a stored block
*/
-void _tr_stored_block(s, buf, stored_len, eof)
+void ZLIB_INTERNAL _tr_stored_block(s, buf, stored_len, last)
deflate_state *s;
charf *buf; /* input block */
ulg stored_len; /* length of input block */
- int eof; /* true if this is the last block for a file */
+ int last; /* one if this is the last block for a file */
{
- send_bits(s, (STORED_BLOCK<<1)+eof, 3); /* send block type */
+ send_bits(s, (STORED_BLOCK<<1)+last, 3); /* send block type */
#ifdef DEBUG
s->compressed_len = (s->compressed_len + 3 + 7) & (ulg)~7L;
s->compressed_len += (stored_len + 4) << 3;
* To simplify the code, we assume the worst case of last real code encoded
* on one bit only.
*/
-void _tr_align(s)
+void ZLIB_INTERNAL _tr_align(s)
deflate_state *s;
{
send_bits(s, STATIC_TREES<<1, 3);
* Determine the best encoding for the current block: dynamic trees, static
* trees or store, and output the encoded block to the zip file.
*/
-void _tr_flush_block(s, buf, stored_len, eof)
+void ZLIB_INTERNAL _tr_flush_block(s, buf, stored_len, last)
deflate_state *s;
charf *buf; /* input block, or NULL if too old */
ulg stored_len; /* length of input block */
- int eof; /* true if this is the last block for a file */
+ int last; /* one if this is the last block for a file */
{
ulg opt_lenb, static_lenb; /* opt_len and static_len in bytes */
int max_blindex = 0; /* index of last bit length code of non zero freq */
/* Build the Huffman trees unless a stored block is forced */
if (s->level > 0) {
- /* Check if the file is ascii or binary */
- if (s->data_type == Z_UNKNOWN) set_data_type(s);
+ /* Check if the file is binary or text */
+ if (s->strm->data_type == Z_UNKNOWN)
+ s->strm->data_type = detect_data_type(s);
- /* Construct the literal and distance trees */
- build_tree(s, (tree_desc *)(&(s->l_desc)));
- Tracev((stderr, "\nlit data: dyn %ld, stat %ld", s->opt_len,
- s->static_len));
+ /* Construct the literal and distance trees */
+ build_tree(s, (tree_desc *)(&(s->l_desc)));
+ Tracev((stderr, "\nlit data: dyn %ld, stat %ld", s->opt_len,
+ s->static_len));
- build_tree(s, (tree_desc *)(&(s->d_desc)));
- Tracev((stderr, "\ndist data: dyn %ld, stat %ld", s->opt_len,
- s->static_len));
- /* At this point, opt_len and static_len are the total bit lengths of
- * the compressed block data, excluding the tree representations.
- */
+ build_tree(s, (tree_desc *)(&(s->d_desc)));
+ Tracev((stderr, "\ndist data: dyn %ld, stat %ld", s->opt_len,
+ s->static_len));
+ /* At this point, opt_len and static_len are the total bit lengths of
+ * the compressed block data, excluding the tree representations.
+ */
- /* Build the bit length tree for the above two trees, and get the index
- * in bl_order of the last bit length code to send.
- */
- max_blindex = build_bl_tree(s);
+ /* Build the bit length tree for the above two trees, and get the index
+ * in bl_order of the last bit length code to send.
+ */
+ max_blindex = build_bl_tree(s);
/* Determine the best encoding. Compute the block lengths in bytes. */
- opt_lenb = (s->opt_len+3+7)>>3;
- static_lenb = (s->static_len+3+7)>>3;
+ opt_lenb = (s->opt_len+3+7)>>3;
+ static_lenb = (s->static_len+3+7)>>3;
- Tracev((stderr, "\nopt %lu(%lu) stat %lu(%lu) stored %lu lit %u ",
- opt_lenb, s->opt_len, static_lenb, s->static_len, stored_len,
- s->last_lit));
+ Tracev((stderr, "\nopt %lu(%lu) stat %lu(%lu) stored %lu lit %u ",
+ opt_lenb, s->opt_len, static_lenb, s->static_len, stored_len,
+ s->last_lit));
- if (static_lenb <= opt_lenb) opt_lenb = static_lenb;
+ if (static_lenb <= opt_lenb) opt_lenb = static_lenb;
} else {
Assert(buf != (char*)0, "lost buf");
- opt_lenb = static_lenb = stored_len + 5; /* force a stored block */
+ opt_lenb = static_lenb = stored_len + 5; /* force a stored block */
}
#ifdef FORCE_STORED
* successful. If LIT_BUFSIZE <= WSIZE, it is never too late to
* transform a block into a stored block.
