zscan 2.0.3 → 2.0.4

data/ext/pack/pack.c

@@ -9,8 +9,6 @@
 
 **********************************************************************/
 
-#include "ruby/ruby.h"
-#include "ruby/encoding.h"
 #include "internal.h"
 #include <sys/types.h>
 #include <ctype.h>
@@ -25,11 +23,11 @@
  * This behavior is consistent with the document of pack/unpack.
  */
 #ifdef HAVE_TRUE_LONG_LONG
-static const char natstr[] = "sSiIlLqQ";
+static const char natstr[] = "sSiIlLqQjJ";
 #else
-static const char natstr[] = "sSiIlL";
+static const char natstr[] = "sSiIlLjJ";
 #endif
-static const char endstr[] = "sSiIlLqQ";
+static const char endstr[] = "sSiIlLqQjJ";
 
 #ifdef HAVE_TRUE_LONG_LONG
 /* It is intentional to use long long instead of LONG_LONG. */
@@ -70,129 +68,18 @@ static const char endstr[] = "sSiIlLqQ";
 # define NATINT_LEN(type,len) ((int)sizeof(type))
 #endif
 
-#if SIZEOF_LONG == 8
-# define INT64toNUM(x) LONG2NUM(x)
-# define UINT64toNUM(x) ULONG2NUM(x)
-#elif defined(HAVE_LONG_LONG) && SIZEOF_LONG_LONG == 8
-# define INT64toNUM(x) LL2NUM(x)
-# define UINT64toNUM(x) ULL2NUM(x)
-#endif
-
-#define define_swapx(x, xtype)		\
-static xtype				\
-TOKEN_PASTE(swap,x)(xtype z)		\
-{					\
-    xtype r;				\
-    xtype *zp;				\
-    unsigned char *s, *t;		\
-    int i;				\
-					\
-    zp = xmalloc(sizeof(xtype));	\
-    *zp = z;				\
-    s = (unsigned char*)zp;		\
-    t = xmalloc(sizeof(xtype));		\
-    for (i=0; i<sizeof(xtype); i++) {	\
-	t[sizeof(xtype)-i-1] = s[i];	\
-    }					\
-    r = *(xtype *)t;			\
-    xfree(t);				\
-    xfree(zp);				\
-    return r;				\
-}
-
-#if GCC_VERSION_SINCE(4,3,0)
-# define swap32(x) __builtin_bswap32(x)
-# define swap64(x) __builtin_bswap64(x)
-#endif
-
-#ifndef swap16
-# define swap16(x)	((uint16_t)((((x)&0xFF)<<8) | (((x)>>8)&0xFF)))
-#endif
-
-#ifndef swap32
-# define swap32(x)	((uint32_t)((((x)&0xFF)<<24)	\
-			|(((x)>>24)&0xFF)	\
-			|(((x)&0x0000FF00)<<8)	\
-			|(((x)&0x00FF0000)>>8)	))
-#endif
-
-#ifndef swap64
-# ifdef HAVE_INT64_T
-#  define byte_in_64bit(n) ((uint64_t)0xff << (n))
-#  define swap64(x)       ((uint64_t)((((x)&byte_in_64bit(0))<<56) 	\
-			   |(((x)>>56)&0xFF)	                \
-			   |(((x)&byte_in_64bit(8))<<40)	\
-			   |(((x)&byte_in_64bit(48))>>40)	\
-			   |(((x)&byte_in_64bit(16))<<24)	\
-			   |(((x)&byte_in_64bit(40))>>24)	\
-			   |(((x)&byte_in_64bit(24))<<8)	\
-			   |(((x)&byte_in_64bit(32))>>8)))
-# endif
-#endif
-
-#if SIZEOF_SHORT == 2
-# define swaps(x)	swap16(x)
-#elif SIZEOF_SHORT == 4
-# define swaps(x)	swap32(x)
-#else
-  define_swapx(s,short)
-#endif
-
-#if SIZEOF_INT == 2
-# define swapi(x)	swap16(x)
-#elif SIZEOF_INT == 4
-# define swapi(x)	swap32(x)
-#else
-  define_swapx(i,int)
-#endif
-
-#if SIZEOF_LONG == 4
-# define swapl(x)	swap32(x)
-#elif SIZEOF_LONG == 8
-# define swapl(x)        swap64(x)
-#else
-  define_swapx(l,long)
-#endif
-
-#ifdef HAVE_LONG_LONG
-# if SIZEOF_LONG_LONG == 8
-#  define swapll(x)        swap64(x)
-# else
-   define_swapx(ll,LONG_LONG)
-# endif
-#endif
-
-#if SIZEOF_FLOAT == 4 && defined(HAVE_INT32_T)
-#   define swapf(x)	swap32(x)
-#   define FLOAT_SWAPPER	uint32_t
-#else
-    define_swapx(f,float)
-#endif
-
-#if SIZEOF_DOUBLE == 8 && defined(HAVE_INT64_T)
-#   define swapd(x)	swap64(x)
-#   define DOUBLE_SWAPPER	uint64_t
-#elif SIZEOF_DOUBLE == 8 && defined(HAVE_INT32_T)
-    static double
-    swapd(const double d)
-    {
-	double dtmp = d;
-	uint32_t utmp[2];
-	uint32_t utmp0;
-
-	utmp[0] = 0; utmp[1] = 0;
-	memcpy(utmp,&dtmp,sizeof(double));
-	utmp0 = utmp[0];
-	utmp[0] = swap32(utmp[1]);
-	utmp[1] = swap32(utmp0);
-	memcpy(&dtmp,utmp,sizeof(double));
-	return dtmp;
-    }
-#else
-    define_swapx(d, double)
-#endif
-
-#undef define_swapx
+typedef union {
+    float f;
+    uint32_t u;
+    char buf[4];
+} FLOAT_SWAPPER;
+typedef union {
+    double d;
+    uint64_t u;
+    char buf[8];
+} DOUBLE_SWAPPER;
+#define swapf(x) swap32(x)
+#define swapd(x) swap64(x)
 
 #define rb_ntohf(x) (BIGENDIAN_P()?(x):swapf(x))
 #define rb_ntohd(x) (BIGENDIAN_P()?(x):swapd(x))
@@ -203,74 +90,19 @@ TOKEN_PASTE(swap,x)(xtype z)		\
 #define rb_vtohf(x) (BIGENDIAN_P()?swapf(x):(x))
 #define rb_vtohd(x) (BIGENDIAN_P()?swapd(x):(x))
 
-#ifdef FLOAT_SWAPPER
-# define FLOAT_CONVWITH(y)	FLOAT_SWAPPER y;
-# define HTONF(x,y)	(memcpy(&(y),&(x),sizeof(float)),	\
-			 (y) = rb_htonf((FLOAT_SWAPPER)(y)),	\
-			 memcpy(&(x),&(y),sizeof(float)),	\
-			 (x))
-# define HTOVF(x,y)	(memcpy(&(y),&(x),sizeof(float)),	\
-			 (y) = rb_htovf((FLOAT_SWAPPER)(y)),	\
-			 memcpy(&(x),&(y),sizeof(float)),	\
-			 (x))
-# define NTOHF(x,y)	(memcpy(&(y),&(x),sizeof(float)),	\
-			 (y) = rb_ntohf((FLOAT_SWAPPER)(y)),	\
-			 memcpy(&(x),&(y),sizeof(float)),	\
-			 (x))
-# define VTOHF(x,y)	(memcpy(&(y),&(x),sizeof(float)),	\
-			 (y) = rb_vtohf((FLOAT_SWAPPER)(y)),	\
-			 memcpy(&(x),&(y),sizeof(float)),	\
-			 (x))
-#else
-# define FLOAT_CONVWITH(y)
-# define HTONF(x,y)	rb_htonf(x)
-# define HTOVF(x,y)	rb_htovf(x)
-# define NTOHF(x,y)	rb_ntohf(x)
-# define VTOHF(x,y)	rb_vtohf(x)
-#endif
+#define FLOAT_CONVWITH(x)	FLOAT_SWAPPER x;
+#define HTONF(x)	((x).u = rb_htonf((x).u))
+#define HTOVF(x)	((x).u = rb_htovf((x).u))
+#define NTOHF(x)	((x).u = rb_ntohf((x).u))
+#define VTOHF(x)	((x).u = rb_vtohf((x).u))
 
