/* pp_pack.c
*
- * Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
- * 2000, 2001, 2002, 2003, 2004, 2005, by Larry Wall and others
+ * Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000,
+ * 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008 by Larry Wall and others
*
* You may distribute under the terms of either the GNU General Public
* License or the Artistic License, as specified in the README file.
* wooden spoon, a short two-pronged fork and some skewers were stowed; and
* hidden at the bottom of the pack in a flat wooden box a dwindling treasure,
* some salt.
+ *
+ * [p.653 of _The Lord of the Rings_, IV/iv: "Of Herbs and Stewed Rabbit"]
*/
/* This file contains pp ("push/pop") functions that
* other pp*.c files for the rest of the pp_ functions.
*/
-
#include "EXTERN.h"
#define PERL_IN_PP_PACK_C
#include "perl.h"
+/* Types used by pack/unpack */
+typedef enum {
+ e_no_len, /* no length */
+ e_number, /* number, [] */
+ e_star /* asterisk */
+} howlen_t;
+
+typedef struct tempsym {
+ const char* patptr; /* current template char */
+ const char* patend; /* one after last char */
+ const char* grpbeg; /* 1st char of ()-group */
+ const char* grpend; /* end of ()-group */
+ I32 code; /* template code (!<>) */
+ I32 length; /* length/repeat count */
+ howlen_t howlen; /* how length is given */
+ int level; /* () nesting level */
+ U32 flags; /* /=4, comma=2, pack=1 */
+ /* and group modifiers */
+ STRLEN strbeg; /* offset of group start */
+ struct tempsym *previous; /* previous group */
+} tempsym_t;
+
+#define TEMPSYM_INIT(symptr, p, e, f) \
+ STMT_START { \
+ (symptr)->patptr = (p); \
+ (symptr)->patend = (e); \
+ (symptr)->grpbeg = NULL; \
+ (symptr)->grpend = NULL; \
+ (symptr)->grpend = NULL; \
+ (symptr)->code = 0; \
+ (symptr)->length = 0; \
+ (symptr)->howlen = e_no_len; \
+ (symptr)->level = 0; \
+ (symptr)->flags = (f); \
+ (symptr)->strbeg = 0; \
+ (symptr)->previous = NULL; \
+ } STMT_END
+
+typedef union {
+ NV nv;
+ U8 bytes[sizeof(NV)];
+} NV_bytes;
+
+#if defined(HAS_LONG_DOUBLE) && defined(USE_LONG_DOUBLE)
+typedef union {
+ long double ld;
+ U8 bytes[sizeof(long double)];
+} ld_bytes;
+#endif
+
+#if PERL_VERSION >= 9
+# define PERL_PACK_CAN_BYTEORDER
+# define PERL_PACK_CAN_SHRIEKSIGN
+#endif
+
+#ifndef CHAR_BIT
+# define CHAR_BIT 8
+#endif
+/* Maximum number of bytes to which a byte can grow due to upgrade */
+#define UTF8_EXPAND 2
+
/*
* Offset for integer pack/unpack.
*
#if U16SIZE > SIZE16 || U32SIZE > SIZE32
# if BYTEORDER == 0x1234 || BYTEORDER == 0x12345678 /* little-endian */
-# define OFF16(p) (char*)(p)
-# define OFF32(p) (char*)(p)
+# define OFF16(p) ((char*)(p))
+# define OFF32(p) ((char*)(p))
# else
# if BYTEORDER == 0x4321 || BYTEORDER == 0x87654321 /* big-endian */
# define OFF16(p) ((char*)(p) + (sizeof(U16) - SIZE16))
# define OFF32(p) ((char*)(p) + (sizeof(U32) - SIZE32))
# else
- }}}} bad cray byte order
+ ++++ bad cray byte order
# endif
# endif
-# define COPY16(s,p) (*(p) = 0, Copy(s, OFF16(p), SIZE16, char))
-# define COPY32(s,p) (*(p) = 0, Copy(s, OFF32(p), SIZE32, char))
-# define COPYNN(s,p,n) (*(p) = 0, Copy(s, (char *)(p), n, char))
-# define CAT16(sv,p) sv_catpvn(sv, OFF16(p), SIZE16)
-# define CAT32(sv,p) sv_catpvn(sv, OFF32(p), SIZE32)
#else
-# define COPY16(s,p) Copy(s, p, SIZE16, char)
-# define COPY32(s,p) Copy(s, p, SIZE32, char)
-# define COPYNN(s,p,n) Copy(s, (char *)(p), n, char)
-# define CAT16(sv,p) sv_catpvn(sv, (char*)(p), SIZE16)
-# define CAT32(sv,p) sv_catpvn(sv, (char*)(p), SIZE32)
+# define OFF16(p) ((char *) (p))
+# define OFF32(p) ((char *) (p))
#endif
+/* Only to be used inside a loop (see the break) */
+#define SHIFT16(utf8, s, strend, p, datumtype) STMT_START { \
+ if (utf8) { \
+ if (!uni_to_bytes(aTHX_ &(s), strend, OFF16(p), SIZE16, datumtype)) break; \
+ } else { \
+ Copy(s, OFF16(p), SIZE16, char); \
+ (s) += SIZE16; \
+ } \
+} STMT_END
+
+/* Only to be used inside a loop (see the break) */
+#define SHIFT32(utf8, s, strend, p, datumtype) STMT_START { \
+ if (utf8) { \
+ if (!uni_to_bytes(aTHX_ &(s), strend, OFF32(p), SIZE32, datumtype)) break; \
+ } else { \
+ Copy(s, OFF32(p), SIZE32, char); \
+ (s) += SIZE32; \
+ } \
+} STMT_END
+
+#define PUSH16(utf8, cur, p) PUSH_BYTES(utf8, cur, OFF16(p), SIZE16)
+#define PUSH32(utf8, cur, p) PUSH_BYTES(utf8, cur, OFF32(p), SIZE32)
+
+/* Only to be used inside a loop (see the break) */
+#define SHIFT_BYTES(utf8, s, strend, buf, len, datumtype) \
+STMT_START { \
+ if (utf8) { \
+ if (!uni_to_bytes(aTHX_ &s, strend, \
+ (char *) (buf), len, datumtype)) break; \
+ } else { \
+ Copy(s, (char *) (buf), len, char); \
+ s += len; \
+ } \
+} STMT_END
+
+#define SHIFT_VAR(utf8, s, strend, var, datumtype) \
+ SHIFT_BYTES(utf8, s, strend, &(var), sizeof(var), datumtype)
+
+#define PUSH_VAR(utf8, aptr, var) \
+ PUSH_BYTES(utf8, aptr, &(var), sizeof(var))
+
/* Avoid stack overflow due to pathological templates. 100 should be plenty. */
#define MAX_SUB_TEMPLATE_LEVEL 100
/* flags (note that type modifiers can also be used as flags!) */
+#define FLAG_WAS_UTF8 0x40
+#define FLAG_PARSE_UTF8 0x20 /* Parse as utf8 */
#define FLAG_UNPACK_ONLY_ONE 0x10
-#define FLAG_UNPACK_DO_UTF8 0x08
+#define FLAG_DO_UTF8 0x08 /* The underlying string is utf8 */
#define FLAG_SLASH 0x04
#define FLAG_COMMA 0x02
#define FLAG_PACK 0x01
STRLEN len;
char *s = SvPV(sv, len);
char *t;
- U32 i = 0;
+
+ PERL_ARGS_ASSERT_MUL128;
if (!strnEQ(s, "0000", 4)) { /* need to grow sv */
- SV *tmpNew = newSVpvn("0000000000", 10);
+ SV * const tmpNew = newSVpvs("0000000000");
sv_catsv(tmpNew, sv);
SvREFCNT_dec(sv); /* free old sv */
while (!*t) /* trailing '\0'? */
t--;
while (t > s) {
- i = ((*t - '0') << 7) + m;
+ const U32 i = ((*t - '0') << 7) + m;
*(t--) = '0' + (char)(i % 10);
m = (char)(i / 10);
}
#define TYPE_IS_SHRIEKING 0x100
#define TYPE_IS_BIG_ENDIAN 0x200
#define TYPE_IS_LITTLE_ENDIAN 0x400
+#define TYPE_IS_PACK 0x800
#define TYPE_ENDIANNESS_MASK (TYPE_IS_BIG_ENDIAN|TYPE_IS_LITTLE_ENDIAN)
-#define TYPE_ENDIANNESS(t) ((t) & TYPE_ENDIANNESS_MASK)
-#define TYPE_NO_ENDIANNESS(t) ((t) & ~TYPE_ENDIANNESS_MASK)
#define TYPE_MODIFIERS(t) ((t) & ~0xFF)
#define TYPE_NO_MODIFIERS(t) ((t) & 0xFF)
-#define ENDIANNESS_ALLOWED_TYPES "sSiIlLqQjJfFdDpP("
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+# define SHRIEKING_ALLOWED_TYPES "sSiIlLxXnNvV@."
+#else
+# define SHRIEKING_ALLOWED_TYPES "sSiIlLxX"
+#endif
+
+#ifndef PERL_PACK_CAN_BYTEORDER
+/* Put "can't" first because it is shorter */
+# define TYPE_ENDIANNESS(t) 0
+# define TYPE_NO_ENDIANNESS(t) (t)
+
+# define ENDIANNESS_ALLOWED_TYPES ""
-#define DO_BO_UNPACK(var, type) \
+# define DO_BO_UNPACK(var, type)
+# define DO_BO_PACK(var, type)
+# define DO_BO_UNPACK_PTR(var, type, pre_cast, post_cast)
+# define DO_BO_PACK_PTR(var, type, pre_cast, post_cast)
+# define DO_BO_UNPACK_N(var, type)
+# define DO_BO_PACK_N(var, type)
+# define DO_BO_UNPACK_P(var)
+# define DO_BO_PACK_P(var)
+# define DO_BO_UNPACK_PC(var)
+# define DO_BO_PACK_PC(var)
+
+#else /* PERL_PACK_CAN_BYTEORDER */
+
+# define TYPE_ENDIANNESS(t) ((t) & TYPE_ENDIANNESS_MASK)
+# define TYPE_NO_ENDIANNESS(t) ((t) & ~TYPE_ENDIANNESS_MASK)
+
+# define ENDIANNESS_ALLOWED_TYPES "sSiIlLqQjJfFdDpP("
+
+# define DO_BO_UNPACK(var, type) \
STMT_START { \
switch (TYPE_ENDIANNESS(datumtype)) { \
case TYPE_IS_BIG_ENDIAN: var = my_betoh ## type (var); break; \
} \
} STMT_END
-#define DO_BO_PACK(var, type) \
+# define DO_BO_PACK(var, type) \
STMT_START { \
switch (TYPE_ENDIANNESS(datumtype)) { \
case TYPE_IS_BIG_ENDIAN: var = my_htobe ## type (var); break; \
} \
} STMT_END
-#define DO_BO_UNPACK_PTR(var, type, pre_cast) \
+# define DO_BO_UNPACK_PTR(var, type, pre_cast, post_cast) \
STMT_START { \
switch (TYPE_ENDIANNESS(datumtype)) { \
case TYPE_IS_BIG_ENDIAN: \
- var = (void *) my_betoh ## type ((pre_cast) var); \
+ var = (post_cast*) my_betoh ## type ((pre_cast) var); \
break; \
case TYPE_IS_LITTLE_ENDIAN: \
- var = (void *) my_letoh ## type ((pre_cast) var); \
+ var = (post_cast *) my_letoh ## type ((pre_cast) var); \
break; \
default: \
break; \
} \
} STMT_END
-#define DO_BO_PACK_PTR(var, type, pre_cast) \
+# define DO_BO_PACK_PTR(var, type, pre_cast, post_cast) \
STMT_START { \
switch (TYPE_ENDIANNESS(datumtype)) { \
case TYPE_IS_BIG_ENDIAN: \
- var = (void *) my_htobe ## type ((pre_cast) var); \
+ var = (post_cast *) my_htobe ## type ((pre_cast) var); \
break; \
case TYPE_IS_LITTLE_ENDIAN: \
- var = (void *) my_htole ## type ((pre_cast) var); \
+ var = (post_cast *) my_htole ## type ((pre_cast) var); \
break; \
default: \
break; \
} \
} STMT_END
-#define BO_CANT_DOIT(action, type) \
+# define BO_CANT_DOIT(action, type) \
STMT_START { \
switch (TYPE_ENDIANNESS(datumtype)) { \
case TYPE_IS_BIG_ENDIAN: \
} \
} STMT_END
-#if PTRSIZE == INTSIZE
-# define DO_BO_UNPACK_P(var) DO_BO_UNPACK_PTR(var, i, int)
-# define DO_BO_PACK_P(var) DO_BO_PACK_PTR(var, i, int)
-#elif PTRSIZE == LONGSIZE
-# define DO_BO_UNPACK_P(var) DO_BO_UNPACK_PTR(var, l, long)
-# define DO_BO_PACK_P(var) DO_BO_PACK_PTR(var, l, long)
-#else
-# define DO_BO_UNPACK_P(var) BO_CANT_DOIT(unpack, pointer)
-# define DO_BO_PACK_P(var) BO_CANT_DOIT(pack, pointer)
-#endif
+# if PTRSIZE == INTSIZE
+# define DO_BO_UNPACK_P(var) DO_BO_UNPACK_PTR(var, i, int, void)
+# define DO_BO_PACK_P(var) DO_BO_PACK_PTR(var, i, int, void)
+# define DO_BO_UNPACK_PC(var) DO_BO_UNPACK_PTR(var, i, int, char)
+# define DO_BO_PACK_PC(var) DO_BO_PACK_PTR(var, i, int, char)
+# elif PTRSIZE == LONGSIZE
+# if LONGSIZE < IVSIZE && IVSIZE == 8
+# define DO_BO_UNPACK_P(var) DO_BO_UNPACK_PTR(var, 64, IV, void)
+# define DO_BO_PACK_P(var) DO_BO_PACK_PTR(var, 64, IV, void)
+# define DO_BO_UNPACK_PC(var) DO_BO_UNPACK_PTR(var, 64, IV, char)
+# define DO_BO_PACK_PC(var) DO_BO_PACK_PTR(var, 64, IV, char)
+# else
+# define DO_BO_UNPACK_P(var) DO_BO_UNPACK_PTR(var, l, IV, void)
+# define DO_BO_PACK_P(var) DO_BO_PACK_PTR(var, l, IV, void)
+# define DO_BO_UNPACK_PC(var) DO_BO_UNPACK_PTR(var, l, IV, char)
+# define DO_BO_PACK_PC(var) DO_BO_PACK_PTR(var, l, IV, char)
+# endif
+# elif PTRSIZE == IVSIZE
+# define DO_BO_UNPACK_P(var) DO_BO_UNPACK_PTR(var, l, IV, void)
+# define DO_BO_PACK_P(var) DO_BO_PACK_PTR(var, l, IV, void)
+# define DO_BO_UNPACK_PC(var) DO_BO_UNPACK_PTR(var, l, IV, char)
+# define DO_BO_PACK_PC(var) DO_BO_PACK_PTR(var, l, IV, char)
+# else
+# define DO_BO_UNPACK_P(var) BO_CANT_DOIT(unpack, pointer)
+# define DO_BO_PACK_P(var) BO_CANT_DOIT(pack, pointer)