*/
- _tr_stored_block(s, buf, stored_len, eof);
+ _tr_stored_block(s, buf, stored_len, last);
#ifdef FORCE_STATIC
} else if (static_lenb >= 0) { /* force static trees */
#else
- } else if (static_lenb == opt_lenb) {
+ } else if (s->strategy == Z_FIXED || static_lenb == opt_lenb) {
#endif
- send_bits(s, (STATIC_TREES<<1)+eof, 3);
+ send_bits(s, (STATIC_TREES<<1)+last, 3);
compress_block(s, (ct_data *)static_ltree, (ct_data *)static_dtree);
#ifdef DEBUG
s->compressed_len += 3 + s->static_len;
#endif
} else {
- send_bits(s, (DYN_TREES<<1)+eof, 3);
+ send_bits(s, (DYN_TREES<<1)+last, 3);
send_all_trees(s, s->l_desc.max_code+1, s->d_desc.max_code+1,
max_blindex+1);
compress_block(s, (ct_data *)s->dyn_ltree, (ct_data *)s->dyn_dtree);
*/
init_block(s);
- if (eof) {
+ if (last) {
bi_windup(s);
#ifdef DEBUG
s->compressed_len += 7; /* align on byte boundary */
#endif
}
Tracev((stderr,"\ncomprlen %lu(%lu) ", s->compressed_len>>3,
- s->compressed_len-7*eof));
+ s->compressed_len-7*last));
}
/* ===========================================================================
* Save the match info and tally the frequency counts. Return true if
* the current block must be flushed.
*/
-int _tr_tally (s, dist, lc)
+int ZLIB_INTERNAL _tr_tally (s, dist, lc)
deflate_state *s;
unsigned dist; /* distance of matched string */
unsigned lc; /* match length-MIN_MATCH or unmatched char (if dist==0) */
} /* literal or match pair ? */
/* Check that the overlay between pending_buf and d_buf+l_buf is ok: */
- Assert(s->pending < s->lit_bufsize + 2*lx, "pendingBuf overflow");
+ Assert((uInt)(s->pending) < s->lit_bufsize + 2*lx,
+ "pendingBuf overflow");
} while (lx < s->last_lit);
}
/* ===========================================================================
- * Set the data type to ASCII or BINARY, using a crude approximation:
- * binary if more than 20% of the bytes are <= 6 or >= 128, ascii otherwise.
- * IN assertion: the fields freq of dyn_ltree are set and the total of all
- * frequencies does not exceed 64K (to fit in an int on 16 bit machines).
+ * Check if the data type is TEXT or BINARY, using the following algorithm:
+ * - TEXT if the two conditions below are satisfied:
+ * a) There are no non-portable control characters belonging to the
+ * "black list" (0..6, 14..25, 28..31).
+ * b) There is at least one printable character belonging to the
+ * "white list" (9 {TAB}, 10 {LF}, 13 {CR}, 32..255).
+ * - BINARY otherwise.
+ * - The following partially-portable control characters form a
+ * "gray list" that is ignored in this detection algorithm:
+ * (7 {BEL}, 8 {BS}, 11 {VT}, 12 {FF}, 26 {SUB}, 27 {ESC}).
+ * IN assertion: the fields Freq of dyn_ltree are set.
*/
-local void set_data_type(s)
+local int detect_data_type(s)
deflate_state *s;
{
- int n = 0;
- unsigned ascii_freq = 0;
- unsigned bin_freq = 0;
- while (n < 7) bin_freq += s->dyn_ltree[n++].Freq;
- while (n < 128) ascii_freq += s->dyn_ltree[n++].Freq;
- while (n < LITERALS) bin_freq += s->dyn_ltree[n++].Freq;
- s->data_type = (Byte)(bin_freq > (ascii_freq >> 2) ? Z_BINARY : Z_ASCII);
+ /* black_mask is the bit mask of black-listed bytes
+ * set bits 0..6, 14..25, and 28..31
+ * 0xf3ffc07f = binary 11110011111111111100000001111111
+ */
+ unsigned long black_mask = 0xf3ffc07fUL;
+ int n;
+
+ /* Check for non-textual ("black-listed") bytes. */
+ for (n = 0; n <= 31; n++, black_mask >>= 1)
+ if ((black_mask & 1) && (s->dyn_ltree[n].Freq != 0))
+ return Z_BINARY;
+
+ /* Check for textual ("white-listed") bytes. */
+ if (s->dyn_ltree[9].Freq != 0 || s->dyn_ltree[10].Freq != 0
+ || s->dyn_ltree[13].Freq != 0)
+ return Z_TEXT;
+ for (n = 32; n < LITERALS; n++)
+ if (s->dyn_ltree[n].Freq != 0)
+ return Z_TEXT;
+
+ /* There are no "black-listed" or "white-listed" bytes:
+ * this stream either is empty or has tolerated ("gray-listed") bytes only.
+ */
+ return Z_BINARY;
}
/* ===========================================================================
s->last_eob_len = 8; /* enough lookahead for inflate */
if (header) {
- put_short(s, (ush)len);
+ put_short(s, (ush)len);
put_short(s, (ush)~len);
#ifdef DEBUG
s->bits_sent += 2*16;
projects / zlib.git / blobdiff