-#ifdef DOUBLE_SWAPPER
-# define DOUBLE_CONVWITH(y)	DOUBLE_SWAPPER y;
-# define HTOND(x,y)	(memcpy(&(y),&(x),sizeof(double)),	\
-			 (y) = rb_htond((DOUBLE_SWAPPER)(y)),	\
-			 memcpy(&(x),&(y),sizeof(double)),	\
-			 (x))
-# define HTOVD(x,y)	(memcpy(&(y),&(x),sizeof(double)),	\
-			 (y) = rb_htovd((DOUBLE_SWAPPER)(y)),	\
-			 memcpy(&(x),&(y),sizeof(double)),	\
-			 (x))
-# define NTOHD(x,y)	(memcpy(&(y),&(x),sizeof(double)),	\
-			 (y) = rb_ntohd((DOUBLE_SWAPPER)(y)),	\
-			 memcpy(&(x),&(y),sizeof(double)),	\
-			 (x))
-# define VTOHD(x,y)	(memcpy(&(y),&(x),sizeof(double)),	\
-			 (y) = rb_vtohd((DOUBLE_SWAPPER)(y)),	\
-			 memcpy(&(x),&(y),sizeof(double)),	\
-			 (x))
-#else
-# define DOUBLE_CONVWITH(y)
-# define HTOND(x,y)	rb_htond(x)
-# define HTOVD(x,y)	rb_htovd(x)
-# define NTOHD(x,y)	rb_ntohd(x)
-# define VTOHD(x,y)	rb_vtohd(x)
-#endif
-
-static unsigned long
-num2i32(VALUE x)
-{
-    x = rb_to_int(x); /* is nil OK? (should not) */
-
-    if (FIXNUM_P(x)) return FIX2LONG(x);
-    if (RB_TYPE_P(x, T_BIGNUM)) {
-	return rb_big2ulong_pack(x);
-    }
-    rb_raise(rb_eTypeError, "can't convert %s to `integer'", rb_obj_classname(x));
-
-    UNREACHABLE;
-}
+#define DOUBLE_CONVWITH(x)	DOUBLE_SWAPPER x;
+#define HTOND(x)	((x).u = rb_htond((x).u))
+#define HTOVD(x)	((x).u = rb_htovd((x).u))
+#define NTOHD(x)	((x).u = rb_ntohd((x).u))
+#define VTOHD(x)	((x).u = rb_vtohd((x).u))
 
 #define MAX_INTEGER_PACK_SIZE 8
-/* #define FORCE_BIG_PACK */
 
 static const char toofew[] = "too few arguments";
 
@@ -279,802 +111,19 @@ static void qpencode(VALUE,VALUE,long);
 
 static unsigned long utf8_to_uv(const char*,long*);
 
-/*
- *  call-seq:
- *     arr.pack ( aTemplateString ) -> aBinaryString
- *
- *  Packs the contents of <i>arr</i> into a binary sequence according to
- *  the directives in <i>aTemplateString</i> (see the table below)
- *  Directives ``A,'' ``a,'' and ``Z'' may be followed by a count,
- *  which gives the width of the resulting field. The remaining
- *  directives also may take a count, indicating the number of array
- *  elements to convert. If the count is an asterisk
- *  (``<code>*</code>''), all remaining array elements will be
- *  converted. Any of the directives ``<code>sSiIlL</code>'' may be
- *  followed by an underscore (``<code>_</code>'') or
- *  exclamation mark (``<code>!</code>'') to use the underlying
- *  platform's native size for the specified type; otherwise, they use a
- *  platform-independent size. Spaces are ignored in the template
- *  string. See also <code>String#unpack</code>.
- *
- *     a = [ "a", "b", "c" ]
- *     n = [ 65, 66, 67 ]
- *     a.pack("A3A3A3")   #=> "a  b  c  "
- *     a.pack("a3a3a3")   #=> "a\000\000b\000\000c\000\000"
- *     n.pack("ccc")      #=> "ABC"
- *
- *  Directives for +pack+.
- *
- *   Integer      | Array   |
- *   Directive    | Element | Meaning
- *   ---------------------------------------------------------------------------
- *      C         | Integer | 8-bit unsigned (unsigned char)
- *      S         | Integer | 16-bit unsigned, native endian (uint16_t)
- *      L         | Integer | 32-bit unsigned, native endian (uint32_t)
- *      Q         | Integer | 64-bit unsigned, native endian (uint64_t)
- *                |         |
- *      c         | Integer | 8-bit signed (signed char)
- *      s         | Integer | 16-bit signed, native endian (int16_t)
- *      l         | Integer | 32-bit signed, native endian (int32_t)
- *      q         | Integer | 64-bit signed, native endian (int64_t)
- *                |         |
- *      S_, S!    | Integer | unsigned short, native endian
- *      I, I_, I! | Integer | unsigned int, native endian
- *      L_, L!    | Integer | unsigned long, native endian
- *      Q_, Q!    | Integer | unsigned long long, native endian (ArgumentError
- *                |         | if the platform has no long long type.)
- *                |         | (Q_ and Q! is available since Ruby 2.1.)
- *                |         |
- *      s_, s!    | Integer | signed short, native endian
- *      i, i_, i! | Integer | signed int, native endian
- *      l_, l!    | Integer | signed long, native endian
- *      q_, q!    | Integer | signed long long, native endian (ArgumentError
- *                |         | if the platform has no long long type.)
- *                |         | (q_ and q! is available since Ruby 2.1.)
- *                |         |
- *      S> L> Q>  | Integer | same as the directives without ">" except
- *      s> l> q>  |         | big endian
- *      S!> I!>   |         | (available since Ruby 1.9.3)
- *      L!> Q!>   |         | "S>" is same as "n"
- *      s!> i!>   |         | "L>" is same as "N"
- *      l!> q!>   |         |
- *                |         |
- *      S< L< Q<  | Integer | same as the directives without "<" except
- *      s< l< q<  |         | little endian
- *      S!< I!<   |         | (available since Ruby 1.9.3)
- *      L!< Q!<   |         | "S<" is same as "v"
- *      s!< i!<   |         | "L<" is same as "V"
- *      l!< q!<   |         |
- *                |         |
- *      n         | Integer | 16-bit unsigned, network (big-endian) byte order
- *      N         | Integer | 32-bit unsigned, network (big-endian) byte order
- *      v         | Integer | 16-bit unsigned, VAX (little-endian) byte order
- *      V         | Integer | 32-bit unsigned, VAX (little-endian) byte order
- *                |         |
- *      U         | Integer | UTF-8 character
- *      w         | Integer | BER-compressed integer
- *
- *   Float        |         |
- *   Directive    |         | Meaning
- *   ---------------------------------------------------------------------------
- *      D, d      | Float   | double-precision, native format
- *      F, f      | Float   | single-precision, native format
- *      E         | Float   | double-precision, little-endian byte order
- *      e         | Float   | single-precision, little-endian byte order
- *      G         | Float   | double-precision, network (big-endian) byte order
- *      g         | Float   | single-precision, network (big-endian) byte order
- *
- *   String       |         |
- *   Directive    |         | Meaning
- *   ---------------------------------------------------------------------------
- *      A         | String  | arbitrary binary string (space padded, count is width)
- *      a         | String  | arbitrary binary string (null padded, count is width)
- *      Z         | String  | same as ``a'', except that null is added with *
- *      B         | String  | bit string (MSB first)
- *      b         | String  | bit string (LSB first)
- *      H         | String  | hex string (high nibble first)
- *      h         | String  | hex string (low nibble first)
- *      u         | String  | UU-encoded string
- *      M         | String  | quoted printable, MIME encoding (see RFC2045)
- *      m         | String  | base64 encoded string (see RFC 2045, count is width)
- *                |         | (if count is 0, no line feed are added, see RFC 4648)
- *      P         | String  | pointer to a structure (fixed-length string)
- *      p         | String  | pointer to a null-terminated string
- *
- *   Misc.        |         |
- *   Directive    |         | Meaning
- *   ---------------------------------------------------------------------------
- *      @         | ---     | moves to absolute position
- *      X         | ---     | back up a byte
- *      x         | ---     | null byte
- */
+static ID id_associated;
 