+# define DO_BO_UNPACK_PC(var) BO_CANT_DOIT(unpack, pointer)
+# define DO_BO_PACK_PC(var) BO_CANT_DOIT(pack, pointer)
+# endif
-#if defined(my_htolen) && defined(my_letohn) && \
+# if defined(my_htolen) && defined(my_letohn) && \
defined(my_htoben) && defined(my_betohn)
-# define DO_BO_UNPACK_N(var, type) \
+# define DO_BO_UNPACK_N(var, type) \
STMT_START { \
switch (TYPE_ENDIANNESS(datumtype)) { \
case TYPE_IS_BIG_ENDIAN: my_betohn(&var, sizeof(type)); break;\
} \
} STMT_END
-# define DO_BO_PACK_N(var, type) \
+# define DO_BO_PACK_N(var, type) \
STMT_START { \
switch (TYPE_ENDIANNESS(datumtype)) { \
case TYPE_IS_BIG_ENDIAN: my_htoben(&var, sizeof(type)); break;\
default: break; \
} \
} STMT_END
-#else
-# define DO_BO_UNPACK_N(var, type) BO_CANT_DOIT(unpack, type)
-# define DO_BO_PACK_N(var, type) BO_CANT_DOIT(pack, type)
-#endif
+# else
+# define DO_BO_UNPACK_N(var, type) BO_CANT_DOIT(unpack, type)
+# define DO_BO_PACK_N(var, type) BO_CANT_DOIT(pack, type)
+# endif
+
+#endif /* PERL_PACK_CAN_BYTEORDER */
#define PACK_SIZE_CANNOT_CSUM 0x80
-#define PACK_SIZE_CANNOT_ONLY_ONE 0x40
+#define PACK_SIZE_UNPREDICTABLE 0x40 /* Not a fixed size element */
#define PACK_SIZE_MASK 0x3F
-
-struct packsize_t {
- const char *array;
- int first;
- int size;
-};
-
-#define PACK_SIZE_NORMAL 0
-#define PACK_SIZE_SHRIEKING 1
-
/* These tables are regenerated by genpacksizetables.pl (and then hand pasted
in). You're unlikely ever to need to regenerate them. */
+
+#if TYPE_IS_SHRIEKING != 0x100
+ ++++shriek offset should be 256
+#endif
+
+typedef U8 packprops_t;
#if 'J'-'I' == 1
/* ASCII */
-unsigned char size_normal[53] = {
- /* C */ sizeof(unsigned char),
+STATIC const packprops_t packprops[512] = {
+ /* normal */
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0,
+ /* C */ sizeof(unsigned char),
#if defined(HAS_LONG_DOUBLE) && defined(USE_LONG_DOUBLE)
- /* D */ LONG_DOUBLESIZE,
+ /* D */ LONG_DOUBLESIZE,
#else
- 0,
+ 0,
#endif
- 0,
- /* F */ NVSIZE,
- 0, 0,
- /* I */ sizeof(unsigned int),
- /* J */ UVSIZE,
- 0,
- /* L */ SIZE32,
- 0,
- /* N */ SIZE32,
- 0, 0,
+ 0,
+ /* F */ NVSIZE,
+ 0, 0,
+ /* I */ sizeof(unsigned int),
+ /* J */ UVSIZE,
+ 0,
+ /* L */ SIZE32,
+ 0,
+ /* N */ SIZE32,
+ 0, 0,
#if defined(HAS_QUAD)
- /* Q */ sizeof(Uquad_t),
+ /* Q */ sizeof(Uquad_t),
#else
- 0,
+ 0,
#endif
- 0,
- /* S */ SIZE16,
- 0,
- /* U */ sizeof(char),
- /* V */ SIZE32,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* c */ sizeof(char),
- /* d */ sizeof(double),
- 0,
- /* f */ sizeof(float),
- 0, 0,
- /* i */ sizeof(int),
- /* j */ IVSIZE,
- 0,
- /* l */ SIZE32,
- 0,
- /* n */ SIZE16,
- 0,
- /* p */ sizeof(char *) | PACK_SIZE_CANNOT_ONLY_ONE | PACK_SIZE_CANNOT_CSUM,
+ 0,
+ /* S */ SIZE16,
+ 0,
+ /* U */ sizeof(char) | PACK_SIZE_UNPREDICTABLE,
+ /* V */ SIZE32,
+ /* W */ sizeof(unsigned char) | PACK_SIZE_UNPREDICTABLE,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* c */ sizeof(char),
+ /* d */ sizeof(double),
+ 0,
+ /* f */ sizeof(float),
+ 0, 0,
+ /* i */ sizeof(int),
+ /* j */ IVSIZE,
+ 0,
+ /* l */ SIZE32,
+ 0,
+ /* n */ SIZE16,
+ 0,
+ /* p */ sizeof(char *) | PACK_SIZE_CANNOT_CSUM,
#if defined(HAS_QUAD)
- /* q */ sizeof(Quad_t),
+ /* q */ sizeof(Quad_t),
#else
- 0,
+ 0,
#endif
- 0,
- /* s */ SIZE16,
- 0, 0,
- /* v */ SIZE16,
- /* w */ sizeof(char) | PACK_SIZE_CANNOT_CSUM
-};
-unsigned char size_shrieking[46] = {
- /* I */ sizeof(unsigned int),
- 0, 0,
- /* L */ sizeof(unsigned long),
- 0,
- /* N */ SIZE32,
- 0, 0, 0, 0,
- /* S */ sizeof(unsigned short),
- 0, 0,
- /* V */ SIZE32,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* i */ sizeof(int),
- 0, 0,
- /* l */ sizeof(long),
- 0,
- /* n */ SIZE16,
- 0, 0, 0, 0,
- /* s */ sizeof(short),
- 0, 0,
- /* v */ SIZE16
-};
-struct packsize_t packsize[2] = {
- {size_normal, 67, 53},
- {size_shrieking, 73, 46}
+ 0,
+ /* s */ SIZE16,
+ 0, 0,
+ /* v */ SIZE16,
+ /* w */ sizeof(char) | PACK_SIZE_UNPREDICTABLE | PACK_SIZE_CANNOT_CSUM,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0,
+ /* shrieking */
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* I */ sizeof(unsigned int),
+ 0, 0,
+ /* L */ sizeof(unsigned long),
+ 0,
+#if defined(PERL_PACK_CAN_SHRIEKSIGN)
+ /* N */ SIZE32,
+#else
+ 0,
+#endif
+ 0, 0, 0, 0,
+ /* S */ sizeof(unsigned short),
+ 0, 0,
+#if defined(PERL_PACK_CAN_SHRIEKSIGN)
+ /* V */ SIZE32,
+#else
+ 0,
+#endif
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0,
+ /* i */ sizeof(int),
+ 0, 0,
+ /* l */ sizeof(long),
+ 0,
+#if defined(PERL_PACK_CAN_SHRIEKSIGN)
+ /* n */ SIZE16,
+#else
+ 0,
+#endif
+ 0, 0, 0, 0,
+ /* s */ sizeof(short),
+ 0, 0,
+#if defined(PERL_PACK_CAN_SHRIEKSIGN)
+ /* v */ SIZE16,
+#else
+ 0,
+#endif
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0
};
#else
/* EBCDIC (or bust) */
-unsigned char size_normal[99] = {
- /* c */ sizeof(char),
- /* d */ sizeof(double),
- 0,
- /* f */ sizeof(float),
- 0, 0,
- /* i */ sizeof(int),
- 0, 0, 0, 0, 0, 0, 0,
- /* j */ IVSIZE,
- 0,
- /* l */ SIZE32,
- 0,
- /* n */ SIZE16,
- 0,
- /* p */ sizeof(char *) | PACK_SIZE_CANNOT_ONLY_ONE | PACK_SIZE_CANNOT_CSUM,
+STATIC const packprops_t packprops[512] = {
+ /* normal */
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0,
+ /* c */ sizeof(char),
+ /* d */ sizeof(double),
+ 0,
+ /* f */ sizeof(float),
+ 0, 0,
+ /* i */ sizeof(int),
+ 0, 0, 0, 0, 0, 0, 0,
+ /* j */ IVSIZE,
+ 0,
+ /* l */ SIZE32,
+ 0,
+ /* n */ SIZE16,
+ 0,
+ /* p */ sizeof(char *) | PACK_SIZE_CANNOT_CSUM,
#if defined(HAS_QUAD)
- /* q */ sizeof(Quad_t),
+ /* q */ sizeof(Quad_t),
#else
- 0,
+ 0,
#endif
- 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* s */ SIZE16,
- 0, 0,
- /* v */ SIZE16,
- /* w */ sizeof(char) | PACK_SIZE_CANNOT_CSUM,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0, 0,
- /* C */ sizeof(unsigned char),
+ 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* s */ SIZE16,
+ 0, 0,
+ /* v */ SIZE16,
+ /* w */ sizeof(char) | PACK_SIZE_UNPREDICTABLE | PACK_SIZE_CANNOT_CSUM,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* C */ sizeof(unsigned char),
#if defined(HAS_LONG_DOUBLE) && defined(USE_LONG_DOUBLE)
- /* D */ LONG_DOUBLESIZE,
+ /* D */ LONG_DOUBLESIZE,
#else
- 0,
+ 0,
#endif
- 0,
- /* F */ NVSIZE,
- 0, 0,
- /* I */ sizeof(unsigned int),
- 0, 0, 0, 0, 0, 0, 0,
- /* J */ UVSIZE,
- 0,
- /* L */ SIZE32,
- 0,
- /* N */ SIZE32,
- 0, 0,
+ 0,
+ /* F */ NVSIZE,
+ 0, 0,
+ /* I */ sizeof(unsigned int),
+ 0, 0, 0, 0, 0, 0, 0,
+ /* J */ UVSIZE,
+ 0,
+ /* L */ SIZE32,
+ 0,
+ /* N */ SIZE32,
+ 0, 0,
#if defined(HAS_QUAD)
- /* Q */ sizeof(Uquad_t),
+ /* Q */ sizeof(Uquad_t),
#else
- 0,
+ 0,
#endif
- 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* S */ SIZE16,
- 0,
- /* U */ sizeof(char),
- /* V */ SIZE32
-};
-unsigned char size_shrieking[93] = {
- /* i */ sizeof(int),
- 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* l */ sizeof(long),
- 0,
- /* n */ SIZE16,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* s */ sizeof(short),
- 0, 0,
- /* v */ SIZE16,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* I */ sizeof(unsigned int),
- 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* L */ sizeof(unsigned long),
- 0,
- /* N */ SIZE32,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- /* S */ sizeof(unsigned short),
- 0, 0,
- /* V */ SIZE32
-};
-struct packsize_t packsize[2] = {
- {size_normal, 131, 99},
- {size_shrieking, 137, 93}
+ 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* S */ SIZE16,
+ 0,
+ /* U */ sizeof(char) | PACK_SIZE_UNPREDICTABLE,
+ /* V */ SIZE32,
+ /* W */ sizeof(unsigned char) | PACK_SIZE_UNPREDICTABLE,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* shrieking */
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* i */ sizeof(int),
+ 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* l */ sizeof(long),
+ 0,
+#if defined(PERL_PACK_CAN_SHRIEKSIGN)
+ /* n */ SIZE16,
+#else
+ 0,
+#endif
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* s */ sizeof(short),
+ 0, 0,
+#if defined(PERL_PACK_CAN_SHRIEKSIGN)
+ /* v */ SIZE16,
+#else
+ 0,
+#endif
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0,
+ /* I */ sizeof(unsigned int),
+ 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* L */ sizeof(unsigned long),
+ 0,
+#if defined(PERL_PACK_CAN_SHRIEKSIGN)
+ /* N */ SIZE32,
+#else
+ 0,
+#endif
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ /* S */ sizeof(unsigned short),
+ 0, 0,
+#if defined(PERL_PACK_CAN_SHRIEKSIGN)
+ /* V */ SIZE32,
+#else
+ 0,
+#endif
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
};
#endif
+STATIC U8
+uni_to_byte(pTHX_ const char **s, const char *end, I32 datumtype)
+{
+ STRLEN retlen;
+ UV val = utf8n_to_uvchr((U8 *) *s, end-*s, &retlen,
+ ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
+ /* We try to process malformed UTF-8 as much as possible (preferrably with
+ warnings), but these two mean we make no progress in the string and
+ might enter an infinite loop */
+ if (retlen == (STRLEN) -1 || retlen == 0)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in '%c' format in unpack",
+ (int) TYPE_NO_MODIFIERS(datumtype));
+ if (val >= 0x100) {
+ Perl_ck_warner(aTHX_ packWARN(WARN_UNPACK),
+ "Character in '%c' format wrapped in unpack",
+ (int) TYPE_NO_MODIFIERS(datumtype));
+ val &= 0xff;
+ }
+ *s += retlen;
+ return (U8)val;
+}
+
+#define SHIFT_BYTE(utf8, s, strend, datumtype) ((utf8) ? \
+ uni_to_byte(aTHX_ &(s), (strend), (datumtype)) : \
+ *(U8 *)(s)++)
+
+STATIC bool
+uni_to_bytes(pTHX_ const char **s, const char *end, const char *buf, int buf_len, I32 datumtype)
+{
+ UV val;
+ STRLEN retlen;
+ const char *from = *s;
+ int bad = 0;
+ const U32 flags = ckWARN(WARN_UTF8) ?
+ UTF8_CHECK_ONLY : (UTF8_CHECK_ONLY | UTF8_ALLOW_ANY);
+ for (;buf_len > 0; buf_len--) {
+ if (from >= end) return FALSE;
+ val = utf8n_to_uvchr((U8 *) from, end-from, &retlen, flags);
+ if (retlen == (STRLEN) -1 || retlen == 0) {
+ from += UTF8SKIP(from);
+ bad |= 1;
+ } else from += retlen;
+ if (val >= 0x100) {
+ bad |= 2;
+ val &= 0xff;
+ }
+ *(U8 *)buf++ = (U8)val;
+ }
+ /* We have enough characters for the buffer. Did we have problems ? */
+ if (bad) {
+ if (bad & 1) {
+ /* Rewalk the string fragment while warning */
+ const char *ptr;
+ const int flags = ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY;
+ for (ptr = *s; ptr < from; ptr += UTF8SKIP(ptr)) {
+ if (ptr >= end) break;
+ utf8n_to_uvuni((U8 *) ptr, end-ptr, &retlen, flags);
+ }
+ if (from > end) from = end;
+ }
+ if ((bad & 2))
+ Perl_ck_warner(aTHX_ packWARN(datumtype & TYPE_IS_PACK ?
