Update zlib to 1.2.5.

third_party/zlib/trees.c

 /* trees.c -- output deflated data using Huffman coding
- * Copyright (C) 1995-2005 Jean-loup Gailly
+ * 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
  */
 
 /*
  *  ALGORITHM
  *
  *      The "deflation" process uses several Huffman trees. The more
  *      common source values are represented by shorter bit sequences.
  *
  *      Each code tree is stored in a compressed form which is itself
  * a Huffman encoding of the lengths of all the code strings (in
  * ascending order by source values).  The actual code strings are
  * reconstructed from the lengths in the inflate process, as described
  * in the deflate specification.
  *
  *  REFERENCES
  *
  *      Deutsch, L.P.,"'Deflate' Compressed Data Format Specification".
  *      Available in ftp.uu.net:/pub/archiving/zip/doc/deflate-1.1.doc
  *
  *      Storer, James A.
  *          Data Compression:  Methods and Theory, pp. 49-50.
  *          Computer Science Press, 1988.  ISBN 0-7167-8156-5.
  *
  *      Sedgewick, R.
  *          Algorithms, p290.
  *          Addison-Wesley, 1983. ISBN 0-201-06672-6.
  */
 
-/* @(#) $Id: trees.c,v 3.6 2005/08/04 19:14:14 tor%cs.brown.edu Exp $ */
+/* @(#) $Id$ */
 
 /* #define GEN_TREES_H */
 
 #include "deflate.h"
 
 #ifdef DEBUG
 #  include <ctype.h>
 #endif
 
 /* ===========================================================================
@@ skipping 102 matching lines @@
 local void gen_bitlen     OF((deflate_state *s, tree_desc *desc));
 local void gen_codes      OF((ct_data *tree, int max_code, ushf *bl_count));
 local void build_tree     OF((deflate_state *s, tree_desc *desc));
 local void scan_tree      OF((deflate_state *s, ct_data *tree, int max_code));
 local void send_tree      OF((deflate_state *s, ct_data *tree, int max_code));
 local int  build_bl_tree  OF((deflate_state *s));
 local void send_all_trees OF((deflate_state *s, int lcodes, int dcodes,
                               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));
 local void copy_block     OF((deflate_state *s, charf *buf, unsigned len,
                               int header));
 
 #ifdef GEN_TREES_H
 local void gen_trees_header OF((void));
 #endif
 
@@ skipping 30 matching lines @@
 {
     Tracevv((stderr," l %2d v %4x ", length, value));
     Assert(length > 0 && length <= 15, "invalid length");
     s->bits_sent += (ulg)length;
 
     /* If not enough room in bi_buf, use (valid) bits from bi_buf and
      * (16 - bi_valid) bits from value, leaving (width - (16-bi_valid))
      * 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;
     }
 }
 #else /* !DEBUG */
 
 #define send_bits(s, value, 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;\
   }\
 }
 #endif /* DEBUG */
 
 
 /* the arguments must not have side effects */
 
 /* ===========================================================================
  * Initialize the various 'constant' tables.
  */
 local void tr_static_init()
 {
 #if defined(GEN_TREES_H) || !defined(STDC)
     static int static_init_done = 0;
     int n;        /* iterates over tree elements */
     int bits;     /* bit counter */
     int length;   /* length value */
     int code;     /* code value */
     int dist;     /* distance index */
     ush bl_count[MAX_BITS+1];
     /* number of codes at each bit length for an optimal tree */
 
     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;
     for (code = 0; code < LENGTH_CODES-1; code++) {
         base_length[code] = length;
         for (n = 0; n < (1<<extra_lbits[code]); n++) {
             _length_code[length++] = (uch)code;
         }
     }
     Assert (length == 256, "tr_static_init: length != 256");
@@ skipping 73 matching lines @@
         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, "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, "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, "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, "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));
     }
 
     fclose(header);
 }
 #endif /* GEN_TREES_H */
 
 /* ===========================================================================
  * 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();
 
     s->l_desc.dyn_tree = s->dyn_ltree;
     s->l_desc.stat_desc = &static_l_desc;
 
     s->d_desc.dyn_tree = s->dyn_dtree;
     s->d_desc.stat_desc = &static_d_desc;
 