-__attribute__ ((unused))
-static VALUE
-pack_pack(VALUE ary, VALUE fmt)
+static void
+str_associate(VALUE str, VALUE add)
 {
-    static const char nul10[] = "\0\0\0\0\0\0\0\0\0\0";
-    static const char spc10[] = "          ";
-    const char *p, *pend;
-    VALUE res, from, associates = 0;
-    char type;
-    long items, len, idx, plen;
-    const char *ptr;
-    int enc_info = 1;		/* 0 - BINARY, 1 - US-ASCII, 2 - UTF-8 */
-#ifdef NATINT_PACK
-    int natint;		/* native integer */
-#endif
-    int integer_size, bigendian_p;
-
-    StringValue(fmt);
-    p = RSTRING_PTR(fmt);
-    pend = p + RSTRING_LEN(fmt);
-    res = rb_str_buf_new(0);
-
-    items = RARRAY_LEN(ary);
-    idx = 0;
-
-#define TOO_FEW (rb_raise(rb_eArgError, toofew), 0)
-#define THISFROM (items > 0 ? RARRAY_PTR(ary)[idx] : TOO_FEW)
-#define NEXTFROM (items-- > 0 ? RARRAY_PTR(ary)[idx++] : TOO_FEW)
-
-    while (p < pend) {
-	int explicit_endian = 0;
-	if (RSTRING_PTR(fmt) + RSTRING_LEN(fmt) != pend) {
-	    rb_raise(rb_eRuntimeError, "format string modified");
-	}
-	type = *p++;		/* get data type */
-#ifdef NATINT_PACK
-	natint = 0;
-#endif
-
-	if (ISSPACE(type)) continue;
-	if (type == '#') {
-	    while ((p < pend) && (*p != '\n')) {
-		p++;
-	    }
-	    continue;
-	}
-
-	{
-          modifiers:
-	    switch (*p) {
-	      case '_':
-	      case '!':
-		if (strchr(natstr, type)) {
-#ifdef NATINT_PACK
-		    natint = 1;
-#endif
-		    p++;
-		}
-		else {
-		    rb_raise(rb_eArgError, "'%c' allowed only after types %s", *p, natstr);
-		}
-		goto modifiers;
-
-	      case '<':
-	      case '>':
-		if (!strchr(endstr, type)) {
-		    rb_raise(rb_eArgError, "'%c' allowed only after types %s", *p, endstr);
-		}
-		if (explicit_endian) {
-		    rb_raise(rb_eRangeError, "Can't use both '<' and '>'");
-		}
-		explicit_endian = *p++;
-		goto modifiers;
-	    }
-	}
-
-	if (*p == '*') {	/* set data length */
-	    len = strchr("@Xxu", type) ? 0
-                : strchr("PMm", type) ? 1
-                : items;
-	    p++;
-	}
-	else if (ISDIGIT(*p)) {
-	    errno = 0;
-	    len = STRTOUL(p, (char**)&p, 10);
-	    if (errno) {
-		rb_raise(rb_eRangeError, "pack length too big");
-	    }
-	}
-	else {
-	    len = 1;
-	}
-
-	switch (type) {
-	  case 'U':
-	    /* if encoding is US-ASCII, upgrade to UTF-8 */
-	    if (enc_info == 1) enc_info = 2;
-	    break;
-	  case 'm': case 'M': case 'u':
-	    /* keep US-ASCII (do nothing) */
-	    break;
-	  default:
-	    /* fall back to BINARY */
-	    enc_info = 0;
-	    break;
-	}
-	switch (type) {
-	  case 'A': case 'a': case 'Z':
-	  case 'B': case 'b':
-	  case 'H': case 'h':
-	    from = NEXTFROM;
-	    if (NIL_P(from)) {
-		ptr = "";
-		plen = 0;
-	    }
-	    else {
-		StringValue(from);
-		ptr = RSTRING_PTR(from);
-		plen = RSTRING_LEN(from);
-		OBJ_INFECT(res, from);
-	    }
-
-	    if (p[-1] == '*')
-		len = plen;
-
-	    switch (type) {
-	      case 'a':		/* arbitrary binary string (null padded)  */
-	      case 'A':         /* arbitrary binary string (ASCII space padded) */
-	      case 'Z':         /* null terminated string  */
-		if (plen >= len) {
-		    rb_str_buf_cat(res, ptr, len);
-		    if (p[-1] == '*' && type == 'Z')
-			rb_str_buf_cat(res, nul10, 1);
-		}
-		else {
-		    rb_str_buf_cat(res, ptr, plen);
-		    len -= plen;
-		    while (len >= 10) {
-			rb_str_buf_cat(res, (type == 'A')?spc10:nul10, 10);
-			len -= 10;
-		    }
-		    rb_str_buf_cat(res, (type == 'A')?spc10:nul10, len);
-		}
-		break;
-
-#define castchar(from) (char)((from) & 0xff)
-
-	      case 'b':		/* bit string (ascending) */
-		{
-		    int byte = 0;
-		    long i, j = 0;
-
-		    if (len > plen) {
-			j = (len - plen + 1)/2;
-			len = plen;
-		    }
-		    for (i=0; i++ < len; ptr++) {
-			if (*ptr & 1)
-			    byte |= 128;
-			if (i & 7)
-			    byte >>= 1;
-			else {
-			    char c = castchar(byte);
-			    rb_str_buf_cat(res, &c, 1);
-			    byte = 0;
-			}
-		    }
-		    if (len & 7) {
-			char c;
-			byte >>= 7 - (len & 7);
-			c = castchar(byte);
-			rb_str_buf_cat(res, &c, 1);
-		    }
-		    len = j;
-		    goto grow;
-		}
-		break;
-
-	      case 'B':		/* bit string (descending) */
-		{
-		    int byte = 0;
-		    long i, j = 0;
-
-		    if (len > plen) {
-			j = (len - plen + 1)/2;
-			len = plen;
-		    }
-		    for (i=0; i++ < len; ptr++) {
-			byte |= *ptr & 1;
-			if (i & 7)
-			    byte <<= 1;
-			else {
-			    char c = castchar(byte);
-			    rb_str_buf_cat(res, &c, 1);
-			    byte = 0;
-			}
-		    }
-		    if (len & 7) {
-			char c;
-			byte <<= 7 - (len & 7);
-			c = castchar(byte);
-			rb_str_buf_cat(res, &c, 1);
-		    }
-		    len = j;
-		    goto grow;
-		}
-		break;
-
-	      case 'h':		/* hex string (low nibble first) */
-		{
-		    int byte = 0;
-		    long i, j = 0;
-
-		    if (len > plen) {
-			j = (len + 1) / 2 - (plen + 1) / 2;
-			len = plen;
-		    }
-		    for (i=0; i++ < len; ptr++) {
-			if (ISALPHA(*ptr))
-			    byte |= (((*ptr & 15) + 9) & 15) << 4;
-			else
-			    byte |= (*ptr & 15) << 4;
-			if (i & 1)
-			    byte >>= 4;
-			else {
-			    char c = castchar(byte);
-			    rb_str_buf_cat(res, &c, 1);
-			    byte = 0;
-			}
-		    }
-		    if (len & 1) {
-			char c = castchar(byte);
-			rb_str_buf_cat(res, &c, 1);
-		    }
-		    len = j;
-		    goto grow;
-		}
-		break;
-
-	      case 'H':		/* hex string (high nibble first) */
-		{
-		    int byte = 0;
-		    long i, j = 0;
-
-		    if (len > plen) {
-			j = (len + 1) / 2 - (plen + 1) / 2;
-			len = plen;
-		    }
-		    for (i=0; i++ < len; ptr++) {
-			if (ISALPHA(*ptr))
-			    byte |= ((*ptr & 15) + 9) & 15;
-			else
-			    byte |= *ptr & 15;
-			if (i & 1)
-			    byte <<= 4;
-			else {
-			    char c = castchar(byte);
-			    rb_str_buf_cat(res, &c, 1);
-			    byte = 0;
-			}
-		    }
-		    if (len & 1) {
-			char c = castchar(byte);
-			rb_str_buf_cat(res, &c, 1);
-		    }
-		    len = j;
-		    goto grow;
-		}
-		break;
-	    }
-	    break;
-
-	  case 'c':		/* signed char */
-	  case 'C':		/* unsigned char */
-	    while (len-- > 0) {
-		char c;
-
-		from = NEXTFROM;
-		c = (char)num2i32(from);
-		rb_str_buf_cat(res, &c, sizeof(char));
-	    }
-	    break;
-
-	  case 's':		/* s for int16_t, s! for signed short */
-            integer_size = NATINT_LEN(short, 2);
-            bigendian_p = BIGENDIAN_P();
-            goto pack_integer;
-
-	  case 'S':		/* S for uint16_t, S! for unsigned short */
-            integer_size = NATINT_LEN(short, 2);
-            bigendian_p = BIGENDIAN_P();
-            goto pack_integer;
-
-	  case 'i':		/* i and i! for signed int */
-            integer_size = (int)sizeof(int);
-            bigendian_p = BIGENDIAN_P();
-            goto pack_integer;
-
-	  case 'I':		/* I and I! for unsigned int */
-            integer_size = (int)sizeof(int);
-            bigendian_p = BIGENDIAN_P();
-            goto pack_integer;
-
-	  case 'l':		/* l for int32_t, l! for signed long */
-            integer_size = NATINT_LEN(long, 4);
-            bigendian_p = BIGENDIAN_P();
-            goto pack_integer;
-
-	  case 'L':		/* L for uint32_t, L! for unsigned long */
-            integer_size = NATINT_LEN(long, 4);
-            bigendian_p = BIGENDIAN_P();
-            goto pack_integer;
-
-	  case 'q':		/* q for int64_t, q! for signed long long */
-	    integer_size = NATINT_LEN_Q;
-            bigendian_p = BIGENDIAN_P();
-            goto pack_integer;
-
-	  case 'Q':		/* Q for uint64_t, Q! for unsigned long long */
-	    integer_size = NATINT_LEN_Q;
-            bigendian_p = BIGENDIAN_P();
-            goto pack_integer;
-
-	  case 'n':		/* 16 bit (2 bytes) integer (network byte-order)  */
-            integer_size = 2;
-            bigendian_p = 1;
-            goto pack_integer;
-
-	  case 'N':		/* 32 bit (4 bytes) integer (network byte-order) */
-            integer_size = 4;
-            bigendian_p = 1;
-            goto pack_integer;
-
-	  case 'v':		/* 16 bit (2 bytes) integer (VAX byte-order) */
-            integer_size = 2;
-            bigendian_p = 0;
-            goto pack_integer;
-
-	  case 'V':		/* 32 bit (4 bytes) integer (VAX byte-order) */
-            integer_size = 4;
-            bigendian_p = 0;
-            goto pack_integer;
-
-          pack_integer:
-	    if (explicit_endian) {
-		bigendian_p = explicit_endian == '>';
-	    }
-
-            switch (integer_size) {
-#if defined(HAVE_INT16_T) && !defined(FORCE_BIG_PACK)