+ WARN_PACK : WARN_UNPACK),
+ "Character(s) in '%c' format wrapped in %s",
+ (int) TYPE_NO_MODIFIERS(datumtype),
+ datumtype & TYPE_IS_PACK ? "pack" : "unpack");
+ }
+ *s = from;
+ return TRUE;
+}
+
+STATIC bool
+next_uni_uu(pTHX_ const char **s, const char *end, I32 *out)
+{
+ dVAR;
+ STRLEN retlen;
+ const UV val = utf8n_to_uvchr((U8 *) *s, end-*s, &retlen, UTF8_CHECK_ONLY);
+ if (val >= 0x100 || !ISUUCHAR(val) ||
+ retlen == (STRLEN) -1 || retlen == 0) {
+ *out = 0;
+ return FALSE;
+ }
+ *out = PL_uudmap[val] & 077;
+ *s += retlen;
+ return TRUE;
+}
+
+STATIC char *
+S_bytes_to_uni(const U8 *start, STRLEN len, char *dest) {
+ const U8 * const end = start + len;
+
+ PERL_ARGS_ASSERT_BYTES_TO_UNI;
+
+ while (start < end) {
+ const UV uv = NATIVE_TO_ASCII(*start);
+ if (UNI_IS_INVARIANT(uv))
+ *dest++ = (char)(U8)UTF_TO_NATIVE(uv);
+ else {
+ *dest++ = (char)(U8)UTF8_EIGHT_BIT_HI(uv);
+ *dest++ = (char)(U8)UTF8_EIGHT_BIT_LO(uv);
+ }
+ start++;
+ }
+ return dest;
+}
+
+#define PUSH_BYTES(utf8, cur, buf, len) \
+STMT_START { \
+ if (utf8) \
+ (cur) = bytes_to_uni((U8 *) buf, len, (cur)); \
+ else { \
+ Copy(buf, cur, len, char); \
+ (cur) += (len); \
+ } \
+} STMT_END
+
+#define GROWING(utf8, cat, start, cur, in_len) \
+STMT_START { \
+ STRLEN glen = (in_len); \
+ if (utf8) glen *= UTF8_EXPAND; \
+ if ((cur) + glen >= (start) + SvLEN(cat)) { \
+ (start) = sv_exp_grow(cat, glen); \
+ (cur) = (start) + SvCUR(cat); \
+ } \
+} STMT_END
+
+#define PUSH_GROWING_BYTES(utf8, cat, start, cur, buf, in_len) \
+STMT_START { \
+ const STRLEN glen = (in_len); \
+ STRLEN gl = glen; \
+ if (utf8) gl *= UTF8_EXPAND; \
+ if ((cur) + gl >= (start) + SvLEN(cat)) { \
+ *cur = '\0'; \
+ SvCUR_set((cat), (cur) - (start)); \
+ (start) = sv_exp_grow(cat, gl); \
+ (cur) = (start) + SvCUR(cat); \
+ } \
+ PUSH_BYTES(utf8, cur, buf, glen); \
+} STMT_END
+
+#define PUSH_BYTE(utf8, s, byte) \
+STMT_START { \
+ if (utf8) { \
+ const U8 au8 = (byte); \
+ (s) = bytes_to_uni(&au8, 1, (s)); \
+ } else *(U8 *)(s)++ = (byte); \
+} STMT_END
+
+/* Only to be used inside a loop (see the break) */
+#define NEXT_UNI_VAL(val, cur, str, end, utf8_flags) \
+STMT_START { \
+ STRLEN retlen; \
+ if (str >= end) break; \
+ val = utf8n_to_uvchr((U8 *) str, end-str, &retlen, utf8_flags); \
+ if (retlen == (STRLEN) -1 || retlen == 0) { \
+ *cur = '\0'; \
+ Perl_croak(aTHX_ "Malformed UTF-8 string in pack"); \
+ } \
+ str += retlen; \
+} STMT_END
+
+static const char *_action( const tempsym_t* symptr )
+{
+ return (const char *)(( symptr->flags & FLAG_PACK ) ? "pack" : "unpack");
+}
/* Returns the sizeof() struct described by pat */
STATIC I32
-S_measure_struct(pTHX_ register tempsym_t* symptr)
+S_measure_struct(pTHX_ tempsym_t* symptr)
{
- register I32 len = 0;
- register I32 total = 0;
- int star;
+ I32 total = 0;
- register int size;
+ PERL_ARGS_ASSERT_MEASURE_STRUCT;
while (next_symbol(symptr)) {
- int which = (symptr->code & TYPE_IS_SHRIEKING)
- ? PACK_SIZE_SHRIEKING : PACK_SIZE_NORMAL;
- int offset
- = TYPE_NO_MODIFIERS(symptr->code) - packsize[which].first;
-
- switch( symptr->howlen ){
- case e_no_len:
- case e_number:
- len = symptr->length;
- break;
- case e_star:
+ I32 len;
+ int size;
+
+ switch (symptr->howlen) {
+ case e_star:
Perl_croak(aTHX_ "Within []-length '*' not allowed in %s",
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ _action( symptr ) );
break;
+ default:
+ /* e_no_len and e_number */
+ len = symptr->length;
+ break;
}
- if ((offset >= 0) && (offset < packsize[which].size))
- size = packsize[which].array[offset] & PACK_SIZE_MASK;
- else
- size = 0;
-
+ size = packprops[TYPE_NO_ENDIANNESS(symptr->code)] & PACK_SIZE_MASK;
if (!size) {
+ int star;
/* endianness doesn't influence the size of a type */
switch(TYPE_NO_ENDIANNESS(symptr->code)) {
default:
Perl_croak(aTHX_ "Invalid type '%c' in %s",
(int)TYPE_NO_MODIFIERS(symptr->code),
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ _action( symptr ) );
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+ case '.' | TYPE_IS_SHRIEKING:
+ case '@' | TYPE_IS_SHRIEKING:
+#endif
case '@':
+ case '.':
case '/':
case 'U': /* XXXX Is it correct? */
case 'w':
case 'u':
Perl_croak(aTHX_ "Within []-length '%c' not allowed in %s",
- (int)symptr->code,
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ (int) TYPE_NO_MODIFIERS(symptr->code),
+ _action( symptr ) );
case '%':
size = 0;
break;
case '(':
- {
- tempsym_t savsym = *symptr;
- symptr->patptr = savsym.grpbeg;
- symptr->patend = savsym.grpend;
- /* XXXX Theoretically, we need to measure many times at
- different positions, since the subexpression may contain
- alignment commands, but be not of aligned length.
- Need to detect this and croak(). */
- size = measure_struct(symptr);
- *symptr = savsym;
- break;
- }
+ {
+ tempsym_t savsym = *symptr;
+ symptr->patptr = savsym.grpbeg;
+ symptr->patend = savsym.grpend;
+ /* XXXX Theoretically, we need to measure many times at
+ different positions, since the subexpression may contain
+ alignment commands, but be not of aligned length.
+ Need to detect this and croak(). */
+ size = measure_struct(symptr);
+ *symptr = savsym;
+ break;
+ }
case 'X' | TYPE_IS_SHRIEKING:
/* XXXX Is this useful? Then need to treat MEASURE_BACKWARDS.
*/
case 'X':
size = -1;
if (total < len)
- Perl_croak(aTHX_ "'X' outside of string in %s",
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ Perl_croak(aTHX_ "'X' outside of string in %s", _action( symptr ) );
break;
case 'x' | TYPE_IS_SHRIEKING:
if (!len) /* Avoid division by 0 */
case 'A':
case 'Z':
case 'a':
- case 'c':
- case 'C':
size = 1;
break;
case 'B':
/* locate matching closing parenthesis or bracket
* returns char pointer to char after match, or NULL
*/
-STATIC char *
-S_group_end(pTHX_ register char *patptr, register char *patend, char ender)
+STATIC const char *
+S_group_end(pTHX_ register const char *patptr, register const char *patend, char ender)
{
+ PERL_ARGS_ASSERT_GROUP_END;
+
while (patptr < patend) {
- char c = *patptr++;
+ const char c = *patptr++;
if (isSPACE(c))
continue;
/* Convert unsigned decimal number to binary.
* Expects a pointer to the first digit and address of length variable
* Advances char pointer to 1st non-digit char and returns number
- */
-STATIC char *
-S_get_num(pTHX_ register char *patptr, I32 *lenptr )
+ */
+STATIC const char *
+S_get_num(pTHX_ register const char *patptr, I32 *lenptr )
{
I32 len = *patptr++ - '0';
+
+ PERL_ARGS_ASSERT_GET_NUM;
+
while (isDIGIT(*patptr)) {
if (len >= 0x7FFFFFFF/10)
Perl_croak(aTHX_ "pack/unpack repeat count overflow");
* locates next template code and count
*/
STATIC bool
-S_next_symbol(pTHX_ register tempsym_t* symptr )
+S_next_symbol(pTHX_ tempsym_t* symptr )
{
- register char* patptr = symptr->patptr;
- register char* patend = symptr->patend;
+ const char* patptr = symptr->patptr;
+ const char* const patend = symptr->patend;
+
+ PERL_ARGS_ASSERT_NEXT_SYMBOL;
symptr->flags &= ~FLAG_SLASH;
if (patptr < patend)
patptr++;
} else {
- /* We should have found a template code */
+ /* We should have found a template code */
I32 code = *patptr++ & 0xFF;
U32 inherited_modifiers = 0;
if (((symptr->flags & FLAG_COMMA) == 0) && ckWARN(WARN_UNPACK)){
symptr->flags |= FLAG_COMMA;
Perl_warner(aTHX_ packWARN(WARN_UNPACK),
- "Invalid type ',' in %s",
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ "Invalid type ',' in %s", _action( symptr ) );
}
continue;
}
-
+
/* for '(', skip to ')' */
- if (code == '(') {
+ if (code == '(') {
if( isDIGIT(*patptr) || *patptr == '*' || *patptr == '[' )
Perl_croak(aTHX_ "()-group starts with a count in %s",
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ _action( symptr ) );
symptr->grpbeg = patptr;
patptr = 1 + ( symptr->grpend = group_end(patptr, patend, ')') );
if( symptr->level >= MAX_SUB_TEMPLATE_LEVEL )
Perl_croak(aTHX_ "Too deeply nested ()-groups in %s",
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ _action( symptr ) );
}
/* look for group modifiers to inherit */
/* look for modifiers */
while (patptr < patend) {
const char *allowed;
- I32 modifier = 0;
+ I32 modifier;
switch (*patptr) {
case '!':
modifier = TYPE_IS_SHRIEKING;
- allowed = "sSiIlLxXnNvV";
+ allowed = SHRIEKING_ALLOWED_TYPES;
break;
+#ifdef PERL_PACK_CAN_BYTEORDER
case '>':
modifier = TYPE_IS_BIG_ENDIAN;
allowed = ENDIANNESS_ALLOWED_TYPES;
modifier = TYPE_IS_LITTLE_ENDIAN;
allowed = ENDIANNESS_ALLOWED_TYPES;
break;
+#endif /* PERL_PACK_CAN_BYTEORDER */
default:
+ allowed = "";
+ modifier = 0;
break;
}
if (!strchr(allowed, TYPE_NO_MODIFIERS(code)))
Perl_croak(aTHX_ "'%c' allowed only after types %s in %s", *patptr,
- allowed, symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ allowed, _action( symptr ) );
if (TYPE_ENDIANNESS(code | modifier) == TYPE_ENDIANNESS_MASK)
Perl_croak(aTHX_ "Can't use both '<' and '>' after type '%c' in %s",
- (int) TYPE_NO_MODIFIERS(code),
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ (int) TYPE_NO_MODIFIERS(code), _action( symptr ) );
else if (TYPE_ENDIANNESS(code | modifier | inherited_modifiers) ==
TYPE_ENDIANNESS_MASK)
Perl_croak(aTHX_ "Can't use '%c' in a group with different byte-order in %s",
- *patptr, symptr->flags & FLAG_PACK ? "pack" : "unpack" );
-
- if (ckWARN(WARN_UNPACK)) {
- if (code & modifier)
- Perl_warner(aTHX_ packWARN(WARN_UNPACK),
- "Duplicate modifier '%c' after '%c' in %s",
- *patptr, (int) TYPE_NO_MODIFIERS(code),
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ *patptr, _action( symptr ) );
+
+ if ((code & modifier)) {
+ Perl_ck_warner(aTHX_ packWARN(WARN_UNPACK),
+ "Duplicate modifier '%c' after '%c' in %s",
+ *patptr, (int) TYPE_NO_MODIFIERS(code),
+ _action( symptr ) );
}
code |= modifier;
/* inherit modifiers */
code |= inherited_modifiers;
- /* look for count and/or / */
+ /* look for count and/or / */
if (patptr < patend) {
if (isDIGIT(*patptr)) {
patptr = get_num( patptr, &symptr->length );
symptr->howlen = e_star;
} else if (*patptr == '[') {
- char* lenptr = ++patptr;
+ const char* lenptr = ++patptr;
symptr->howlen = e_number;
patptr = group_end( patptr, patend, ']' ) + 1;
/* what kind of [] is it? */
lenptr = get_num( lenptr, &symptr->length );
if( *lenptr != ']' )
Perl_croak(aTHX_ "Malformed integer in [] in %s",
- symptr->flags & FLAG_PACK ? "pack" : "unpack");
+ _action( symptr ) );
} else {
tempsym_t savsym = *symptr;
symptr->patend = patptr-1;
if (patptr < patend &&
(isDIGIT(*patptr) || *patptr == '*' || *patptr == '['))
Perl_croak(aTHX_ "'/' does not take a repeat count in %s",
- symptr->flags & FLAG_PACK ? "pack" : "unpack" );
+ _action( symptr ) );
}
break;
}
}
symptr->code = code;
- symptr->patptr = patptr;
+ symptr->patptr = patptr;
return TRUE;
}
}
- symptr->patptr = patptr;
+ symptr->patptr = patptr;
return FALSE;
}
/*
-=for apidoc unpack_str
-
-The engine implementing unpack() Perl function. Note: parameters strbeg, new_s
-and ocnt are not used. This call should not be used, use unpackstring instead.