@@ skipping 464 matching lines @@
     send_tree(s, (ct_data *)s->dyn_ltree, lcodes-1); /* literal tree */
     Tracev((stderr, "\nlit tree: sent %ld", s->bits_sent));
 
     send_tree(s, (ct_data *)s->dyn_dtree, dcodes-1); /* distance tree */
     Tracev((stderr, "\ndist tree: sent %ld", s->bits_sent));
 }
 
 /* ===========================================================================
  * 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;
 #endif
     copy_block(s, buf, (unsigned)stored_len, 1); /* with header */
 }
 
 /* ===========================================================================
  * Send one empty static block to give enough lookahead for inflate.
  * This takes 10 bits, of which 7 may remain in the bit buffer.
  * The current inflate code requires 9 bits of lookahead. If the
  * last two codes for the previous block (real code plus EOB) were coded
  * on 5 bits or less, inflate may have only 5+3 bits of lookahead to decode
  * the last real code. In this case we send two empty static blocks instead
  * of one. (There are no problems if the previous block is stored or fixed.)
  * 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);
     send_code(s, END_BLOCK, static_ltree);
 #ifdef DEBUG
     s->compressed_len += 10L; /* 3 for block type, 7 for EOB */
 #endif
     bi_flush(s);
     /* Of the 10 bits for the empty block, we have already sent
      * (10 - bi_valid) bits. The lookahead for the last real code (before
      * the EOB of the previous block) was thus at least one plus the length
      * of the EOB plus what we have just sent of the empty static block.
      */
     if (1 + s->last_eob_len + 10 - s->bi_valid < 9) {
         send_bits(s, STATIC_TREES<<1, 3);
         send_code(s, END_BLOCK, static_ltree);
 #ifdef DEBUG
         s->compressed_len += 10L;
 #endif
         bi_flush(s);
     }
     s->last_eob_len = 7;
 }
 
 /* ===========================================================================
  * 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 binary or text */
-        if (stored_len > 0 && s->strm->data_type == Z_UNKNOWN)
-            set_data_type(s);
+        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));
 
         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
@@ skipping 25 matching lines @@
 #else
     if (stored_len+4 <= opt_lenb && buf != (char*)0) {
                        /* 4: two words for the lengths */
 #endif
         /* The test buf != NULL is only necessary if LIT_BUFSIZE > WSIZE.
          * Otherwise we can't have processed more than WSIZE input bytes since
          * the last block flush, because compression would have been
          * 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 (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);
 #ifdef DEBUG
         s->compressed_len += 3 + s->opt_len;
 #endif
     }
     Assert (s->compressed_len == s->bits_sent, "bad compressed size");
     /* The above check is made mod 2^32, for files larger than 512 MB
      * and uLong implemented on 32 bits.
      */
     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) */
 {
     s->d_buf[s->last_lit] = (ush)dist;
     s->l_buf[s->last_lit++] = (uch)lc;
     if (dist == 0) {
         /* lc is the unmatched char */
         s->dyn_ltree[lc].Freq++;
     } else {
@@ skipping 78 matching lines @@
         Assert((uInt)(s->pending) < s->lit_bufsize + 2*lx,
                "pendingBuf overflow");
 
     } while (lx < s->last_lit);
 
     send_code(s, END_BLOCK, ltree);
     s->last_eob_len = ltree[END_BLOCK].Len;
 }
 
 /* ===========================================================================
- * Set the data type to BINARY or TEXT, using a crude approximation:
- * set it to Z_TEXT if all symbols are either printable characters (33 to 255)
- * or white spaces (9 to 13, or 32); or set it to Z_BINARY otherwise.
+ * 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;
 {
+    /* 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;
 
-    for (n = 0; n < 9; 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)
-            break;
-    if (n == 9)
-        for (n = 14; n < 32; n++)
-            if (s->dyn_ltree[n].Freq != 0)
-                break;
-    s->strm->data_type = (n == 32) ? Z_TEXT : Z_BINARY;
+            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;
 }
 
 /* ===========================================================================
  * Reverse the first len bits of a code, using straightforward code (a faster
  * method would use a table)
  * IN assertion: 1 <= len <= 15
  */
 local unsigned bi_reverse(code, len)
     unsigned code; /* the value to invert */
     int len;       /* its bit length */
@@ skipping 61 matching lines @@
         s->bits_sent += 2*16;
 #endif
     }
 #ifdef DEBUG
     s->bits_sent += (ulg)len<<3;
 #endif
     while (len--) {
         put_byte(s, *buf++);
     }
 }