-              case SIZEOF_INT16_T:
-		while (len-- > 0) {
-                    union {
-                        int16_t i;
-                        char a[sizeof(int16_t)];
-                    } v;
-
-		    from = NEXTFROM;
-		    v.i = (int16_t)num2i32(from);
-		    if (bigendian_p != BIGENDIAN_P()) v.i = swap16(v.i);
-		    rb_str_buf_cat(res, v.a, sizeof(int16_t));
-		}
-		break;
-#endif
-
-#if defined(HAVE_INT32_T) && !defined(FORCE_BIG_PACK)
-              case SIZEOF_INT32_T:
-		while (len-- > 0) {
-		    union {
-                        int32_t i;
-                        char a[sizeof(int32_t)];
-                    } v;
-
-		    from = NEXTFROM;
-		    v.i = (int32_t)num2i32(from);
-		    if (bigendian_p != BIGENDIAN_P()) v.i = swap32(v.i);
-		    rb_str_buf_cat(res, v.a, sizeof(int32_t));
-		}
-		break;
-#endif
-
-#if defined(HAVE_INT64_T) && SIZEOF_LONG == SIZEOF_INT64_T && !defined(FORCE_BIG_PACK)
-              case SIZEOF_INT64_T:
-		while (len-- > 0) {
-		    union {
-                        int64_t i;
-                        char a[sizeof(int64_t)];
-                    } v;
-
-		    from = NEXTFROM;
-		    v.i = num2i32(from); /* can return 64bit value if SIZEOF_LONG == SIZEOF_INT64_T */
-		    if (bigendian_p != BIGENDIAN_P()) v.i = swap64(v.i);
-		    rb_str_buf_cat(res, v.a, sizeof(int64_t));
-		}
-		break;
-#endif
-
-	      default:
-                if (integer_size > MAX_INTEGER_PACK_SIZE)
-                    rb_bug("unexpected intger size for pack: %d", integer_size);
-                while (len-- > 0) {
-                    union {
-                        unsigned long i[(MAX_INTEGER_PACK_SIZE+SIZEOF_LONG-1)/SIZEOF_LONG];
-                        char a[(MAX_INTEGER_PACK_SIZE+SIZEOF_LONG-1)/SIZEOF_LONG*SIZEOF_LONG];
-                    } v;
-                    int num_longs = (integer_size+SIZEOF_LONG-1)/SIZEOF_LONG;
-                    int i;
-
-                    from = NEXTFROM;
-                    rb_big_pack(from, v.i, num_longs);
-                    if (bigendian_p) {
-                        for (i = 0; i < num_longs/2; i++) {
-                            unsigned long t = v.i[i];
-                            v.i[i] = v.i[num_longs-1-i];
-                            v.i[num_longs-1-i] = t;
-                        }
-                    }
-		    if (bigendian_p != BIGENDIAN_P()) {
-                        for (i = 0; i < num_longs; i++)
-                            v.i[i] = swapl(v.i[i]);
-                    }
-                    rb_str_buf_cat(res,
-                                   bigendian_p ?
-                                     v.a + sizeof(long)*num_longs - integer_size :
-                                     v.a,
-                                   integer_size);
-                }
-                break;
-	    }
-	    break;
-
-	  case 'f':		/* single precision float in native format */
-	  case 'F':		/* ditto */
-	    while (len-- > 0) {
-		float f;
-
-		from = NEXTFROM;
-		f = (float)RFLOAT_VALUE(rb_to_float(from));
-		rb_str_buf_cat(res, (char*)&f, sizeof(float));
-	    }
-	    break;
-
-	  case 'e':		/* single precision float in VAX byte-order */
-	    while (len-- > 0) {
-		float f;
-		FLOAT_CONVWITH(ftmp);
-
-		from = NEXTFROM;
-		f = (float)RFLOAT_VALUE(rb_to_float(from));
-		f = HTOVF(f,ftmp);
-		rb_str_buf_cat(res, (char*)&f, sizeof(float));
-	    }
-	    break;
-
-	  case 'E':		/* double precision float in VAX byte-order */
-	    while (len-- > 0) {
-		double d;
-		DOUBLE_CONVWITH(dtmp);
-
-		from = NEXTFROM;
-		d = RFLOAT_VALUE(rb_to_float(from));
-		d = HTOVD(d,dtmp);
-		rb_str_buf_cat(res, (char*)&d, sizeof(double));
-	    }
-	    break;
-
-	  case 'd':		/* double precision float in native format */
-	  case 'D':		/* ditto */
-	    while (len-- > 0) {
-		double d;
-
-		from = NEXTFROM;
-		d = RFLOAT_VALUE(rb_to_float(from));
-		rb_str_buf_cat(res, (char*)&d, sizeof(double));
-	    }
-	    break;
-
-	  case 'g':		/* single precision float in network byte-order */
-	    while (len-- > 0) {
-		float f;
-		FLOAT_CONVWITH(ftmp);
-
-		from = NEXTFROM;
-		f = (float)RFLOAT_VALUE(rb_to_float(from));
-		f = HTONF(f,ftmp);
-		rb_str_buf_cat(res, (char*)&f, sizeof(float));
-	    }
-	    break;
-
-	  case 'G':		/* double precision float in network byte-order */
-	    while (len-- > 0) {
-		double d;
-		DOUBLE_CONVWITH(dtmp);
-
-		from = NEXTFROM;
-		d = RFLOAT_VALUE(rb_to_float(from));
-		d = HTOND(d,dtmp);
-		rb_str_buf_cat(res, (char*)&d, sizeof(double));
-	    }
-	    break;
-
-	  case 'x':		/* null byte */
-	  grow:
-	    while (len >= 10) {
-		rb_str_buf_cat(res, nul10, 10);
-		len -= 10;
-	    }
-	    rb_str_buf_cat(res, nul10, len);
-	    break;
-
-	  case 'X':		/* back up byte */
-	  shrink:
-	    plen = RSTRING_LEN(res);
-	    if (plen < len)
-		rb_raise(rb_eArgError, "X outside of string");
-	    rb_str_set_len(res, plen - len);
-	    break;
-
-	  case '@':		/* null fill to absolute position */
-	    len -= RSTRING_LEN(res);
-	    if (len > 0) goto grow;
-	    len = -len;
-	    if (len > 0) goto shrink;
-	    break;
-
-	  case '%':
-	    rb_raise(rb_eArgError, "%% is not supported");
-	    break;
-
-	  case 'U':		/* Unicode character */
-	    while (len-- > 0) {
-		SIGNED_VALUE l;
-		char buf[8];
-		int le;
-
-		from = NEXTFROM;
-		from = rb_to_int(from);
-		l = NUM2LONG(from);
-		if (l < 0) {
-		    rb_raise(rb_eRangeError, "pack(U): value out of range");
-		}
-		le = rb_uv_to_utf8(buf, l);
-		rb_str_buf_cat(res, (char*)buf, le);
-	    }
-	    break;
-
-	  case 'u':		/* uuencoded string */
-	  case 'm':		/* base64 encoded string */
-	    from = NEXTFROM;
-	    StringValue(from);
-	    ptr = RSTRING_PTR(from);
-	    plen = RSTRING_LEN(from);
-
-	    if (len == 0 && type == 'm') {
-		encodes(res, ptr, plen, type, 0);
-		ptr += plen;
-		break;
-	    }
-	    if (len <= 2)
-		len = 45;
-	    else if (len > 63 && type == 'u')
-		len = 63;
-	    else
-		len = len / 3 * 3;
-	    while (plen > 0) {
-		long todo;
-
-		if (plen > len)
-		    todo = len;
-		else
-		    todo = plen;
-		encodes(res, ptr, todo, type, 1);
-		plen -= todo;
-		ptr += todo;
-	    }
-	    break;
-
-	  case 'M':		/* quoted-printable encoded string */
-	    from = rb_obj_as_string(NEXTFROM);
-	    if (len <= 1)
-		len = 72;
-	    qpencode(res, from, len);
-	    break;
-
-	  case 'P':		/* pointer to packed byte string */
-	    from = THISFROM;
-	    if (!NIL_P(from)) {
-		StringValue(from);
-		if (RSTRING_LEN(from) < len) {
-		    rb_raise(rb_eArgError, "too short buffer for P(%ld for %ld)",
-			     RSTRING_LEN(from), len);
-		}
-	    }
-	    len = 1;
-	    /* FALL THROUGH */
-	  case 'p':		/* pointer to string */
-	    while (len-- > 0) {
-		char *t;
-		from = NEXTFROM;
-		if (NIL_P(from)) {
-		    t = 0;
-		}
-		else {
-		    t = StringValuePtr(from);
-		}
-		if (!associates) {
-		    associates = rb_ary_new();
-		}
-		rb_ary_push(associates, from);
-		rb_obj_taint(from);
-		rb_str_buf_cat(res, (char*)&t, sizeof(char*));
-	    }
-	    break;
-
-	  case 'w':		/* BER compressed integer  */
-	    while (len-- > 0) {
-		unsigned long ul;
-		VALUE buf = rb_str_new(0, 0);
-		char c, *bufs, *bufe;
-
-		from = NEXTFROM;
-		if (RB_TYPE_P(from, T_BIGNUM)) {
-		    VALUE big128 = rb_uint2big(128);
-		    while (RB_TYPE_P(from, T_BIGNUM)) {
-			from = rb_big_divmod(from, big128);
-			c = castchar(NUM2INT(RARRAY_PTR(from)[1]) | 0x80); /* mod */
-			rb_str_buf_cat(buf, &c, sizeof(char));
-			from = RARRAY_PTR(from)[0]; /* div */
-		    }
-		}
-
-		{
-		    long l = NUM2LONG(from);
-		    if (l < 0) {
-			rb_raise(rb_eArgError, "can't compress negative numbers");
-		    }
-		    ul = l;
-		}
-
-		while (ul) {
-		    c = castchar((ul & 0x7f) | 0x80);
-		    rb_str_buf_cat(buf, &c, sizeof(char));
-		    ul >>=  7;
-		}
-
-		if (RSTRING_LEN(buf)) {
-		    bufs = RSTRING_PTR(buf);
-		    bufe = bufs + RSTRING_LEN(buf) - 1;
-		    *bufs &= 0x7f; /* clear continue bit */
-		    while (bufs < bufe) { /* reverse */
-			c = *bufs;
-			*bufs++ = *bufe;
-			*bufe-- = c;
-		    }
-		    rb_str_buf_cat(res, RSTRING_PTR(buf), RSTRING_LEN(buf));
-		}
-		else {
-		    c = 0;
-		    rb_str_buf_cat(res, &c, sizeof(char));
-		}
-	    }
-	    break;
-
-	  default:
-	    rb_warning("unknown pack directive '%c' in '%s'",
-		type, RSTRING_PTR(fmt));
-	    break;
-	}
-    }
+    /* assert(NIL_P(rb_attr_get(str, id_associated))); */
+    rb_ivar_set(str, id_associated, add);
+}
 