-
-=cut */
-
-I32
-Perl_unpack_str(pTHX_ char *pat, register char *patend, register char *s, char *strbeg, char *strend, char **new_s, I32 ocnt, U32 flags)
+ There is no way to cleanly handle the case where we should process the
+ string per byte in its upgraded form while it's really in downgraded form
+ (e.g. estimates like strend-s as an upper bound for the number of
+ characters left wouldn't work). So if we foresee the need of this
+ (pattern starts with U or contains U0), we want to work on the encoded
+ version of the string. Users are advised to upgrade their pack string
+ themselves if they need to do a lot of unpacks like this on it
+*/
+STATIC bool
+need_utf8(const char *pat, const char *patend)
{
- tempsym_t sym = { 0 };
- sym.patptr = pat;
- sym.patend = patend;
- sym.flags = flags;
+ bool first = TRUE;
+
+ PERL_ARGS_ASSERT_NEED_UTF8;
+
+ while (pat < patend) {
+ if (pat[0] == '#') {
+ pat++;
+ pat = (const char *) memchr(pat, '\n', patend-pat);
+ if (!pat) return FALSE;
+ } else if (pat[0] == 'U') {
+ if (first || pat[1] == '0') return TRUE;
+ } else first = FALSE;
+ pat++;
+ }
+ return FALSE;
+}
- return unpack_rec(&sym, s, s, strend, NULL );
+STATIC char
+first_symbol(const char *pat, const char *patend) {
+ PERL_ARGS_ASSERT_FIRST_SYMBOL;
+
+ while (pat < patend) {
+ if (pat[0] != '#') return pat[0];
+ pat++;
+ pat = (const char *) memchr(pat, '\n', patend-pat);
+ if (!pat) return 0;
+ pat++;
+ }
+ return 0;
}
/*
=cut */
I32
-Perl_unpackstring(pTHX_ char *pat, register char *patend, register char *s, char *strend, U32 flags)
+Perl_unpackstring(pTHX_ const char *pat, const char *patend, const char *s, const char *strend, U32 flags)
{
- tempsym_t sym = { 0 };
- sym.patptr = pat;
- sym.patend = patend;
- sym.flags = flags;
+ tempsym_t sym;
+
+ PERL_ARGS_ASSERT_UNPACKSTRING;
+
+ if (flags & FLAG_DO_UTF8) flags |= FLAG_WAS_UTF8;
+ else if (need_utf8(pat, patend)) {
+ /* We probably should try to avoid this in case a scalar context call
+ wouldn't get to the "U0" */
+ STRLEN len = strend - s;
+ s = (char *) bytes_to_utf8((U8 *) s, &len);
+ SAVEFREEPV(s);
+ strend = s + len;
+ flags |= FLAG_DO_UTF8;
+ }
+
+ if (first_symbol(pat, patend) != 'U' && (flags & FLAG_DO_UTF8))
+ flags |= FLAG_PARSE_UTF8;
+
+ TEMPSYM_INIT(&sym, pat, patend, flags);
return unpack_rec(&sym, s, s, strend, NULL );
}
-STATIC
-I32
-S_unpack_rec(pTHX_ register tempsym_t* symptr, register char *s, char *strbeg, char *strend, char **new_s )
+STATIC I32
+S_unpack_rec(pTHX_ tempsym_t* symptr, const char *s, const char *strbeg, const char *strend, const char **new_s )
{
- dSP;
- I32 datumtype;
- register I32 len = 0;
- register I32 bits = 0;
- register char *str;
- SV *sv;
- I32 start_sp_offset = SP - PL_stack_base;
+ dVAR; dSP;
+ SV *sv = NULL;
+ const I32 start_sp_offset = SP - PL_stack_base;
howlen_t howlen;
-
- /* These must not be in registers: */
- I16 ai16;
- U16 au16;
- I32 ai32;
- U32 au32;
-#ifdef HAS_QUAD
- Quad_t aquad;
- Uquad_t auquad;
-#endif
-#if SHORTSIZE != SIZE16
- short ashort;
- unsigned short aushort;
-#endif
- int aint;
- unsigned int auint;
- long along;
-#if LONGSIZE != SIZE32
- unsigned long aulong;
-#endif
- char *aptr;
- float afloat;
- double adouble;
-#if defined(HAS_LONG_DOUBLE) && defined(USE_LONG_DOUBLE)
- long double aldouble;
-#endif
- IV aiv;
- UV auv;
- NV anv;
-
I32 checksum = 0;
UV cuv = 0;
NV cdouble = 0.0;
- const int bits_in_uv = 8 * sizeof(cuv);
- char* strrelbeg = s;
+ const int bits_in_uv = CHAR_BIT * sizeof(cuv);
bool beyond = FALSE;
- bool unpack_only_one = (symptr->flags & FLAG_UNPACK_ONLY_ONE) != 0;
+ bool explicit_length;
+ const bool unpack_only_one = (symptr->flags & FLAG_UNPACK_ONLY_ONE) != 0;
+ bool utf8 = (symptr->flags & FLAG_PARSE_UTF8) ? 1 : 0;
+
+ PERL_ARGS_ASSERT_UNPACK_REC;
+
+ symptr->strbeg = s - strbeg;
while (next_symbol(symptr)) {
- datumtype = symptr->code;
+ packprops_t props;
+ I32 len;
+ I32 datumtype = symptr->code;
/* do first one only unless in list context
- / is implemented by unpacking the count, then poping it from the
+ / is implemented by unpacking the count, then popping it from the
stack, so must check that we're not in the middle of a / */
if ( unpack_only_one
&& (SP - PL_stack_base == start_sp_offset + 1)
&& (datumtype != '/') ) /* XXX can this be omitted */
break;
- switch( howlen = symptr->howlen ){
- case e_no_len:
- case e_number:
- len = symptr->length;
+ switch (howlen = symptr->howlen) {
+ case e_star:
+ len = strend - strbeg; /* long enough */
break;
- case e_star:
- len = strend - strbeg; /* long enough */
+ default:
+ /* e_no_len and e_number */
+ len = symptr->length;
break;
}
+ explicit_length = TRUE;
redo_switch:
beyond = s >= strend;
- {
- int which = (symptr->code & TYPE_IS_SHRIEKING)
- ? PACK_SIZE_SHRIEKING : PACK_SIZE_NORMAL;
- int offset = TYPE_NO_MODIFIERS(datumtype) - packsize[which].first;
-
- if (offset >= 0 && offset < packsize[which].size) {
- /* Data about this template letter */
- unsigned char data = packsize[which].array[offset];
-
- if (data) {
- /* data nonzero means we can process this letter. */
- long size = data & PACK_SIZE_MASK;
- long howmany = (strend - s) / size;
- if (len > howmany)
- len = howmany;
-
- if (!checksum || (data & PACK_SIZE_CANNOT_CSUM)) {
- if (len && unpack_only_one &&
- !(data & PACK_SIZE_CANNOT_ONLY_ONE))
- len = 1;
- EXTEND(SP, len);
- EXTEND_MORTAL(len);
- }
- }
+
+ props = packprops[TYPE_NO_ENDIANNESS(datumtype)];
+ if (props) {
+ /* props nonzero means we can process this letter. */
+ const long size = props & PACK_SIZE_MASK;
+ const long howmany = (strend - s) / size;
+ if (len > howmany)
+ len = howmany;
+
+ if (!checksum || (props & PACK_SIZE_CANNOT_CSUM)) {
+ if (len && unpack_only_one) len = 1;
+ EXTEND(SP, len);
+ EXTEND_MORTAL(len);
}
}
+
switch(TYPE_NO_ENDIANNESS(datumtype)) {
default:
Perl_croak(aTHX_ "Invalid type '%c' in unpack", (int)TYPE_NO_MODIFIERS(datumtype) );
break;
case '(':
{
- char *ss = s; /* Move from register */
tempsym_t savsym = *symptr;
- U32 group_modifiers = TYPE_MODIFIERS(datumtype & ~symptr->flags);
+ const U32 group_modifiers = TYPE_MODIFIERS(datumtype & ~symptr->flags);
symptr->flags |= group_modifiers;
symptr->patend = savsym.grpend;
+ symptr->previous = &savsym;
symptr->level++;
PUTBACK;
+ if (len && unpack_only_one) len = 1;
while (len--) {
symptr->patptr = savsym.grpbeg;
- unpack_rec(symptr, ss, strbeg, strend, &ss );
- if (ss == strend && savsym.howlen == e_star)
+ if (utf8) symptr->flags |= FLAG_PARSE_UTF8;
+ else symptr->flags &= ~FLAG_PARSE_UTF8;
+ unpack_rec(symptr, s, strbeg, strend, &s);
+ if (s == strend && savsym.howlen == e_star)
break; /* No way to continue */
}
SPAGAIN;
- s = ss;
- symptr->flags &= ~group_modifiers;
- savsym.flags = symptr->flags;
+ savsym.flags = symptr->flags & ~group_modifiers;
*symptr = savsym;
break;
}
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+ case '.' | TYPE_IS_SHRIEKING:
+#endif
+ case '.': {
+ const char *from;
+ SV *sv;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+ const bool u8 = utf8 && !(datumtype & TYPE_IS_SHRIEKING);
+#else /* PERL_PACK_CAN_SHRIEKSIGN */
+ const bool u8 = utf8;
+#endif
+ if (howlen == e_star) from = strbeg;
+ else if (len <= 0) from = s;
+ else {
+ tempsym_t *group = symptr;
+
+ while (--len && group) group = group->previous;
+ from = group ? strbeg + group->strbeg : strbeg;
+ }
+ sv = from <= s ?
+ newSVuv( u8 ? (UV) utf8_length((const U8*)from, (const U8*)s) : (UV) (s-from)) :
+ newSViv(-(u8 ? (IV) utf8_length((const U8*)s, (const U8*)from) : (IV) (from-s)));
+ mXPUSHs(sv);
+ break;
+ }
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+ case '@' | TYPE_IS_SHRIEKING:
+#endif
case '@':
- if (len > strend - strrelbeg)
- Perl_croak(aTHX_ "'@' outside of string in unpack");
- s = strrelbeg + len;
+ s = strbeg + symptr->strbeg;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+ if (utf8 && !(datumtype & TYPE_IS_SHRIEKING))
+#else /* PERL_PACK_CAN_SHRIEKSIGN */
+ if (utf8)
+#endif
+ {
+ while (len > 0) {
+ if (s >= strend)
+ Perl_croak(aTHX_ "'@' outside of string in unpack");
+ s += UTF8SKIP(s);
+ len--;
+ }
+ if (s > strend)
+ Perl_croak(aTHX_ "'@' outside of string with malformed UTF-8 in unpack");
+ } else {
+ if (strend-s < len)
+ Perl_croak(aTHX_ "'@' outside of string in unpack");
+ s += len;
+ }
break;
case 'X' | TYPE_IS_SHRIEKING:
if (!len) /* Avoid division by 0 */
len = 1;
- len = (s - strbeg) % len;
+ if (utf8) {
+ const char *hop, *last;
+ I32 l = len;
+ hop = last = strbeg;
+ while (hop < s) {
+ hop += UTF8SKIP(hop);
+ if (--l == 0) {
+ last = hop;
+ l = len;
+ }
+ }
+ if (last > s)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in unpack");
+ s = last;
+ break;
+ }
+ len = (s - strbeg) % len;
/* FALL THROUGH */
case 'X':
- if (len > s - strbeg)
- Perl_croak(aTHX_ "'X' outside of string in unpack" );
- s -= len;
+ if (utf8) {
+ while (len > 0) {
+ if (s <= strbeg)
+ Perl_croak(aTHX_ "'X' outside of string in unpack");
+ while (--s, UTF8_IS_CONTINUATION(*s)) {
+ if (s <= strbeg)
+ Perl_croak(aTHX_ "'X' outside of string in unpack");
+ }
+ len--;
+ }
+ } else {
+ if (len > s - strbeg)
+ Perl_croak(aTHX_ "'X' outside of string in unpack" );
+ s -= len;
+ }
break;
- case 'x' | TYPE_IS_SHRIEKING:
+ case 'x' | TYPE_IS_SHRIEKING: {
+ I32 ai32;
if (!len) /* Avoid division by 0 */
len = 1;
- aint = (s - strbeg) % len;
- if (aint) /* Other portable ways? */
- len = len - aint;
- else
- len = 0;
+ if (utf8) ai32 = utf8_length((U8 *) strbeg, (U8 *) s) % len;
+ else ai32 = (s - strbeg) % len;
+ if (ai32 == 0) break;
+ len -= ai32;
+ }
/* FALL THROUGH */
case 'x':
- if (len > strend - s)
- Perl_croak(aTHX_ "'x' outside of string in unpack");
- s += len;
+ if (utf8) {
+ while (len>0) {
+ if (s >= strend)
+ Perl_croak(aTHX_ "'x' outside of string in unpack");
+ s += UTF8SKIP(s);
+ len--;
+ }
+ } else {
+ if (len > strend - s)
+ Perl_croak(aTHX_ "'x' outside of string in unpack");
+ s += len;
+ }
break;
case '/':
Perl_croak(aTHX_ "'/' must follow a numeric type in unpack");
case 'A':
case 'Z':
case 'a':
- if (len > strend - s)
- len = strend - s;
- if (checksum)
- goto uchar_checksum;
- sv = newSVpvn(s, len);
- if (len > 0 && (datumtype == 'A' || datumtype == 'Z')) {
- aptr = s; /* borrow register */
- if (datumtype == 'Z') { /* 'Z' strips stuff after first null */
- s = SvPVX(sv);
- while (*s)
- s++;
- if (howlen == e_star) /* exact for 'Z*' */
- len = s - SvPVX(sv) + 1;
+ if (checksum) {
+ /* Preliminary length estimate is assumed done in 'W' */
+ if (len > strend - s) len = strend - s;
+ goto W_checksum;
+ }
+ if (utf8) {
+ I32 l;
+ const char *hop;
+ for (l=len, hop=s; l>0; l--, hop += UTF8SKIP(hop)) {
+ if (hop >= strend) {
+ if (hop > strend)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in unpack");
+ break;
+ }
}
- else { /* 'A' strips both nulls and spaces */
- s = SvPVX(sv) + len - 1;
- while (s >= SvPVX(sv) && (!*s || isSPACE(*s)))
- s--;
- *++s = '\0';
+ if (hop > strend)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in unpack");
+ len = hop - s;
+ } else if (len > strend - s)
+ len = strend - s;
+
+ if (datumtype == 'Z') {
+ /* 'Z' strips stuff after first null */
+ const char *ptr, *end;
+ end = s + len;
+ for (ptr = s; ptr < end; ptr++) if (*ptr == 0) break;
+ sv = newSVpvn(s, ptr-s);
+ if (howlen == e_star) /* exact for 'Z*' */
+ len = ptr-s + (ptr != strend ? 1 : 0);
+ } else if (datumtype == 'A') {
+ /* 'A' strips both nulls and spaces */
+ const char *ptr;
+ if (utf8 && (symptr->flags & FLAG_WAS_UTF8)) {
+ for (ptr = s+len-1; ptr >= s; ptr--)
+ if (*ptr != 0 && !UTF8_IS_CONTINUATION(*ptr) &&
+ !is_utf8_space((U8 *) ptr)) break;
+ if (ptr >= s) ptr += UTF8SKIP(ptr);
+ else ptr++;
+ if (ptr > s+len)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in unpack");
+ } else {
+ for (ptr = s+len-1; ptr >= s; ptr--)
+ if (*ptr != 0 && !isSPACE(*ptr)) break;
+ ptr++;
}
- SvCUR_set(sv, s - SvPVX(sv));
- s = aptr; /* unborrow register */
+ sv = newSVpvn(s, ptr-s);
+ } else sv = newSVpvn(s, len);
+
+ if (utf8) {
+ SvUTF8_on(sv);
+ /* Undo any upgrade done due to need_utf8() */
+ if (!(symptr->flags & FLAG_WAS_UTF8))
+ sv_utf8_downgrade(sv, 0);
}
+ mXPUSHs(sv);
s += len;
- XPUSHs(sv_2mortal(sv));
break;
case 'B':
- case 'b':
+ case 'b': {
+ char *str;
if (howlen == e_star || len > (strend - s) * 8)
len = (strend - s) * 8;
if (checksum) {
- if (!PL_bitcount) {
- Newz(601, PL_bitcount, 256, char);
- for (bits = 1; bits < 256; bits++) {
- if (bits & 1) PL_bitcount[bits]++;
- if (bits & 2) PL_bitcount[bits]++;
- if (bits & 4) PL_bitcount[bits]++;
- if (bits & 8) PL_bitcount[bits]++;
- if (bits & 16) PL_bitcount[bits]++;
- if (bits & 32) PL_bitcount[bits]++;
- if (bits & 64) PL_bitcount[bits]++;
- if (bits & 128) PL_bitcount[bits]++;
+ if (utf8)
+ while (len >= 8 && s < strend) {
+ cuv += PL_bitcount[uni_to_byte(aTHX_ &s, strend, datumtype)];
+ len -= 8;
}
- }
- while (len >= 8) {
- cuv += PL_bitcount[*(unsigned char*)s++];
- len -= 8;
- }
- if (len) {
- bits = *s;
- if (datumtype == 'b') {
+ else
+ while (len >= 8) {
+ cuv += PL_bitcount[*(U8 *)s++];
+ len -= 8;
+ }
+ if (len && s < strend) {
+ U8 bits;
+ bits = SHIFT_BYTE(utf8, s, strend, datumtype);
+ if (datumtype == 'b')
while (len-- > 0) {
if (bits & 1) cuv++;
bits >>= 1;
}
- }
- else {
+ else
while (len-- > 0) {
- if (bits & 128) cuv++;
+ if (bits & 0x80) cuv++;
bits <<= 1;
}
- }
}
break;
}
- sv = NEWSV(35, len + 1);
- SvCUR_set(sv, len);
+
+ sv = sv_2mortal(newSV(len ? len : 1));
SvPOK_on(sv);
str = SvPVX(sv);
if (datumtype == 'b') {
- aint = len;
- for (len = 0; len < aint; len++) {
- if (len & 7) /*SUPPRESS 595*/
- bits >>= 1;
- else
- bits = *s++;
- *str++ = '0' + (bits & 1);
+ U8 bits = 0;
+ const I32 ai32 = len;
+ for (len = 0; len < ai32; len++) {
+ if (len & 7) bits >>= 1;
+ else if (utf8) {
+ if (s >= strend) break;
+ bits = uni_to_byte(aTHX_ &s, strend, datumtype);
+ } else bits = *(U8 *) s++;
+ *str++ = bits & 1 ? '1' : '0';
}
- }
- else {
- aint = len;
- for (len = 0; len < aint; len++) {
- if (len & 7)
- bits <<= 1;
- else
- bits = *s++;
- *str++ = '0' + ((bits & 128) != 0);
+ } else {
+ U8 bits = 0;
+ const I32 ai32 = len;
+ for (len = 0; len < ai32; len++) {
+ if (len & 7) bits <<= 1;
+ else if (utf8) {
+ if (s >= strend) break;
+ bits = uni_to_byte(aTHX_ &s, strend, datumtype);
+ } else bits = *(U8 *) s++;
+ *str++ = bits & 0x80 ? '1' : '0';
}
}
*str = '\0';
- XPUSHs(sv_2mortal(sv));
+ SvCUR_set(sv, str - SvPVX_const(sv));
+ XPUSHs(sv);
break;
+ }
case 'H':
- case 'h':
+ case 'h': {
+ char *str = NULL;
+ /* Preliminary length estimate, acceptable for utf8 too */
if (howlen == e_star || len > (strend - s) * 2)
len = (strend - s) * 2;
- sv = NEWSV(35, len + 1);
- SvCUR_set(sv, len);
- SvPOK_on(sv);
- str = SvPVX(sv);
+ if (!checksum) {
+ sv = sv_2mortal(newSV(len ? len : 1));
+ SvPOK_on(sv);
+ str = SvPVX(sv);
+ }
if (datumtype == 'h') {
- aint = len;
- for (len = 0; len < aint; len++) {
- if (len & 1)
- bits >>= 4;
- else
- bits = *s++;
- *str++ = PL_hexdigit[bits & 15];
+ U8 bits = 0;
+ I32 ai32 = len;
+ for (len = 0; len < ai32; len++) {
+ if (len & 1) bits >>= 4;
+ else if (utf8) {
+ if (s >= strend) break;
+ bits = uni_to_byte(aTHX_ &s, strend, datumtype);
+ } else bits = * (U8 *) s++;
+ if (!checksum)
+ *str++ = PL_hexdigit[bits & 15];
}
- }
- else {
- aint = len;
- for (len = 0; len < aint; len++) {
- if (len & 1)
- bits <<= 4;
- else
- bits = *s++;
- *str++ = PL_hexdigit[(bits >> 4) & 15];
+ } else {
+ U8 bits = 0;
+ const I32 ai32 = len;
+ for (len = 0; len < ai32; len++) {
+ if (len & 1) bits <<= 4;
+ else if (utf8) {
+ if (s >= strend) break;
+ bits = uni_to_byte(aTHX_ &s, strend, datumtype);
+ } else bits = *(U8 *) s++;
+ if (!checksum)
+ *str++ = PL_hexdigit[(bits >> 4) & 15];
}
}
- *str = '\0';
- XPUSHs(sv_2mortal(sv));
+ if (!checksum) {
+ *str = '\0';
+ SvCUR_set(sv, str - SvPVX_const(sv));
+ XPUSHs(sv);
+ }
break;
+ }
+ case 'C':
+ if (len == 0) {
+ if (explicit_length)
+ /* Switch to "character" mode */
+ utf8 = (symptr->flags & FLAG_DO_UTF8) ? 1 : 0;
+ break;
+ }
+ /* FALL THROUGH */
case 'c':
- while (len-- > 0) {
- aint = *s++;
- if (aint >= 128) /* fake up signed chars */
+ while (len-- > 0 && s < strend) {
+ int aint;
+ if (utf8)
+ {
+ STRLEN retlen;
+ aint = utf8n_to_uvchr((U8 *) s, strend-s, &retlen,
+ ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
+ if (retlen == (STRLEN) -1 || retlen == 0)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in unpack");
+ s += retlen;
+ }
+ else
+ aint = *(U8 *)(s)++;
+ if (aint >= 128 && datumtype != 'C') /* fake up signed chars */
aint -= 256;
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((IV)aint)));
- }
+ if (!checksum)
+ mPUSHi(aint);
else if (checksum > bits_in_uv)
cdouble += (NV)aint;
else
cuv += aint;
}
break;
- case 'C':
- unpack_C: /* unpack U will jump here if not UTF-8 */
- if (len == 0) {
- symptr->flags &= ~FLAG_UNPACK_DO_UTF8;
- break;
- }
- if (checksum) {
- uchar_checksum:
- while (len-- > 0) {
- auint = *s++ & 255;
- cuv += auint;
+ case 'W':
+ W_checksum:
+ if (utf8) {
+ while (len-- > 0 && s < strend) {
+ STRLEN retlen;
+ const UV val = utf8n_to_uvchr((U8 *) s, strend-s, &retlen,
+ ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANY);
+ if (retlen == (STRLEN) -1 || retlen == 0)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in unpack");
+ s += retlen;
+ if (!checksum)
+ mPUSHu(val);
+ else if (checksum > bits_in_uv)
+ cdouble += (NV) val;
+ else
+ cuv += val;
}
- }
- else {
+ } else if (!checksum)
while (len-- > 0) {
- auint = *s++ & 255;
- PUSHs(sv_2mortal(newSViv((IV)auint)));
- }
+ const U8 ch = *(U8 *) s++;
+ mPUSHu(ch);
}
+ else if (checksum > bits_in_uv)
+ while (len-- > 0) cdouble += (NV) *(U8 *) s++;
+ else
+ while (len-- > 0) cuv += *(U8 *) s++;
break;
case 'U':
if (len == 0) {
- symptr->flags |= FLAG_UNPACK_DO_UTF8;
+ if (explicit_length) {
+ /* Switch to "bytes in UTF-8" mode */
+ if (symptr->flags & FLAG_DO_UTF8) utf8 = 0;
+ else
+ /* Should be impossible due to the need_utf8() test */
+ Perl_croak(aTHX_ "U0 mode on a byte string");
+ }
break;
}
- if ((symptr->flags & FLAG_UNPACK_DO_UTF8) == 0)
- goto unpack_C;
+ if (len > strend - s) len = strend - s;
+ if (!checksum) {
+ if (len && unpack_only_one) len = 1;
+ EXTEND(SP, len);
+ EXTEND_MORTAL(len);
+ }
while (len-- > 0 && s < strend) {
- STRLEN alen;
- auint = NATIVE_TO_UNI(utf8n_to_uvchr((U8*)s, strend - s, &alen, ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANYUV));
- along = alen;
- s += along;
- if (!checksum) {
- PUSHs(sv_2mortal(newSVuv((UV)auint)));
+ STRLEN retlen;
+ UV auv;
+ if (utf8) {
+ U8 result[UTF8_MAXLEN];
+ const char *ptr = s;
+ STRLEN len;
+ /* Bug: warns about bad utf8 even if we are short on bytes
+ and will break out of the loop */
+ if (!uni_to_bytes(aTHX_ &ptr, strend, (char *) result, 1,
+ 'U'))
+ break;
+ len = UTF8SKIP(result);
+ if (!uni_to_bytes(aTHX_ &ptr, strend,
+ (char *) &result[1], len-1, 'U')) break;
+ auv = utf8n_to_uvuni(result, len, &retlen, ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANYUV);
+ s = ptr;
+ } else {
+ auv = utf8n_to_uvuni((U8*)s, strend - s, &retlen, ckWARN(WARN_UTF8) ? 0 : UTF8_ALLOW_ANYUV);
+ if (retlen == (STRLEN) -1 || retlen == 0)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in unpack");
+ s += retlen;
}
+ if (!checksum)
+ mPUSHu(auv);
else if (checksum > bits_in_uv)
- cdouble += (NV)auint;
+ cdouble += (NV) auv;
else
- cuv += auint;
+ cuv += auv;
}
break;
case 's' | TYPE_IS_SHRIEKING:
#if SHORTSIZE != SIZE16
while (len-- > 0) {
- COPYNN(s, &ashort, sizeof(short));
+ short ashort;
+ SHIFT_VAR(utf8, s, strend, ashort, datumtype);
DO_BO_UNPACK(ashort, s);
- s += sizeof(short);
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((IV)ashort)));
- }
+ if (!checksum)
+ mPUSHi(ashort);
else if (checksum > bits_in_uv)
cdouble += (NV)ashort;
else
#endif
case 's':
while (len-- > 0) {
- COPY16(s, &ai16);
+ I16 ai16;
+
+#if U16SIZE > SIZE16
+ ai16 = 0;
+#endif
+ SHIFT16(utf8, s, strend, &ai16, datumtype);
DO_BO_UNPACK(ai16, 16);
#if U16SIZE > SIZE16
if (ai16 > 32767)
ai16 -= 65536;
#endif
- s += SIZE16;
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((IV)ai16)));
- }
+ if (!checksum)
+ mPUSHi(ai16);
else if (checksum > bits_in_uv)
cdouble += (NV)ai16;
else
case 'S' | TYPE_IS_SHRIEKING:
#if SHORTSIZE != SIZE16
while (len-- > 0) {
- COPYNN(s, &aushort, sizeof(unsigned short));
+ unsigned short aushort;
+ SHIFT_VAR(utf8, s, strend, aushort, datumtype);
DO_BO_UNPACK(aushort, s);
- s += sizeof(unsigned short);
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((UV)aushort)));
- }
+ if (!checksum)
+ mPUSHu(aushort);
else if (checksum > bits_in_uv)
cdouble += (NV)aushort;
else
case 'n':
case 'S':
while (len-- > 0) {
- COPY16(s, &au16);
+ U16 au16;
+#if U16SIZE > SIZE16
+ au16 = 0;
+#endif
+ SHIFT16(utf8, s, strend, &au16, datumtype);
DO_BO_UNPACK(au16, 16);
- s += SIZE16;
#ifdef HAS_NTOHS
if (datumtype == 'n')
au16 = PerlSock_ntohs(au16);
if (datumtype == 'v')
au16 = vtohs(au16);
#endif
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((UV)au16)));
- }
+ if (!checksum)
+ mPUSHu(au16);
else if (checksum > bits_in_uv)
- cdouble += (NV)au16;
+ cdouble += (NV) au16;
else
cuv += au16;
}
break;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
case 'v' | TYPE_IS_SHRIEKING:
case 'n' | TYPE_IS_SHRIEKING:
while (len-- > 0) {
- COPY16(s, &ai16);
- s += SIZE16;
-#ifdef HAS_NTOHS
+ I16 ai16;
+# if U16SIZE > SIZE16
+ ai16 = 0;
+# endif
+ SHIFT16(utf8, s, strend, &ai16, datumtype);
+# ifdef HAS_NTOHS
if (datumtype == ('n' | TYPE_IS_SHRIEKING))
- ai16 = (I16)PerlSock_ntohs((U16)ai16);
-#endif
-#ifdef HAS_VTOHS
+ ai16 = (I16) PerlSock_ntohs((U16) ai16);
+# endif /* HAS_NTOHS */
+# ifdef HAS_VTOHS
if (datumtype == ('v' | TYPE_IS_SHRIEKING))
- ai16 = (I16)vtohs((U16)ai16);
-#endif
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((IV)ai16)));
- }
+ ai16 = (I16) vtohs((U16) ai16);
+# endif /* HAS_VTOHS */
+ if (!checksum)
+ mPUSHi(ai16);
else if (checksum > bits_in_uv)
- cdouble += (NV)ai16;
+ cdouble += (NV) ai16;
else
cuv += ai16;
}
break;
+#endif /* PERL_PACK_CAN_SHRIEKSIGN */
case 'i':
case 'i' | TYPE_IS_SHRIEKING:
while (len-- > 0) {
- Copy(s, &aint, 1, int);
+ int aint;
+ SHIFT_VAR(utf8, s, strend, aint, datumtype);
DO_BO_UNPACK(aint, i);
- s += sizeof(int);
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((IV)aint)));
- }
+ if (!checksum)
+ mPUSHi(aint);
else if (checksum > bits_in_uv)
cdouble += (NV)aint;
else
case 'I':
case 'I' | TYPE_IS_SHRIEKING:
while (len-- > 0) {
- Copy(s, &auint, 1, unsigned int);
+ unsigned int auint;
+ SHIFT_VAR(utf8, s, strend, auint, datumtype);
DO_BO_UNPACK(auint, i);
- s += sizeof(unsigned int);
- if (!checksum) {
- PUSHs(sv_2mortal(newSVuv((UV)auint)));
- }
+ if (!checksum)
+ mPUSHu(auint);
else if (checksum > bits_in_uv)
cdouble += (NV)auint;
else
break;
case 'j':
while (len-- > 0) {
- Copy(s, &aiv, 1, IV);
+ IV aiv;
+ SHIFT_VAR(utf8, s, strend, aiv, datumtype);
#if IVSIZE == INTSIZE
DO_BO_UNPACK(aiv, i);
#elif IVSIZE == LONGSIZE
DO_BO_UNPACK(aiv, l);
#elif defined(HAS_QUAD) && IVSIZE == U64SIZE
DO_BO_UNPACK(aiv, 64);
+#else
+ Perl_croak(aTHX_ "'j' not supported on this platform");
#endif
- s += IVSIZE;
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv(aiv)));
- }
+ if (!checksum)
+ mPUSHi(aiv);
else if (checksum > bits_in_uv)
cdouble += (NV)aiv;
else
break;
case 'J':
while (len-- > 0) {
- Copy(s, &auv, 1, UV);
-#if UVSIZE == INTSIZE
+ UV auv;
+ SHIFT_VAR(utf8, s, strend, auv, datumtype);
+#if IVSIZE == INTSIZE
DO_BO_UNPACK(auv, i);
-#elif UVSIZE == LONGSIZE
+#elif IVSIZE == LONGSIZE
DO_BO_UNPACK(auv, l);
-#elif defined(HAS_QUAD) && UVSIZE == U64SIZE
+#elif defined(HAS_QUAD) && IVSIZE == U64SIZE
DO_BO_UNPACK(auv, 64);
+#else
+ Perl_croak(aTHX_ "'J' not supported on this platform");
#endif
- s += UVSIZE;
- if (!checksum) {
- PUSHs(sv_2mortal(newSVuv(auv)));
- }
+ if (!checksum)
+ mPUSHu(auv);
else if (checksum > bits_in_uv)
cdouble += (NV)auv;
else
case 'l' | TYPE_IS_SHRIEKING:
#if LONGSIZE != SIZE32
while (len-- > 0) {
- COPYNN(s, &along, sizeof(long));
+ long along;
+ SHIFT_VAR(utf8, s, strend, along, datumtype);
DO_BO_UNPACK(along, l);
- s += sizeof(long);
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((IV)along)));
- }
+ if (!checksum)
+ mPUSHi(along);
else if (checksum > bits_in_uv)
cdouble += (NV)along;
else
#endif
case 'l':
while (len-- > 0) {
- COPY32(s, &ai32);
+ I32 ai32;
+#if U32SIZE > SIZE32
+ ai32 = 0;
+#endif
+ SHIFT32(utf8, s, strend, &ai32, datumtype);
DO_BO_UNPACK(ai32, 32);
#if U32SIZE > SIZE32
- if (ai32 > 2147483647)
- ai32 -= 4294967296;
+ if (ai32 > 2147483647) ai32 -= 4294967296;
#endif
- s += SIZE32;
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((IV)ai32)));
- }
+ if (!checksum)
+ mPUSHi(ai32);
else if (checksum > bits_in_uv)
cdouble += (NV)ai32;
else
case 'L' | TYPE_IS_SHRIEKING:
#if LONGSIZE != SIZE32
while (len-- > 0) {
- COPYNN(s, &aulong, sizeof(unsigned long));
+ unsigned long aulong;
+ SHIFT_VAR(utf8, s, strend, aulong, datumtype);
DO_BO_UNPACK(aulong, l);
- s += sizeof(unsigned long);
- if (!checksum) {
- PUSHs(sv_2mortal(newSVuv((UV)aulong)));
- }
+ if (!checksum)
+ mPUSHu(aulong);
else if (checksum > bits_in_uv)
cdouble += (NV)aulong;
else
case 'N':
case 'L':
while (len-- > 0) {
- COPY32(s, &au32);
+ U32 au32;
+#if U32SIZE > SIZE32
+ au32 = 0;
+#endif
+ SHIFT32(utf8, s, strend, &au32, datumtype);
DO_BO_UNPACK(au32, 32);
- s += SIZE32;
#ifdef HAS_NTOHL
if (datumtype == 'N')
au32 = PerlSock_ntohl(au32);
if (datumtype == 'V')
au32 = vtohl(au32);
#endif
- if (!checksum) {
- PUSHs(sv_2mortal(newSVuv((UV)au32)));
- }
- else if (checksum > bits_in_uv)
- cdouble += (NV)au32;
- else
- cuv += au32;
+ if (!checksum)
+ mPUSHu(au32);
+ else if (checksum > bits_in_uv)
+ cdouble += (NV)au32;
+ else
+ cuv += au32;
}
break;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
case 'V' | TYPE_IS_SHRIEKING:
case 'N' | TYPE_IS_SHRIEKING:
while (len-- > 0) {
- COPY32(s, &ai32);
- s += SIZE32;
-#ifdef HAS_NTOHL
+ I32 ai32;
+# if U32SIZE > SIZE32
+ ai32 = 0;
+# endif
+ SHIFT32(utf8, s, strend, &ai32, datumtype);
+# ifdef HAS_NTOHL
if (datumtype == ('N' | TYPE_IS_SHRIEKING))
ai32 = (I32)PerlSock_ntohl((U32)ai32);
-#endif
-#ifdef HAS_VTOHL
+# endif
+# ifdef HAS_VTOHL
if (datumtype == ('V' | TYPE_IS_SHRIEKING))
ai32 = (I32)vtohl((U32)ai32);
-#endif
- if (!checksum) {
- PUSHs(sv_2mortal(newSViv((IV)ai32)));
- }
+# endif
+ if (!checksum)
+ mPUSHi(ai32);
else if (checksum > bits_in_uv)
cdouble += (NV)ai32;
else
cuv += ai32;
}
break;
+#endif /* PERL_PACK_CAN_SHRIEKSIGN */
case 'p':
while (len-- > 0) {
- if (sizeof(char*) > strend - s)
- break;
- else {
- Copy(s, &aptr, 1, char*);
- DO_BO_UNPACK_P(aptr);
- s += sizeof(char*);
- }
+ const char *aptr;
+ SHIFT_VAR(utf8, s, strend, aptr, datumtype);
+ DO_BO_UNPACK_PC(aptr);
/* newSVpv generates undef if aptr is NULL */
- PUSHs(sv_2mortal(newSVpv(aptr, 0)));
+ mPUSHs(newSVpv(aptr, 0));
}
break;
case 'w':
{
UV auv = 0;
U32 bytes = 0;
-
- while ((len > 0) && (s < strend)) {
- auv = (auv << 7) | (*s & 0x7f);
+
+ while (len > 0 && s < strend) {
+ U8 ch;
+ ch = SHIFT_BYTE(utf8, s, strend, datumtype);
+ auv = (auv << 7) | (ch & 0x7f);
/* UTF8_IS_XXXXX not right here - using constant 0x80 */
- if ((U8)(*s++) < 0x80) {
+ if (ch < 0x80) {
bytes = 0;
- PUSHs(sv_2mortal(newSVuv(auv)));
+ mPUSHu(auv);
len--;
auv = 0;
+ continue;
}
- else if (++bytes >= sizeof(UV)) { /* promote to string */
- char *t;
- STRLEN n_a;
+ if (++bytes >= sizeof(UV)) { /* promote to string */
+ const char *t;
- sv = Perl_newSVpvf(aTHX_ "%.*"UVf, (int)TYPE_DIGITS(UV), auv);
+ sv = Perl_newSVpvf(aTHX_ "%.*"UVuf, (int)TYPE_DIGITS(UV), auv);
while (s < strend) {
- sv = mul128(sv, (U8)(*s & 0x7f));
- if (!(*s++ & 0x80)) {
+ ch = SHIFT_BYTE(utf8, s, strend, datumtype);
+ sv = mul128(sv, (U8)(ch & 0x7f));
+ if (!(ch & 0x80)) {
bytes = 0;
break;
}
}
- t = SvPV(sv, n_a);
+ t = SvPV_nolen_const(sv);
while (*t == '0')
t++;
sv_chop(sv, t);
- PUSHs(sv_2mortal(sv));
+ mPUSHs(sv);
len--;
auv = 0;
}
if (symptr->howlen == e_star)
Perl_croak(aTHX_ "'P' must have an explicit size in unpack");
EXTEND(SP, 1);
- if (sizeof(char*) > strend - s)
- break;
- else {
- Copy(s, &aptr, 1, char*);
- DO_BO_UNPACK_P(aptr);
- s += sizeof(char*);
+ if (s + sizeof(char*) <= strend) {
+ char *aptr;
+ SHIFT_VAR(utf8, s, strend, aptr, datumtype);
+ DO_BO_UNPACK_PC(aptr);
+ /* newSVpvn generates undef if aptr is NULL */
+ PUSHs(newSVpvn_flags(aptr, len, SVs_TEMP));
}
- /* newSVpvn generates undef if aptr is NULL */
- PUSHs(sv_2mortal(newSVpvn(aptr, len)));
break;
#ifdef HAS_QUAD
case 'q':
while (len-- > 0) {
- assert (s + sizeof(Quad_t) <= strend);
- Copy(s, &aquad, 1, Quad_t);
+ Quad_t aquad;
+ SHIFT_VAR(utf8, s, strend, aquad, datumtype);
DO_BO_UNPACK(aquad, 64);
- s += sizeof(Quad_t);
- if (!checksum) {
- PUSHs(sv_2mortal((aquad >= IV_MIN && aquad <= IV_MAX) ?