-    if (associates) {
-	rb_str_associate(res, associates);
-    }
-    OBJ_INFECT(res, fmt);
-    switch (enc_info) {
-      case 1:
-	ENCODING_CODERANGE_SET(res, rb_usascii_encindex(), ENC_CODERANGE_7BIT);
-	break;
-      case 2:
-	rb_enc_set_index(res, rb_utf8_encindex());
-	break;
-      default:
-	/* do nothing, keep ASCII-8BIT */
-	break;
-    }
-    return res;
+static VALUE
+str_associated(VALUE str)
+{
+    return rb_ivar_lookup(str, id_associated, Qfalse);
 }
 
 static const char uu_table[] =
@@ -1083,12 +132,14 @@ static const char b64_table[] =
 "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
 
 static void
-encodes(VALUE str, const char *s, long len, int type, int tail_lf)
+encodes(VALUE str, const char *s0, long len, int type, int tail_lf)
 {
-    char buff[4096];
+    enum {buff_size = 4096, encoded_unit = 4, input_unit = 3};
+    char buff[buff_size + 1];	/* +1 for tail_lf */
     long i = 0;
-    const char *trans = type == 'u' ? uu_table : b64_table;
+    const char *const trans = type == 'u' ? uu_table : b64_table;
     char padding;
+    const unsigned char *s = (const unsigned char *)s0;
 
     if (type == 'u') {
 	buff[i++] = (char)len + ' ';
@@ -1097,16 +148,16 @@ encodes(VALUE str, const char *s, long len, int type, int tail_lf)
     else {
 	padding = '=';
     }
-    while (len >= 3) {
-        while (len >= 3 && sizeof(buff)-i >= 4) {
+    while (len >= input_unit) {
+        while (len >= input_unit && buff_size-i >= encoded_unit) {
             buff[i++] = trans[077 & (*s >> 2)];
             buff[i++] = trans[077 & (((*s << 4) & 060) | ((s[1] >> 4) & 017))];
             buff[i++] = trans[077 & (((s[1] << 2) & 074) | ((s[2] >> 6) & 03))];
             buff[i++] = trans[077 & s[2]];
-            s += 3;
-            len -= 3;
+            s += input_unit;
+            len -= input_unit;
         }
-        if (sizeof(buff)-i < 4) {
+        if (buff_size-i < encoded_unit) {
             rb_str_buf_cat(str, buff, i);
             i = 0;
         }
@@ -1126,6 +177,7 @@ encodes(VALUE str, const char *s, long len, int type, int tail_lf)
     }
     if (tail_lf) buff[i++] = '\n';
     rb_str_buf_cat(str, buff, i);
+    if ((size_t)i > sizeof(buff)) rb_bug("encodes() buffer overrun");
 }
 
 static const char hex_table[] = "0123456789ABCDEF";
@@ -1186,19 +238,11 @@ qpencode(VALUE str, VALUE from, long len)
 static inline int
 hex2num(char c)
 {
-    switch (c) {
-      case '0': case '1': case '2': case '3': case '4':
-      case '5': case '6': case '7': case '8': case '9':
-        return c - '0';
-      case 'a': case 'b': case 'c':
-      case 'd': case 'e': case 'f':
-	return c - 'a' + 10;
-      case 'A': case 'B': case 'C':
-      case 'D': case 'E': case 'F':
-	return c - 'A' + 10;
-      default:
-	return -1;
-    }
+    int n;
+    n = ruby_digit36_to_number_table[(unsigned char)c];
+    if (16 <= n)
+        n = -1;
+    return n;
 }
 
 #define PACK_LENGTH_ADJUST_SIZE(sz) do {	\
@@ -1212,10 +256,19 @@ hex2num(char c)
 } while (0)
 
 #define PACK_ITEM_ADJUST() do { \
-    if (tmp_len > 0 && !block_p) \
+    if (tmp_len > 0 && mode == UNPACK_ARRAY) \
 	rb_ary_store(ary, RARRAY_LEN(ary)+tmp_len-1, Qnil); \
 } while (0)
 
+/* Workaround for Oracle Solaris Studio 12.4 C compiler optimization bug
+ * with "-xO4" optimization option.
+ */
+#if defined(__SUNPRO_C) && __SUNPRO_C == 0x5130
+# define AVOID_CC_BUG volatile
+#else
+# define AVOID_CC_BUG
+#endif
+
 static VALUE
 infected_str_new(const char *ptr, long len, VALUE str)
 {
@@ -1225,137 +278,31 @@ infected_str_new(const char *ptr, long len, VALUE str)
     return s;
 }
 