- newSViv((IV)aquad) : newSVnv((NV)aquad)));
- }
+ if (!checksum)
+ mPUSHs(aquad >= IV_MIN && aquad <= IV_MAX ?
+ newSViv((IV)aquad) : newSVnv((NV)aquad));
else if (checksum > bits_in_uv)
cdouble += (NV)aquad;
else
break;
case 'Q':
while (len-- > 0) {
- assert (s + sizeof(Uquad_t) <= strend);
- Copy(s, &auquad, 1, Uquad_t);
+ Uquad_t auquad;
+ SHIFT_VAR(utf8, s, strend, auquad, datumtype);
DO_BO_UNPACK(auquad, 64);
- s += sizeof(Uquad_t);
- if (!checksum) {
- PUSHs(sv_2mortal((auquad <= UV_MAX) ?
- newSVuv((UV)auquad) : newSVnv((NV)auquad)));
- }
+ if (!checksum)
+ mPUSHs(auquad <= UV_MAX ?
+ newSVuv((UV)auquad) : newSVnv((NV)auquad));
else if (checksum > bits_in_uv)
cdouble += (NV)auquad;
else
cuv += auquad;
}
break;
-#endif
+#endif /* HAS_QUAD */
/* float and double added gnb@melba.bby.oz.au 22/11/89 */
case 'f':
while (len-- > 0) {
- Copy(s, &afloat, 1, float);
+ float afloat;
+ SHIFT_VAR(utf8, s, strend, afloat, datumtype);
DO_BO_UNPACK_N(afloat, float);
- s += sizeof(float);
- if (!checksum) {
- PUSHs(sv_2mortal(newSVnv((NV)afloat)));
- }
- else {
+ if (!checksum)
+ mPUSHn(afloat);
+ else
cdouble += afloat;
- }
}
break;
case 'd':
while (len-- > 0) {
- Copy(s, &adouble, 1, double);
+ double adouble;
+ SHIFT_VAR(utf8, s, strend, adouble, datumtype);
DO_BO_UNPACK_N(adouble, double);
- s += sizeof(double);
- if (!checksum) {
- PUSHs(sv_2mortal(newSVnv((NV)adouble)));
- }
- else {
+ if (!checksum)
+ mPUSHn(adouble);
+ else
cdouble += adouble;
- }
}
break;
case 'F':
while (len-- > 0) {
- Copy(s, &anv, 1, NV);
- DO_BO_UNPACK_N(anv, NV);
- s += NVSIZE;
- if (!checksum) {
- PUSHs(sv_2mortal(newSVnv(anv)));
- }
- else {
- cdouble += anv;
- }
+ NV_bytes anv;
+ SHIFT_BYTES(utf8, s, strend, anv.bytes, sizeof(anv.bytes), datumtype);
+ DO_BO_UNPACK_N(anv.nv, NV);
+ if (!checksum)
+ mPUSHn(anv.nv);
+ else
+ cdouble += anv.nv;
}
break;
#if defined(HAS_LONG_DOUBLE) && defined(USE_LONG_DOUBLE)
case 'D':
while (len-- > 0) {
- Copy(s, &aldouble, 1, long double);
- DO_BO_UNPACK_N(aldouble, long double);
- s += LONG_DOUBLESIZE;
- if (!checksum) {
- PUSHs(sv_2mortal(newSVnv((NV)aldouble)));
- }
- else {cdouble += aldouble;
- }
+ ld_bytes aldouble;
+ SHIFT_BYTES(utf8, s, strend, aldouble.bytes, sizeof(aldouble.bytes), datumtype);
+ DO_BO_UNPACK_N(aldouble.ld, long double);
+ if (!checksum)
+ mPUSHn(aldouble.ld);
+ else
+ cdouble += aldouble.ld;
}
break;
#endif
case 'u':
- /* MKS:
- * Initialise the decode mapping. By using a table driven
- * algorithm, the code will be character-set independent
- * (and just as fast as doing character arithmetic)
- */
- if (PL_uudmap['M'] == 0) {
- int i;
-
- for (i = 0; i < sizeof(PL_uuemap); i += 1)
- PL_uudmap[(U8)PL_uuemap[i]] = i;
- /*
- * Because ' ' and '`' map to the same value,
- * we need to decode them both the same.
- */
- PL_uudmap[' '] = 0;
- }
-
- along = (strend - s) * 3 / 4;
- sv = NEWSV(42, along);
- if (along)
- SvPOK_on(sv);
- while (s < strend && *s > ' ' && ISUUCHAR(*s)) {
- I32 a, b, c, d;
- char hunk[4];
-
- hunk[3] = '\0';
- len = PL_uudmap[*(U8*)s++] & 077;
- while (len > 0) {
- if (s < strend && ISUUCHAR(*s))
- a = PL_uudmap[*(U8*)s++] & 077;
- else
- a = 0;
- if (s < strend && ISUUCHAR(*s))
- b = PL_uudmap[*(U8*)s++] & 077;
- else
- b = 0;
- if (s < strend && ISUUCHAR(*s))
- c = PL_uudmap[*(U8*)s++] & 077;
- else
- c = 0;
- if (s < strend && ISUUCHAR(*s))
- d = PL_uudmap[*(U8*)s++] & 077;
- else
- d = 0;
- hunk[0] = (char)((a << 2) | (b >> 4));
- hunk[1] = (char)((b << 4) | (c >> 2));
- hunk[2] = (char)((c << 6) | d);
- sv_catpvn(sv, hunk, (len > 3) ? 3 : len);
- len -= 3;
+ if (!checksum) {
+ const STRLEN l = (STRLEN) (strend - s) * 3 / 4;
+ sv = sv_2mortal(newSV(l));
+ if (l) SvPOK_on(sv);
+ }
+ if (utf8) {
+ while (next_uni_uu(aTHX_ &s, strend, &len)) {
+ I32 a, b, c, d;
+ char hunk[3];
+
+ while (len > 0) {
+ next_uni_uu(aTHX_ &s, strend, &a);
+ next_uni_uu(aTHX_ &s, strend, &b);
+ next_uni_uu(aTHX_ &s, strend, &c);
+ next_uni_uu(aTHX_ &s, strend, &d);
+ hunk[0] = (char)((a << 2) | (b >> 4));
+ hunk[1] = (char)((b << 4) | (c >> 2));
+ hunk[2] = (char)((c << 6) | d);
+ if (!checksum)
+ sv_catpvn(sv, hunk, (len > 3) ? 3 : len);
+ len -= 3;
+ }
+ if (s < strend) {
+ if (*s == '\n') {
+ s++;
+ }
+ else {
+ /* possible checksum byte */
+ const char *skip = s+UTF8SKIP(s);
+ if (skip < strend && *skip == '\n')
+ s = skip+1;
+ }
+ }
+ }
+ } else {
+ while (s < strend && *s > ' ' && ISUUCHAR(*s)) {
+ I32 a, b, c, d;
+ char hunk[3];
+
+ len = PL_uudmap[*(U8*)s++] & 077;
+ while (len > 0) {
+ if (s < strend && ISUUCHAR(*s))
+ a = PL_uudmap[*(U8*)s++] & 077;
+ else
+ a = 0;
+ if (s < strend && ISUUCHAR(*s))
+ b = PL_uudmap[*(U8*)s++] & 077;
+ else
+ b = 0;
+ if (s < strend && ISUUCHAR(*s))
+ c = PL_uudmap[*(U8*)s++] & 077;
+ else
+ c = 0;
+ if (s < strend && ISUUCHAR(*s))
+ d = PL_uudmap[*(U8*)s++] & 077;
+ else
+ d = 0;
+ hunk[0] = (char)((a << 2) | (b >> 4));
+ hunk[1] = (char)((b << 4) | (c >> 2));
+ hunk[2] = (char)((c << 6) | d);
+ if (!checksum)
+ sv_catpvn(sv, hunk, (len > 3) ? 3 : len);
+ len -= 3;
+ }
+ if (*s == '\n')
+ s++;
+ else /* possible checksum byte */
+ if (s + 1 < strend && s[1] == '\n')
+ s += 2;
}
- if (*s == '\n')
- s++;
- else /* possible checksum byte */
- if (s + 1 < strend && s[1] == '\n')
- s += 2;
}
- XPUSHs(sv_2mortal(sv));
+ if (!checksum)
+ XPUSHs(sv);
break;
}
if (checksum) {
if (strchr("fFdD", TYPE_NO_MODIFIERS(datumtype)) ||
(checksum > bits_in_uv &&
- strchr("csSiIlLnNUvVqQjJ", TYPE_NO_MODIFIERS(datumtype))) ) {
- NV trouble;
+ strchr("cCsSiIlLnNUWvVqQjJ", TYPE_NO_MODIFIERS(datumtype))) ) {
+ NV trouble, anv;
- adouble = (NV) (1 << (checksum & 15));
+ anv = (NV) (1 << (checksum & 15));
while (checksum >= 16) {
checksum -= 16;
- adouble *= 65536.0;
+ anv *= 65536.0;
}
while (cdouble < 0.0)
- cdouble += adouble;
- cdouble = Perl_modf(cdouble / adouble, &trouble) * adouble;
+ cdouble += anv;
+ cdouble = Perl_modf(cdouble / anv, &trouble) * anv;
sv = newSVnv(cdouble);
}
else {
}
sv = newSVuv(cuv);
}
- XPUSHs(sv_2mortal(sv));
+ mXPUSHs(sv);
checksum = 0;
}
-
+
if (symptr->flags & FLAG_SLASH){
if (SP - PL_stack_base - start_sp_offset <= 0)
Perl_croak(aTHX_ "'/' must follow a numeric type in unpack");
Perl_croak(aTHX_ "Code missing after '/' in unpack" );
}
datumtype = symptr->code;
+ explicit_length = FALSE;
goto redo_switch;
}
}
PP(pp_unpack)
{
+ dVAR;
dSP;
dPOPPOPssrl;
I32 gimme = GIMME_V;
STRLEN llen;
STRLEN rlen;
- register char *pat = SvPV(left, llen);
-#ifdef PACKED_IS_OCTETS
- /* Packed side is assumed to be octets - so force downgrade if it
- has been UTF-8 encoded by accident
- */
- register char *s = SvPVbyte(right, rlen);
-#else
- register char *s = SvPV(right, rlen);
-#endif
- char *strend = s + rlen;
- register char *patend = pat + llen;
- register I32 cnt;
+ const char *pat = SvPV_const(left, llen);
+ const char *s = SvPV_const(right, rlen);
+ const char *strend = s + rlen;
+ const char *patend = pat + llen;
+ I32 cnt;
PUTBACK;
cnt = unpackstring(pat, patend, s, strend,
((gimme == G_SCALAR) ? FLAG_UNPACK_ONLY_ONE : 0)
- | (DO_UTF8(right) ? FLAG_UNPACK_DO_UTF8 : 0));
+ | (DO_UTF8(right) ? FLAG_DO_UTF8 : 0));
SPAGAIN;
if ( !cnt && gimme == G_SCALAR )
RETURN;
}
-STATIC void
-S_doencodes(pTHX_ register SV *sv, register char *s, register I32 len)
+STATIC U8 *
+doencodes(U8 *h, const char *s, I32 len)
{
- char hunk[5];
-
- *hunk = PL_uuemap[len];
- sv_catpvn(sv, hunk, 1);
- hunk[4] = '\0';
+ *h++ = PL_uuemap[len];
while (len > 2) {
- hunk[0] = PL_uuemap[(077 & (*s >> 2))];
- hunk[1] = PL_uuemap[(077 & (((*s << 4) & 060) | ((s[1] >> 4) & 017)))];
- hunk[2] = PL_uuemap[(077 & (((s[1] << 2) & 074) | ((s[2] >> 6) & 03)))];
- hunk[3] = PL_uuemap[(077 & (s[2] & 077))];
- sv_catpvn(sv, hunk, 4);
+ *h++ = PL_uuemap[(077 & (s[0] >> 2))];
+ *h++ = PL_uuemap[(077 & (((s[0] << 4) & 060) | ((s[1] >> 4) & 017)))];
+ *h++ = PL_uuemap[(077 & (((s[1] << 2) & 074) | ((s[2] >> 6) & 03)))];
+ *h++ = PL_uuemap[(077 & (s[2] & 077))];
s += 3;
len -= 3;
}
if (len > 0) {
- char r = (len > 1 ? s[1] : '\0');
- hunk[0] = PL_uuemap[(077 & (*s >> 2))];
- hunk[1] = PL_uuemap[(077 & (((*s << 4) & 060) | ((r >> 4) & 017)))];
- hunk[2] = PL_uuemap[(077 & ((r << 2) & 074))];
- hunk[3] = PL_uuemap[0];
- sv_catpvn(sv, hunk, 4);
+ const char r = (len > 1 ? s[1] : '\0');
+ *h++ = PL_uuemap[(077 & (s[0] >> 2))];
+ *h++ = PL_uuemap[(077 & (((s[0] << 4) & 060) | ((r >> 4) & 017)))];
+ *h++ = PL_uuemap[(077 & ((r << 2) & 074))];
+ *h++ = PL_uuemap[0];
}
- sv_catpvn(sv, "\n", 1);
+ *h++ = '\n';
+ return h;
}
STATIC SV *
-S_is_an_int(pTHX_ char *s, STRLEN l)
+S_is_an_int(pTHX_ const char *s, STRLEN l)
{
- STRLEN n_a;
- SV *result = newSVpvn(s, l);
- char *result_c = SvPV(result, n_a); /* convenience */
- char *out = result_c;
- bool skip = 1;
- bool ignore = 0;
+ SV *result = newSVpvn(s, l);
+ char *const result_c = SvPV_nolen(result); /* convenience */
+ char *out = result_c;
+ bool skip = 1;
+ bool ignore = 0;
+
+ PERL_ARGS_ASSERT_IS_AN_INT;
while (*s) {
switch (*s) {
STATIC int
S_div128(pTHX_ SV *pnum, bool *done)
{
- STRLEN len;
- char *s = SvPV(pnum, len);
- int m = 0;
- int r = 0;
- char *t = s;
-
- *done = 1;
- while (*t) {
- int i;
-
- i = m * 10 + (*t - '0');
- m = i & 0x7F;
- r = (i >> 7); /* r < 10 */
- if (r) {
- *done = 0;
+ STRLEN len;
+ char * const s = SvPV(pnum, len);
+ char *t = s;
+ int m = 0;
+
+ PERL_ARGS_ASSERT_DIV128;
+
+ *done = 1;
+ while (*t) {
+ const int i = m * 10 + (*t - '0');
+ const int r = (i >> 7); /* r < 10 */
+ m = i & 0x7F;
+ if (r) {
+ *done = 0;
+ }
+ *(t++) = '0' + r;
}
- *(t++) = '0' + r;
- }
- *(t++) = '\0';
- SvCUR_set(pnum, (STRLEN) (t - s));
- return (m);
+ *(t++) = '\0';
+ SvCUR_set(pnum, (STRLEN) (t - s));
+ return (m);
}
-
-
/*
-=for apidoc pack_cat
-
-The engine implementing pack() Perl function. Note: parameters next_in_list and
-flags are not used. This call should not be used; use packlist instead.
+=for apidoc packlist
-=cut */
+The engine implementing pack() Perl function.
+=cut
+*/
void
-Perl_pack_cat(pTHX_ SV *cat, char *pat, register char *patend, register SV **beglist, SV **endlist, SV ***next_in_list, U32 flags)
+Perl_packlist(pTHX_ SV *cat, const char *pat, const char *patend, register SV **beglist, SV **endlist )
{
- tempsym_t sym = { 0 };
- sym.patptr = pat;
- sym.patend = patend;
- sym.flags = FLAG_PACK;
+ dVAR;
+ tempsym_t sym;
+
+ PERL_ARGS_ASSERT_PACKLIST;
+
+ TEMPSYM_INIT(&sym, pat, patend, FLAG_PACK);
+
+ /* We're going to do changes through SvPVX(cat). Make sure it's valid.
+ Also make sure any UTF8 flag is loaded */
+ SvPV_force_nolen(cat);
+ if (DO_UTF8(cat))
+ sym.flags |= FLAG_PARSE_UTF8 | FLAG_DO_UTF8;
(void)pack_rec( cat, &sym, beglist, endlist );
}
+/* like sv_utf8_upgrade, but also repoint the group start markers */
+STATIC void
+marked_upgrade(pTHX_ SV *sv, tempsym_t *sym_ptr) {
+ STRLEN len;
+ tempsym_t *group;
+ const char *from_ptr, *from_start, *from_end, **marks, **m;
+ char *to_start, *to_ptr;
+
+ if (SvUTF8(sv)) return;
+
+ from_start = SvPVX_const(sv);
+ from_end = from_start + SvCUR(sv);
+ for (from_ptr = from_start; from_ptr < from_end; from_ptr++)
+ if (!NATIVE_IS_INVARIANT(*from_ptr)) break;
+ if (from_ptr == from_end) {
+ /* Simple case: no character needs to be changed */
+ SvUTF8_on(sv);
+ return;
+ }
-/*
-=for apidoc packlist
-
-The engine implementing pack() Perl function.
+ len = (from_end-from_ptr)*UTF8_EXPAND+(from_ptr-from_start)+1;
+ Newx(to_start, len, char);
+ Copy(from_start, to_start, from_ptr-from_start, char);
+ to_ptr = to_start + (from_ptr-from_start);
+
+ Newx(marks, sym_ptr->level+2, const char *);
+ for (group=sym_ptr; group; group = group->previous)
+ marks[group->level] = from_start + group->strbeg;
+ marks[sym_ptr->level+1] = from_end+1;
+ for (m = marks; *m < from_ptr; m++)
+ *m = to_start + (*m-from_start);
+
+ for (;from_ptr < from_end; from_ptr++) {
+ while (*m == from_ptr) *m++ = to_ptr;
+ to_ptr = (char *) uvchr_to_utf8((U8 *) to_ptr, *(U8 *) from_ptr);
+ }
+ *to_ptr = 0;
-=cut */
+ while (*m == from_ptr) *m++ = to_ptr;
+ if (m != marks + sym_ptr->level+1) {
+ Safefree(marks);
+ Safefree(to_start);
+ Perl_croak(aTHX_ "panic: marks beyond string end");
+ }
+ for (group=sym_ptr; group; group = group->previous)
+ group->strbeg = marks[group->level] - to_start;
+ Safefree(marks);
+
+ if (SvOOK(sv)) {
+ if (SvIVX(sv)) {
+ SvLEN_set(sv, SvLEN(sv) + SvIVX(sv));
+ from_start -= SvIVX(sv);
+ SvIV_set(sv, 0);
+ }
+ SvFLAGS(sv) &= ~SVf_OOK;
+ }
+ if (SvLEN(sv) != 0)
+ Safefree(from_start);
+ SvPV_set(sv, to_start);
+ SvCUR_set(sv, to_ptr - to_start);
+ SvLEN_set(sv, len);
+ SvUTF8_on(sv);
+}
+/* Exponential string grower. Makes string extension effectively O(n)
+ needed says how many extra bytes we need (not counting the final '\0')
+ Only grows the string if there is an actual lack of space
+*/
+STATIC char *
+S_sv_exp_grow(pTHX_ SV *sv, STRLEN needed) {
+ const STRLEN cur = SvCUR(sv);
+ const STRLEN len = SvLEN(sv);
+ STRLEN extend;
-void
-Perl_packlist(pTHX_ SV *cat, char *pat, register char *patend, register SV **beglist, SV **endlist )
-{
- tempsym_t sym = { 0 };
- sym.patptr = pat;
- sym.patend = patend;
- sym.flags = FLAG_PACK;
+ PERL_ARGS_ASSERT_SV_EXP_GROW;
- (void)pack_rec( cat, &sym, beglist, endlist );
+ if (len - cur > needed) return SvPVX(sv);
+ extend = needed > len ? needed : len;
+ return SvGROW(sv, len+extend+1);
}
-
STATIC
SV **
-S_pack_rec(pTHX_ SV *cat, register tempsym_t* symptr, register SV **beglist, SV **endlist )
+S_pack_rec(pTHX_ SV *cat, tempsym_t* symptr, SV **beglist, SV **endlist )
{
- register I32 items;
- STRLEN fromlen;
- register I32 len = 0;
- SV *fromstr;
- /*SUPPRESS 442*/
- static char null10[] = {0,0,0,0,0,0,0,0,0,0};
- static char *space10 = " ";
- bool found;
-
- /* These must not be in registers: */
- char achar;
- I16 ai16;
- U16 au16;
- I32 ai32;
- U32 au32;
-#ifdef HAS_QUAD
- Quad_t aquad;
- Uquad_t auquad;
-#endif
-#if SHORTSIZE != SIZE16
- short ashort;
- unsigned short aushort;
-#endif
- int aint;
- unsigned int auint;
-#if LONGSIZE != SIZE32
- long along;
- unsigned long aulong;
-#endif
- char *aptr;
- float afloat;
- double adouble;
-#if defined(HAS_LONG_DOUBLE) && defined(USE_LONG_DOUBLE)
- long double aldouble;
-#endif
- IV aiv;
- UV auv;
- NV anv;
-
- int strrelbeg = SvCUR(cat);
+ dVAR;
tempsym_t lookahead;
+ I32 items = endlist - beglist;
+ bool found = next_symbol(symptr);
+ bool utf8 = (symptr->flags & FLAG_PARSE_UTF8) ? 1 : 0;
+ bool warn_utf8 = ckWARN(WARN_UTF8);
- items = endlist - beglist;
- found = next_symbol( symptr );
+ PERL_ARGS_ASSERT_PACK_REC;
-#ifndef PACKED_IS_OCTETS
- if (symptr->level == 0 && found && symptr->code == 'U' ){
- SvUTF8_on(cat);
+ if (symptr->level == 0 && found && symptr->code == 'U') {
+ marked_upgrade(aTHX_ cat, symptr);
+ symptr->flags |= FLAG_DO_UTF8;
+ utf8 = 0;
}
-#endif
+ symptr->strbeg = SvCUR(cat);
while (found) {
- SV *lengthcode = Nullsv;
-#define NEXTFROM ( lengthcode ? lengthcode : items-- > 0 ? *beglist++ : &PL_sv_no)
-
+ SV *fromstr;
+ STRLEN fromlen;
+ I32 len;
+ SV *lengthcode = NULL;
I32 datumtype = symptr->code;
- howlen_t howlen;
+ howlen_t howlen = symptr->howlen;
+ char *start = SvPVX(cat);
+ char *cur = start + SvCUR(cat);
- switch( howlen = symptr->howlen ){
- case e_no_len:
- case e_number:
- len = symptr->length;
+#define NEXTFROM (lengthcode ? lengthcode : items-- > 0 ? *beglist++ : &PL_sv_no)
+
+ switch (howlen) {
+ case e_star:
+ len = strchr("@Xxu", TYPE_NO_MODIFIERS(datumtype)) ?
+ 0 : items;
break;
- case e_star:
- len = strchr("@Xxu", TYPE_NO_MODIFIERS(datumtype)) ? 0 : items;
+ default:
+ /* e_no_len and e_number */
+ len = symptr->length;
break;
}
+ if (len) {
+ packprops_t props = packprops[TYPE_NO_ENDIANNESS(datumtype)];
+
+ if (props && !(props & PACK_SIZE_UNPREDICTABLE)) {
+ /* We can process this letter. */
+ STRLEN size = props & PACK_SIZE_MASK;
+ GROWING(utf8, cat, start, cur, (STRLEN) len * size);
+ }
+ }
+
/* Look ahead for next symbol. Do we have code/code? */
lookahead = *symptr;
found = next_symbol(&lookahead);
- if ( symptr->flags & FLAG_SLASH ) {
- if (found){
- if ( 0 == strchr( "aAZ", lookahead.code ) ||
- e_star != lookahead.howlen )
- Perl_croak(aTHX_ "'/' must be followed by 'a*', 'A*' or 'Z*' in pack");
- lengthcode = sv_2mortal(newSViv(sv_len(items > 0
- ? *beglist : &PL_sv_no)
- + (lookahead.code == 'Z' ? 1 : 0)));
+ if (symptr->flags & FLAG_SLASH) {
+ IV count;
+ if (!found) Perl_croak(aTHX_ "Code missing after '/' in pack");
+ if (strchr("aAZ", lookahead.code)) {
+ if (lookahead.howlen == e_number) count = lookahead.length;
+ else {
+ if (items > 0) {
+ if (SvGAMAGIC(*beglist)) {
+ /* Avoid reading the active data more than once
+ by copying it to a temporary. */
+ STRLEN len;
+ const char *const pv = SvPV_const(*beglist, len);
+ SV *const temp
+ = newSVpvn_flags(pv, len,
+ SVs_TEMP | SvUTF8(*beglist));
+ *beglist = temp;
+ }
+ count = DO_UTF8(*beglist) ?