-/*
- *  call-seq:
- *     str.unpack(format)    ->  anArray
- *
- *  Decodes <i>str</i> (which may contain binary data) according to the
- *  format string, returning an array of each value extracted. The
- *  format string consists of a sequence of single-character directives,
- *  summarized in the table at the end of this entry.
- *  Each directive may be followed
- *  by a number, indicating the number of times to repeat with this
- *  directive. An asterisk (``<code>*</code>'') will use up all
- *  remaining elements. The directives <code>sSiIlL</code> may each be
- *  followed by an underscore (``<code>_</code>'') or
- *  exclamation mark (``<code>!</code>'') to use the underlying
- *  platform's native size for the specified type; otherwise, it uses a
- *  platform-independent consistent size. Spaces are ignored in the
- *  format string. See also <code>Array#pack</code>.
- *
- *     "abc \0\0abc \0\0".unpack('A6Z6')   #=> ["abc", "abc "]
- *     "abc \0\0".unpack('a3a3')           #=> ["abc", " \000\000"]
- *     "abc \0abc \0".unpack('Z*Z*')       #=> ["abc ", "abc "]
- *     "aa".unpack('b8B8')                 #=> ["10000110", "01100001"]
- *     "aaa".unpack('h2H2c')               #=> ["16", "61", 97]
- *     "\xfe\xff\xfe\xff".unpack('sS')     #=> [-2, 65534]
- *     "now=20is".unpack('M*')             #=> ["now is"]
- *     "whole".unpack('xax2aX2aX1aX2a')    #=> ["h", "e", "l", "l", "o"]
- *
- *  This table summarizes the various formats and the Ruby classes
- *  returned by each.
- *
- *   Integer      |         |
- *   Directive    | Returns | Meaning
- *   -----------------------------------------------------------------
- *      C         | Integer | 8-bit unsigned (unsigned char)
- *      S         | Integer | 16-bit unsigned, native endian (uint16_t)
- *      L         | Integer | 32-bit unsigned, native endian (uint32_t)
- *      Q         | Integer | 64-bit unsigned, native endian (uint64_t)
- *                |         |
- *      c         | Integer | 8-bit signed (signed char)
- *      s         | Integer | 16-bit signed, native endian (int16_t)
- *      l         | Integer | 32-bit signed, native endian (int32_t)
- *      q         | Integer | 64-bit signed, native endian (int64_t)
- *                |         |
- *      S_, S!    | Integer | unsigned short, native endian
- *      I, I_, I! | Integer | unsigned int, native endian
- *      L_, L!    | Integer | unsigned long, native endian
- *      Q_, Q!    | Integer | unsigned long long, native endian (ArgumentError
- *                |         | if the platform has no long long type.)
- *                |         | (Q_ and Q! is available since Ruby 2.1.)
- *                |         |
- *      s_, s!    | Integer | signed short, native endian
- *      i, i_, i! | Integer | signed int, native endian
- *      l_, l!    | Integer | signed long, native endian
- *      q_, q!    | Integer | signed long long, native endian (ArgumentError
- *                |         | if the platform has no long long type.)
- *                |         | (q_ and q! is available since Ruby 2.1.)
- *                |         |
- *      S> L> Q>  | Integer | same as the directives without ">" except
- *      s> l> q>  |         | big endian
- *      S!> I!>   |         | (available since Ruby 1.9.3)
- *      L!> Q!>   |         | "S>" is same as "n"
- *      s!> i!>   |         | "L>" is same as "N"
- *      l!> q!>   |         |
- *                |         |
- *      S< L< Q<  | Integer | same as the directives without "<" except
- *      s< l< q<  |         | little endian
- *      S!< I!<   |         | (available since Ruby 1.9.3)
- *      L!< Q!<   |         | "S<" is same as "v"
- *      s!< i!<   |         | "L<" is same as "V"
- *      l!< q!<   |         |
- *                |         |
- *      n         | Integer | 16-bit unsigned, network (big-endian) byte order
- *      N         | Integer | 32-bit unsigned, network (big-endian) byte order
- *      v         | Integer | 16-bit unsigned, VAX (little-endian) byte order
- *      V         | Integer | 32-bit unsigned, VAX (little-endian) byte order
- *                |         |
- *      U         | Integer | UTF-8 character
- *      w         | Integer | BER-compressed integer (see Array.pack)
- *
- *   Float        |         |
- *   Directive    | Returns | Meaning
- *   -----------------------------------------------------------------
- *      D, d      | Float   | double-precision, native format
- *      F, f      | Float   | single-precision, native format
- *      E         | Float   | double-precision, little-endian byte order
- *      e         | Float   | single-precision, little-endian byte order
- *      G         | Float   | double-precision, network (big-endian) byte order
- *      g         | Float   | single-precision, network (big-endian) byte order
- *
- *   String       |         |
- *   Directive    | Returns | Meaning
- *   -----------------------------------------------------------------
- *      A         | String  | arbitrary binary string (remove trailing nulls and ASCII spaces)
- *      a         | String  | arbitrary binary string
- *      Z         | String  | null-terminated string
- *      B         | String  | bit string (MSB first)
- *      b         | String  | bit string (LSB first)
- *      H         | String  | hex string (high nibble first)
- *      h         | String  | hex string (low nibble first)
- *      u         | String  | UU-encoded string
- *      M         | String  | quoted-printable, MIME encoding (see RFC2045)
- *      m         | String  | base64 encoded string (RFC 2045) (default)
- *                |         | base64 encoded string (RFC 4648) if followed by 0
- *      P         | String  | pointer to a structure (fixed-length string)
- *      p         | String  | pointer to a null-terminated string
- *
- *   Misc.        |         |
- *   Directive    | Returns | Meaning
- *   -----------------------------------------------------------------
- *      @         | ---     | skip to the offset given by the length argument
- *      X         | ---     | skip backward one byte
- *      x         | ---     | skip forward one byte
- */
+/* unpack mode */
+#define UNPACK_ARRAY 0
+#define UNPACK_BLOCK 1
+#define UNPACK_1 2
+
+#define castchar(from) (char)((from) & 0xff)
 
 VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 {
-    static const char hexdigits[] = "0123456789abcdef";
+#define hexdigits ruby_hexdigits
     char *init_s, *s, *send;
     char *p, *pend;
     VALUE ary;
     char type;
-    long len, tmp_len;
+    long len;
+    AVOID_CC_BUG long tmp_len;
     int star;
 #ifdef NATINT_PACK
     int natint;			/* native integer */
 #endif
-    int block_p = rb_block_given_p();
     int signed_p, integer_size, bigendian_p;
+    int mode = (rb_block_given_p() ? UNPACK_BLOCK : UNPACK_ARRAY);
 #define UNPACK_PUSH(item) do {\
 	VALUE item_val = (item);\
-	if (block_p) {\
+	if ((mode) == UNPACK_BLOCK) {\
 	    rb_yield(item_val);\
 	}\
 	else {\
@@ -1363,14 +310,14 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 	}\
     } while (0)
 
-    // StringValue(str);
+    StringValue(str);
     StringValue(fmt);
     init_s = s = RSTRING_PTR(str);
     send = s + RSTRING_LEN(str);
     p = RSTRING_PTR(fmt);
     pend = p + RSTRING_LEN(fmt);
 
-    ary = block_p ? Qnil : rb_ary_new();
+    ary = mode == UNPACK_ARRAY ? rb_ary_new() : Qnil;
     while (p < pend) {
 	int explicit_endian = 0;
 	type = *p++;
@@ -1483,13 +430,14 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 		if (p[-1] == '*' || len > (send - s) * 8)
 		    len = (send - s) * 8;
 		bits = 0;
-		UNPACK_PUSH(bitstr = rb_usascii_str_new(0, len));
+		bitstr = rb_usascii_str_new(0, len);
 		t = RSTRING_PTR(bitstr);
 		for (i=0; i<len; i++) {
 		    if (i & 7) bits >>= 1;
-		    else bits = *s++;
+		    else bits = (unsigned char)*s++;
 		    *t++ = (bits & 1) ? '1' : '0';
 		}
+		UNPACK_PUSH(bitstr);
 	    }
 	    break;
 
@@ -1503,13 +451,14 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 		if (p[-1] == '*' || len > (send - s) * 8)
 		    len = (send - s) * 8;
 		bits = 0;
-		UNPACK_PUSH(bitstr = rb_usascii_str_new(0, len));
+		bitstr = rb_usascii_str_new(0, len);
 		t = RSTRING_PTR(bitstr);
 		for (i=0; i<len; i++) {
 		    if (i & 7) bits <<= 1;
-		    else bits = *s++;
+		    else bits = (unsigned char)*s++;
 		    *t++ = (bits & 128) ? '1' : '0';
 		}
+		UNPACK_PUSH(bitstr);
 	    }
 	    break;
 
@@ -1523,15 +472,16 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 		if (p[-1] == '*' || len > (send - s) * 2)
 		    len = (send - s) * 2;
 		bits = 0;
-		UNPACK_PUSH(bitstr = rb_usascii_str_new(0, len));
+		bitstr = rb_usascii_str_new(0, len);
 		t = RSTRING_PTR(bitstr);
 		for (i=0; i<len; i++) {
 		    if (i & 1)
 			bits >>= 4;
 		    else
-			bits = *s++;
+			bits = (unsigned char)*s++;
 		    *t++ = hexdigits[bits & 15];
 		}
+		UNPACK_PUSH(bitstr);
 	    }
 	    break;
 
@@ -1545,36 +495,30 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 		if (p[-1] == '*' || len > (send - s) * 2)
 		    len = (send - s) * 2;
 		bits = 0;
-		UNPACK_PUSH(bitstr = rb_usascii_str_new(0, len));
+		bitstr = rb_usascii_str_new(0, len);
 		t = RSTRING_PTR(bitstr);
 		for (i=0; i<len; i++) {
 		    if (i & 1)
 			bits <<= 4;
 		    else
-			bits = *s++;
+			bits = (unsigned char)*s++;
 		    *t++ = hexdigits[(bits >> 4) & 15];
 		}
+		UNPACK_PUSH(bitstr);
 	    }
 	    break;
 
 	  case 'c':
-	    PACK_LENGTH_ADJUST_SIZE(sizeof(char));
-	    while (len-- > 0) {
-                int c = *s++;
-                if (c > (char)127) c-=256;
-		UNPACK_PUSH(INT2FIX(c));
-	    }
-	    PACK_ITEM_ADJUST();
-	    break;
+	    signed_p = 1;
+	    integer_size = 1;
+	    bigendian_p = BIGENDIAN_P(); /* not effective */
+	    goto unpack_integer;
 
 	  case 'C':
-	    PACK_LENGTH_ADJUST_SIZE(sizeof(unsigned char));
-	    while (len-- > 0) {
-		unsigned char c = *s++;
-		UNPACK_PUSH(INT2FIX(c));
-	    }
-	    PACK_ITEM_ADJUST();
-	    break;
+	    signed_p = 0;
+	    integer_size = 1;
+	    bigendian_p = BIGENDIAN_P(); /* not effective */
+	    goto unpack_integer;
 
 	  case 's':
 	    signed_p = 1;
@@ -1624,6 +568,18 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 	    bigendian_p = BIGENDIAN_P();
 	    goto unpack_integer;
 
+	  case 'j':
+	    signed_p = 1;
+	    integer_size = sizeof(intptr_t);
+	    bigendian_p = BIGENDIAN_P();
+	    goto unpack_integer;
+
+	  case 'J':
+	    signed_p = 0;
+	    integer_size = sizeof(uintptr_t);
+	    bigendian_p = BIGENDIAN_P();
+	    goto unpack_integer;
+
 	  case 'n':
 	    signed_p = 0;
 	    integer_size = 2;
@@ -1652,144 +608,17 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 	    if (explicit_endian) {
 		bigendian_p = explicit_endian == '>';
 	    }
-
-	    switch (integer_size) {
-#if defined(HAVE_INT16_T) && !defined(FORCE_BIG_PACK)
-	      case SIZEOF_INT16_T:
-		if (signed_p) {
-		    PACK_LENGTH_ADJUST_SIZE(sizeof(int16_t));
-		    while (len-- > 0) {
-			union {
-                            int16_t i;
-                            char a[sizeof(int16_t)];
-                        } v;
-			memcpy(v.a, s, sizeof(int16_t));
-			if (bigendian_p != BIGENDIAN_P()) v.i = swap16(v.i);
-			s += sizeof(int16_t);
-			UNPACK_PUSH(INT2FIX(v.i));
-		    }
-		    PACK_ITEM_ADJUST();
-		}
-		else {
-		    PACK_LENGTH_ADJUST_SIZE(sizeof(uint16_t));
-		    while (len-- > 0) {
-			union {
-                            uint16_t i;
-                            char a[sizeof(uint16_t)];
-                        } v;
-			memcpy(v.a, s, sizeof(uint16_t));
-			if (bigendian_p != BIGENDIAN_P()) v.i = swap16(v.i);
-			s += sizeof(uint16_t);
-			UNPACK_PUSH(INT2FIX(v.i));
-		    }
-		    PACK_ITEM_ADJUST();
-		}
-		break;
-#endif
-
-#if defined(HAVE_INT32_T) && !defined(FORCE_BIG_PACK)
-	      case SIZEOF_INT32_T:
-		if (signed_p) {
-		    PACK_LENGTH_ADJUST_SIZE(sizeof(int32_t));
-		    while (len-- > 0) {
-			union {
-                            int32_t i;
-                            char a[sizeof(int32_t)];
-                        } v;
-			memcpy(v.a, s, sizeof(int32_t));
-			if (bigendian_p != BIGENDIAN_P()) v.i = swap32(v.i);
-			s += sizeof(int32_t);
-			UNPACK_PUSH(INT2NUM(v.i));
-		    }
-		    PACK_ITEM_ADJUST();
-		}
-		else {
-		    PACK_LENGTH_ADJUST_SIZE(sizeof(uint32_t));
-		    while (len-- > 0) {
-			union {
-                            uint32_t i;
-                            char a[sizeof(uint32_t)];
-                        } v;
-			memcpy(v.a, s, sizeof(uint32_t));
-			if (bigendian_p != BIGENDIAN_P()) v.i = swap32(v.i);
-			s += sizeof(uint32_t);
-			UNPACK_PUSH(UINT2NUM(v.i));
-		    }
-		    PACK_ITEM_ADJUST();
-		}
-		break;
-#endif
-
-#if defined(HAVE_INT64_T) && !defined(FORCE_BIG_PACK)
-	      case SIZEOF_INT64_T:
-		if (signed_p) {
-		    PACK_LENGTH_ADJUST_SIZE(sizeof(int64_t));
-		    while (len-- > 0) {
-			union {
-                            int64_t i;
-                            char a[sizeof(int64_t)];
-                        } v;
-			memcpy(v.a, s, sizeof(int64_t));
-			if (bigendian_p != BIGENDIAN_P()) v.i = swap64(v.i);
-			s += sizeof(int64_t);
-			UNPACK_PUSH(INT64toNUM(v.i));
-		    }
-		    PACK_ITEM_ADJUST();
-		}
-		else {
-		    PACK_LENGTH_ADJUST_SIZE(sizeof(uint64_t));
-		    while (len-- > 0) {
-			union {
-                            uint64_t i;
-                            char a[sizeof(uint64_t)];
-                        } v;
-			memcpy(v.a, s, sizeof(uint64_t));
-			if (bigendian_p != BIGENDIAN_P()) v.i = swap64(v.i);
-			s += sizeof(uint64_t);
-			UNPACK_PUSH(UINT64toNUM(v.i));
-		    }
-		    PACK_ITEM_ADJUST();
-		}
-		break;
-#endif
-
-              default:
-                if (integer_size > MAX_INTEGER_PACK_SIZE)
-                    rb_bug("unexpected integer size for pack: %d", integer_size);
-                PACK_LENGTH_ADJUST_SIZE(integer_size);
-                while (len-- > 0) {
-                    union {
-                        unsigned long i[(MAX_INTEGER_PACK_SIZE+SIZEOF_LONG)/SIZEOF_LONG];
-                        char a[(MAX_INTEGER_PACK_SIZE+SIZEOF_LONG)/SIZEOF_LONG*SIZEOF_LONG];
-                    } v;
-                    int num_longs = (integer_size+SIZEOF_LONG)/SIZEOF_LONG;
-                    int i;
-
-                    if (signed_p && (signed char)s[bigendian_p ? 0 : (integer_size-1)] < 0)
-                        memset(v.a, 0xff, sizeof(long)*num_longs);
-                    else
-                        memset(v.a, 0, sizeof(long)*num_longs);
-                    if (bigendian_p)
-                        memcpy(v.a + sizeof(long)*num_longs - integer_size, s, integer_size);
-                    else
-                        memcpy(v.a, s, integer_size);
-                    if (bigendian_p) {
-                        for (i = 0; i < num_longs/2; i++) {
-                            unsigned long t = v.i[i];
-                            v.i[i] = v.i[num_longs-1-i];
-                            v.i[num_longs-1-i] = t;
-                        }
-                    }
-                    if (bigendian_p != BIGENDIAN_P()) {
-                        for (i = 0; i < num_longs; i++)
-                            v.i[i] = swapl(v.i[i]);
-                    }
-                    s += integer_size;
-                    UNPACK_PUSH(rb_big_unpack(v.i, num_longs));
-                }
-                PACK_ITEM_ADJUST();
-		break;
-	    }
+            PACK_LENGTH_ADJUST_SIZE(integer_size);
+            while (len-- > 0) {
+                int flags = bigendian_p ? INTEGER_PACK_BIG_ENDIAN : INTEGER_PACK_LITTLE_ENDIAN;
+                VALUE val;
+                if (signed_p)
+                    flags |= INTEGER_PACK_2COMP;
+                val = rb_integer_unpack(s, integer_size, 1, 0, flags);
+                UNPACK_PUSH(val);
+                s += integer_size;
+            }
+            PACK_ITEM_ADJUST();
             break;
 
 	  case 'f':
@@ -1807,13 +636,11 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 	  case 'e':
 	    PACK_LENGTH_ADJUST_SIZE(sizeof(float));
 	    while (len-- > 0) {
-	        float tmp;
-		FLOAT_CONVWITH(ftmp);
-
-		memcpy(&tmp, s, sizeof(float));
+		FLOAT_CONVWITH(tmp);
+		memcpy(tmp.buf, s, sizeof(float));
 		s += sizeof(float);
-		tmp = VTOHF(tmp,ftmp);
-		UNPACK_PUSH(DBL2NUM((double)tmp));
+		VTOHF(tmp);
+		UNPACK_PUSH(DBL2NUM(tmp.f));
 	    }
 	    PACK_ITEM_ADJUST();
 	    break;
@@ -1821,13 +648,11 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 	  case 'E':
 	    PACK_LENGTH_ADJUST_SIZE(sizeof(double));
 	    while (len-- > 0) {
-		double tmp;
-		DOUBLE_CONVWITH(dtmp);
-
-		memcpy(&tmp, s, sizeof(double));
+		DOUBLE_CONVWITH(tmp);
+		memcpy(tmp.buf, s, sizeof(double));
 		s += sizeof(double);
-		tmp = VTOHD(tmp,dtmp);
-		UNPACK_PUSH(DBL2NUM(tmp));
+		VTOHD(tmp);
+		UNPACK_PUSH(DBL2NUM(tmp.d));
 	    }
 	    PACK_ITEM_ADJUST();
 	    break;
@@ -1847,13 +672,11 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 	  case 'g':
 	    PACK_LENGTH_ADJUST_SIZE(sizeof(float));
 	    while (len-- > 0) {
-	        float tmp;
-		FLOAT_CONVWITH(ftmp);
-
-		memcpy(&tmp, s, sizeof(float));
+		FLOAT_CONVWITH(tmp);
+		memcpy(tmp.buf, s, sizeof(float));
 		s += sizeof(float);
-		tmp = NTOHF(tmp,ftmp);
-		UNPACK_PUSH(DBL2NUM((double)tmp));
+		NTOHF(tmp);
+		UNPACK_PUSH(DBL2NUM(tmp.f));
 	    }
 	    PACK_ITEM_ADJUST();
 	    break;
@@ -1861,13 +684,11 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 	  case 'G':
 	    PACK_LENGTH_ADJUST_SIZE(sizeof(double));
 	    while (len-- > 0) {
-		double tmp;
-		DOUBLE_CONVWITH(dtmp);
-
-		memcpy(&tmp, s, sizeof(double));
+		DOUBLE_CONVWITH(tmp);
+		memcpy(tmp.buf, s, sizeof(double));
 		s += sizeof(double);
-		tmp = NTOHD(tmp,dtmp);
-		UNPACK_PUSH(DBL2NUM(tmp));
+		NTOHD(tmp);
+		UNPACK_PUSH(DBL2NUM(tmp.d));
 	    }
 	    PACK_ITEM_ADJUST();
 	    break;
@@ -1890,12 +711,12 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 		char *ptr = RSTRING_PTR(buf);
 		long total = 0;
 
-		while (s < send && *s > ' ' && *s < 'a') {
+		while (s < send && (unsigned char)*s > ' ' && (unsigned char)*s < 'a') {
 		    long a,b,c,d;
-		    char hunk[4];
+		    char hunk[3];
+
+		    len = ((unsigned char)*s++ - ' ') & 077;
 
-		    hunk[3] = '\0';
-		    len = (*s++ - ' ') & 077;
 		    total += len;
 		    if (total > RSTRING_LEN(buf)) {
 			len -= total - RSTRING_LEN(buf);
@@ -1905,20 +726,20 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 		    while (len > 0) {
 			long mlen = len > 3 ? 3 : len;
 
-			if (s < send && *s >= ' ')
-			    a = (*s++ - ' ') & 077;
+			if (s < send && (unsigned char)*s >= ' ' && (unsigned char)*s < 'a')
+			    a = ((unsigned char)*s++ - ' ') & 077;
 			else
 			    a = 0;
-			if (s < send && *s >= ' ')
-			    b = (*s++ - ' ') & 077;
+			if (s < send && (unsigned char)*s >= ' ' && (unsigned char)*s < 'a')
+			    b = ((unsigned char)*s++ - ' ') & 077;
 			else
 			    b = 0;
-			if (s < send && *s >= ' ')
-			    c = (*s++ - ' ') & 077;
+			if (s < send && (unsigned char)*s >= ' ' && (unsigned char)*s < 'a')
+			    c = ((unsigned char)*s++ - ' ') & 077;
 			else
 			    c = 0;
-			if (s < send && *s >= ' ')
-			    d = (*s++ - ' ') & 077;
+			if (s < send && (unsigned char)*s >= ' ' && (unsigned char)*s < 'a')
+			    d = ((unsigned char)*s++ - ' ') & 077;
 			else
 			    d = 0;
 			hunk[0] = (char)(a << 2 | b >> 4);
@@ -1928,10 +749,10 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 			ptr += mlen;
 			len -= mlen;
 		    }
-		    if (*s == '\r') s++;
-		    if (*s == '\n') s++;
-		    else if (s < send && (s+1 == send || s[1] == '\n'))
-			s += 2;	/* possible checksum byte */
+		    if (s < send && (unsigned char)*s != '\r' && *s != '\n')
+			s++;	/* possible checksum byte */
+		    if (s < send && *s == '\r') s++;
+		    if (s < send && *s == '\n') s++;
 		}
 
 		rb_str_set_len(buf, total);
@@ -2078,18 +899,19 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 		s += sizeof(char *);
 
 		if (t) {
-		    VALUE a, *p, *pend;
+		    VALUE a;
+		    const VALUE *p, *pend;
 
-		    if (!(a = rb_str_associated(str))) {
+		    if (!(a = str_associated(str))) {
 			rb_raise(rb_eArgError, "no associated pointer");
 		    }
-		    p = RARRAY_PTR(a);
+		    p = RARRAY_CONST_PTR(a);
 		    pend = p + RARRAY_LEN(a);
 		    while (p < pend) {
 			if (RB_TYPE_P(*p, T_STRING) && RSTRING_PTR(*p) == t) {
 			    if (len < RSTRING_LEN(*p)) {
 				tmp = rb_tainted_str_new(t, len);
-				rb_str_associate(tmp, a);
+				str_associate(tmp, a);
 			    }
 			    else {
 				tmp = *p;
@@ -2120,12 +942,13 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 		    s += sizeof(char *);
 
 		    if (t) {
-			VALUE a, *p, *pend;
+			VALUE a;
+			const VALUE *p, *pend;
 
-			if (!(a = rb_str_associated(str))) {
+			if (!(a = str_associated(str))) {
 			    rb_raise(rb_eArgError, "no associated pointer");
 			}
-			p = RARRAY_PTR(a);
+			p = RARRAY_CONST_PTR(a);
 			pend = p + RARRAY_LEN(a);
 			while (p < pend) {
 			    if (RB_TYPE_P(*p, T_STRING) && RSTRING_PTR(*p) == t) {
@@ -2145,32 +968,18 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
 
 	  case 'w':
 	    {
-		unsigned long ul = 0;
-		unsigned long ulmask = 0xfeUL << ((sizeof(unsigned long) - 1) * 8);
-
-		while (len > 0 && s < send) {
-		    ul <<= 7;
-		    ul |= (*s & 0x7f);
-		    if (!(*s++ & 0x80)) {
-			UNPACK_PUSH(ULONG2NUM(ul));
-			len--;
-			ul = 0;
-		    }
-		    else if (ul & ulmask) {
-			VALUE big = rb_uint2big(ul);
-			VALUE big128 = rb_uint2big(128);
-			while (s < send) {
-			    big = rb_big_mul(big, big128);
-			    big = rb_big_plus(big, rb_uint2big(*s & 0x7f));
-			    if (!(*s++ & 0x80)) {
-				UNPACK_PUSH(big);
-				len--;
-				ul = 0;
-				break;
-			    }
-			}
-		    }
-		}
+                char *s0 = s;
+                while (len > 0 && s < send) {
+                    if (*s & 0x80) {
+                        s++;
+                    }
+                    else {
+                        s++;
+                        UNPACK_PUSH(rb_integer_unpack(s0, s-s0, 1, 1, INTEGER_PACK_BIG_ENDIAN));
+                        len--;
+                        s0 = s;
+                    }
+                }
 	    }
 	    break;
 
@@ -2185,8 +994,6 @@ VALUE zscan_internal_unpack(VALUE str, VALUE fmt, long* parsed_len)
     return ary;
 }
 
-#define BYTEWIDTH 8
-
 int
 rb_uv_to_utf8(char buf[6], unsigned long uv)
 {