+ sv_len_utf8(*beglist) : sv_len(*beglist);
+ }
+ else count = 0;
+ if (lookahead.code == 'Z') count++;
+ }
} else {
- Perl_croak(aTHX_ "Code missing after '/' in pack");
- }
+ if (lookahead.howlen == e_number && lookahead.length < items)
+ count = lookahead.length;
+ else count = items;
+ }
+ lookahead.howlen = e_number;
+ lookahead.length = count;
+ lengthcode = sv_2mortal(newSViv(count));
}
+ /* Code inside the switch must take care to properly update
+ cat (CUR length and '\0' termination) if it updated *cur and
+ doesn't simply leave using break */
switch(TYPE_NO_ENDIANNESS(datumtype)) {
default:
- Perl_croak(aTHX_ "Invalid type '%c' in pack", (int)TYPE_NO_MODIFIERS(datumtype));
+ Perl_croak(aTHX_ "Invalid type '%c' in pack",
+ (int) TYPE_NO_MODIFIERS(datumtype));
case '%':
Perl_croak(aTHX_ "'%%' may not be used in pack");
+ {
+ char *from;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+ case '.' | TYPE_IS_SHRIEKING:
+#endif
+ case '.':
+ if (howlen == e_star) from = start;
+ else if (len == 0) from = cur;
+ else {
+ tempsym_t *group = symptr;
+
+ while (--len && group) group = group->previous;
+ from = group ? start + group->strbeg : start;
+ }
+ fromstr = NEXTFROM;
+ len = SvIV(fromstr);
+ goto resize;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+ case '@' | TYPE_IS_SHRIEKING:
+#endif
case '@':
- len += strrelbeg - SvCUR(cat);
- if (len > 0)
- goto grow;
- len = -len;
- if (len > 0)
+ from = start + symptr->strbeg;
+ resize:
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
+ if (utf8 && !(datumtype & TYPE_IS_SHRIEKING))
+#else /* PERL_PACK_CAN_SHRIEKSIGN */
+ if (utf8)
+#endif
+ if (len >= 0) {
+ while (len && from < cur) {
+ from += UTF8SKIP(from);
+ len--;
+ }
+ if (from > cur)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in pack");
+ if (len) {
+ /* Here we know from == cur */
+ grow:
+ GROWING(0, cat, start, cur, len);
+ Zero(cur, len, char);
+ cur += len;
+ } else if (from < cur) {
+ len = cur - from;
+ goto shrink;
+ } else goto no_change;
+ } else {
+ cur = from;
+ len = -len;
+ goto utf8_shrink;
+ }
+ else {
+ len -= cur - from;
+ if (len > 0) goto grow;
+ if (len == 0) goto no_change;
+ len = -len;
goto shrink;
+ }
break;
- case '(':
- {
+ }
+ case '(': {
tempsym_t savsym = *symptr;
U32 group_modifiers = TYPE_MODIFIERS(datumtype & ~symptr->flags);
symptr->flags |= group_modifiers;
symptr->patend = savsym.grpend;
symptr->level++;
+ symptr->previous = &lookahead;
while (len--) {
+ U32 was_utf8;
+ if (utf8) symptr->flags |= FLAG_PARSE_UTF8;
+ else symptr->flags &= ~FLAG_PARSE_UTF8;
+ was_utf8 = SvUTF8(cat);
symptr->patptr = savsym.grpbeg;
- beglist = pack_rec(cat, symptr, beglist, endlist );
+ beglist = pack_rec(cat, symptr, beglist, endlist);
+ if (SvUTF8(cat) != was_utf8)
+ /* This had better be an upgrade while in utf8==0 mode */
+ utf8 = 1;
+
if (savsym.howlen == e_star && beglist == endlist)
break; /* No way to continue */
}
- symptr->flags &= ~group_modifiers;
- lookahead.flags = symptr->flags;
- *symptr = savsym;
- break;
+ items = endlist - beglist;
+ lookahead.flags = symptr->flags & ~group_modifiers;
+ goto no_change;
}
case 'X' | TYPE_IS_SHRIEKING:
if (!len) /* Avoid division by 0 */
len = 1;
- len = (SvCUR(cat)) % len;
+ if (utf8) {
+ char *hop, *last;
+ I32 l = len;
+ hop = last = start;
+ while (hop < cur) {
+ hop += UTF8SKIP(hop);
+ if (--l == 0) {
+ last = hop;
+ l = len;
+ }
+ }
+ if (last > cur)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in pack");
+ cur = last;
+ break;
+ }
+ len = (cur-start) % len;
/* FALL THROUGH */
case 'X':
- shrink:
- if ((I32)SvCUR(cat) < len)
- Perl_croak(aTHX_ "'X' outside of string in pack");
- SvCUR(cat) -= len;
- *SvEND(cat) = '\0';
- break;
- case 'x' | TYPE_IS_SHRIEKING:
+ if (utf8) {
+ if (len < 1) goto no_change;
+ utf8_shrink:
+ while (len > 0) {
+ if (cur <= start)
+ Perl_croak(aTHX_ "'%c' outside of string in pack",
+ (int) TYPE_NO_MODIFIERS(datumtype));
+ while (--cur, UTF8_IS_CONTINUATION(*cur)) {
+ if (cur <= start)
+ Perl_croak(aTHX_ "'%c' outside of string in pack",
+ (int) TYPE_NO_MODIFIERS(datumtype));
+ }
+ len--;
+ }
+ } else {
+ shrink:
+ if (cur - start < len)
+ Perl_croak(aTHX_ "'%c' outside of string in pack",
+ (int) TYPE_NO_MODIFIERS(datumtype));
+ cur -= len;
+ }
+ if (cur < start+symptr->strbeg) {
+ /* Make sure group starts don't point into the void */
+ tempsym_t *group;
+ const STRLEN length = cur-start;
+ for (group = symptr;
+ group && length < group->strbeg;
+ group = group->previous) group->strbeg = length;
+ lookahead.strbeg = length;
+ }
+ break;
+ case 'x' | TYPE_IS_SHRIEKING: {
+ I32 ai32;
if (!len) /* Avoid division by 0 */
len = 1;
- aint = (SvCUR(cat)) % len;
- if (aint) /* Other portable ways? */
- len = len - aint;
- else
- len = 0;
- /* FALL THROUGH */
-
+ if (utf8) ai32 = utf8_length((U8 *) start, (U8 *) cur) % len;
+ else ai32 = (cur - start) % len;
+ if (ai32 == 0) goto no_change;
+ len -= ai32;
+ }
+ /* FALL THROUGH */
case 'x':
- grow:
- while (len >= 10) {
- sv_catpvn(cat, null10, 10);
- len -= 10;
- }
- sv_catpvn(cat, null10, len);
- break;
+ goto grow;
case 'A':
case 'Z':
- case 'a':
+ case 'a': {
+ const char *aptr;
+
fromstr = NEXTFROM;
- aptr = SvPV(fromstr, fromlen);
- if (howlen == e_star) {
- len = fromlen;
- if (datumtype == 'Z')
- ++len;
- }
- if ((I32)fromlen >= len) {
- sv_catpvn(cat, aptr, len);
- if (datumtype == 'Z')
- *(SvEND(cat)-1) = '\0';
- }
- else {
- sv_catpvn(cat, aptr, fromlen);
+ aptr = SvPV_const(fromstr, fromlen);
+ if (DO_UTF8(fromstr)) {
+ const char *end, *s;
+
+ if (!utf8 && !SvUTF8(cat)) {
+ marked_upgrade(aTHX_ cat, symptr);
+ lookahead.flags |= FLAG_DO_UTF8;
+ lookahead.strbeg = symptr->strbeg;
+ utf8 = 1;
+ start = SvPVX(cat);
+ cur = start + SvCUR(cat);
+ }
+ if (howlen == e_star) {
+ if (utf8) goto string_copy;
+ len = fromlen+1;
+ }
+ s = aptr;
+ end = aptr + fromlen;
+ fromlen = datumtype == 'Z' ? len-1 : len;
+ while ((I32) fromlen > 0 && s < end) {
+ s += UTF8SKIP(s);
+ fromlen--;
+ }
+ if (s > end)
+ Perl_croak(aTHX_ "Malformed UTF-8 string in pack");
+ if (utf8) {
+ len = fromlen;
+ if (datumtype == 'Z') len++;
+ fromlen = s-aptr;
+ len += fromlen;
+
+ goto string_copy;
+ }
+ fromlen = len - fromlen;
+ if (datumtype == 'Z') fromlen--;
+ if (howlen == e_star) {
+ len = fromlen;
+ if (datumtype == 'Z') len++;
+ }
+ GROWING(0, cat, start, cur, len);
+ if (!uni_to_bytes(aTHX_ &aptr, end, cur, fromlen,
+ datumtype | TYPE_IS_PACK))
+ Perl_croak(aTHX_ "panic: predicted utf8 length not available");
+ cur += fromlen;
len -= fromlen;
- if (datumtype == 'A') {
- while (len >= 10) {
- sv_catpvn(cat, space10, 10);
- len -= 10;
- }
- sv_catpvn(cat, space10, len);
+ } else if (utf8) {
+ if (howlen == e_star) {
+ len = fromlen;
+ if (datumtype == 'Z') len++;
}
- else {
- while (len >= 10) {
- sv_catpvn(cat, null10, 10);
- len -= 10;
- }
- sv_catpvn(cat, null10, len);
+ if (len <= (I32) fromlen) {
+ fromlen = len;
+ if (datumtype == 'Z' && fromlen > 0) fromlen--;
+ }
+ /* assumes a byte expands to at most UTF8_EXPAND bytes on
+ upgrade, so:
+ expected_length <= from_len*UTF8_EXPAND + (len-from_len) */
+ GROWING(0, cat, start, cur, fromlen*(UTF8_EXPAND-1)+len);
+ len -= fromlen;
+ while (fromlen > 0) {
+ cur = (char *) uvchr_to_utf8((U8 *) cur, * (U8 *) aptr);
+ aptr++;
+ fromlen--;
+ }
+ } else {
+ string_copy:
+ if (howlen == e_star) {
+ len = fromlen;
+ if (datumtype == 'Z') len++;
+ }
+ if (len <= (I32) fromlen) {
+ fromlen = len;
+ if (datumtype == 'Z' && fromlen > 0) fromlen--;
}
+ GROWING(0, cat, start, cur, len);
+ Copy(aptr, cur, fromlen, char);
+ cur += fromlen;
+ len -= fromlen;
}
+ memset(cur, datumtype == 'A' ? ' ' : '\0', len);
+ cur += len;
+ SvTAINT(cat);
break;
+ }
case 'B':
- case 'b':
- {
- register char *str;
- I32 saveitems;
+ case 'b': {
+ const char *str, *end;
+ I32 l, field_len;
+ U8 bits;
+ bool utf8_source;
+ U32 utf8_flags;
- fromstr = NEXTFROM;
- saveitems = items;
- str = SvPV(fromstr, fromlen);
- if (howlen == e_star)
- len = fromlen;
- aint = SvCUR(cat);
- SvCUR(cat) += (len+7)/8;
- SvGROW(cat, SvCUR(cat) + 1);
- aptr = SvPVX(cat) + aint;
- if (len > (I32)fromlen)
- len = fromlen;
- aint = len;
- items = 0;
- if (datumtype == 'B') {
- for (len = 0; len++ < aint;) {
- items |= *str++ & 1;
- if (len & 7)
- items <<= 1;
- else {
- *aptr++ = items & 0xff;
- items = 0;
- }
+ fromstr = NEXTFROM;
+ str = SvPV_const(fromstr, fromlen);
+ end = str + fromlen;
+ if (DO_UTF8(fromstr)) {
+ utf8_source = TRUE;
+ utf8_flags = warn_utf8 ? 0 : UTF8_ALLOW_ANY;
+ } else {
+ utf8_source = FALSE;
+ utf8_flags = 0; /* Unused, but keep compilers happy */
+ }
+ if (howlen == e_star) len = fromlen;
+ field_len = (len+7)/8;
+ GROWING(utf8, cat, start, cur, field_len);
+ if (len > (I32)fromlen) len = fromlen;
+ bits = 0;
+ l = 0;
+ if (datumtype == 'B')
+ while (l++ < len) {
+ if (utf8_source) {
+ UV val = 0;
+ NEXT_UNI_VAL(val, cur, str, end, utf8_flags);
+ bits |= val & 1;
+ } else bits |= *str++ & 1;
+ if (l & 7) bits <<= 1;
+ else {
+ PUSH_BYTE(utf8, cur, bits);
+ bits = 0;
}
}
- else {
- for (len = 0; len++ < aint;) {
- if (*str++ & 1)
- items |= 128;
- if (len & 7)
- items >>= 1;
- else {
- *aptr++ = items & 0xff;
- items = 0;
- }
+ else
+ /* datumtype == 'b' */
+ while (l++ < len) {
+ if (utf8_source) {
+ UV val = 0;
+ NEXT_UNI_VAL(val, cur, str, end, utf8_flags);
+ if (val & 1) bits |= 0x80;
+ } else if (*str++ & 1)
+ bits |= 0x80;
+ if (l & 7) bits >>= 1;
+ else {
+ PUSH_BYTE(utf8, cur, bits);
+ bits = 0;
}
}
- if (aint & 7) {
- if (datumtype == 'B')
- items <<= 7 - (aint & 7);
- else
- items >>= 7 - (aint & 7);
- *aptr++ = items & 0xff;
- }
- str = SvPVX(cat) + SvCUR(cat);
- while (aptr <= str)
- *aptr++ = '\0';
-
- items = saveitems;
+ l--;
+ if (l & 7) {
+ if (datumtype == 'B')
+ bits <<= 7 - (l & 7);
+ else
+ bits >>= 7 - (l & 7);
+ PUSH_BYTE(utf8, cur, bits);
+ l += 7;
}
+ /* Determine how many chars are left in the requested field */
+ l /= 8;
+ if (howlen == e_star) field_len = 0;
+ else field_len -= l;
+ Zero(cur, field_len, char);
+ cur += field_len;
break;
+ }
case 'H':
- case 'h':
- {
- register char *str;
- I32 saveitems;
+ case 'h': {
+ const char *str, *end;
+ I32 l, field_len;
+ U8 bits;
+ bool utf8_source;
+ U32 utf8_flags;
- fromstr = NEXTFROM;
- saveitems = items;
- str = SvPV(fromstr, fromlen);
- if (howlen == e_star)
- len = fromlen;
- aint = SvCUR(cat);
- SvCUR(cat) += (len+1)/2;
- SvGROW(cat, SvCUR(cat) + 1);
- aptr = SvPVX(cat) + aint;
- if (len > (I32)fromlen)
- len = fromlen;
- aint = len;
- items = 0;
- if (datumtype == 'H') {
- for (len = 0; len++ < aint;) {
- if (isALPHA(*str))
- items |= ((*str++ & 15) + 9) & 15;
+ fromstr = NEXTFROM;
+ str = SvPV_const(fromstr, fromlen);
+ end = str + fromlen;
+ if (DO_UTF8(fromstr)) {
+ utf8_source = TRUE;
+ utf8_flags = warn_utf8 ? 0 : UTF8_ALLOW_ANY;
+ } else {
+ utf8_source = FALSE;
+ utf8_flags = 0; /* Unused, but keep compilers happy */
+ }
+ if (howlen == e_star) len = fromlen;
+ field_len = (len+1)/2;
+ GROWING(utf8, cat, start, cur, field_len);
+ if (!utf8 && len > (I32)fromlen) len = fromlen;
+ bits = 0;
+ l = 0;
+ if (datumtype == 'H')
+ while (l++ < len) {
+ if (utf8_source) {
+ UV val = 0;
+ NEXT_UNI_VAL(val, cur, str, end, utf8_flags);
+ if (val < 256 && isALPHA(val))
+ bits |= (val + 9) & 0xf;
else
- items |= *str++ & 15;
- if (len & 1)
- items <<= 4;
- else {
- *aptr++ = items & 0xff;
- items = 0;
- }
+ bits |= val & 0xf;
+ } else if (isALPHA(*str))
+ bits |= (*str++ + 9) & 0xf;
+ else
+ bits |= *str++ & 0xf;
+ if (l & 1) bits <<= 4;
+ else {
+ PUSH_BYTE(utf8, cur, bits);
+ bits = 0;
}
}
- else {
- for (len = 0; len++ < aint;) {
- if (isALPHA(*str))
- items |= (((*str++ & 15) + 9) & 15) << 4;
+ else
+ while (l++ < len) {
+ if (utf8_source) {
+ UV val = 0;
+ NEXT_UNI_VAL(val, cur, str, end, utf8_flags);
+ if (val < 256 && isALPHA(val))
+ bits |= ((val + 9) & 0xf) << 4;
else
- items |= (*str++ & 15) << 4;
- if (len & 1)
- items >>= 4;
- else {
- *aptr++ = items & 0xff;
- items = 0;
- }
+ bits |= (val & 0xf) << 4;
+ } else if (isALPHA(*str))
+ bits |= ((*str++ + 9) & 0xf) << 4;
+ else
+ bits |= (*str++ & 0xf) << 4;
+ if (l & 1) bits >>= 4;
+ else {
+ PUSH_BYTE(utf8, cur, bits);
+ bits = 0;
}
}
- if (aint & 1)
- *aptr++ = items & 0xff;
- str = SvPVX(cat) + SvCUR(cat);
- while (aptr <= str)
- *aptr++ = '\0';
-
- items = saveitems;
+ l--;
+ if (l & 1) {
+ PUSH_BYTE(utf8, cur, bits);
+ l++;
+ }
+ /* Determine how many chars are left in the requested field */
+ l /= 2;
+ if (howlen == e_star) field_len = 0;
+ else field_len -= l;
+ Zero(cur, field_len, char);
+ cur += field_len;
+ break;
+ }
+ case 'c':
+ while (len-- > 0) {
+ IV aiv;
+ fromstr = NEXTFROM;
+ aiv = SvIV(fromstr);
+ if ((-128 > aiv || aiv > 127))
+ Perl_ck_warner(aTHX_ packWARN(WARN_PACK),
+ "Character in 'c' format wrapped in pack");
+ PUSH_BYTE(utf8, cur, (U8)(aiv & 0xff));
}
break;
case 'C':
- case 'c':
+ if (len == 0) {
+ utf8 = (symptr->flags & FLAG_DO_UTF8) ? 1 : 0;
+ break;
+ }
+ while (len-- > 0) {
+ IV aiv;
+ fromstr = NEXTFROM;
+ aiv = SvIV(fromstr);
+ if ((0 > aiv || aiv > 0xff))
+ Perl_ck_warner(aTHX_ packWARN(WARN_PACK),
+ "Character in 'C' format wrapped in pack");
+ PUSH_BYTE(utf8, cur, (U8)(aiv & 0xff));
+ }
+ break;
+ case 'W': {
+ char *end;
+ U8 in_bytes = (U8)IN_BYTES;
+
+ end = start+SvLEN(cat)-1;
+ if (utf8) end -= UTF8_MAXLEN-1;
while (len-- > 0) {
+ UV auv;
fromstr = NEXTFROM;
- switch (TYPE_NO_MODIFIERS(datumtype)) {
- case 'C':
- aint = SvIV(fromstr);
- if ((aint < 0 || aint > 255) &&
- ckWARN(WARN_PACK))
- Perl_warner(aTHX_ packWARN(WARN_PACK),
- "Character in 'C' format wrapped in pack");
- achar = aint & 255;
- sv_catpvn(cat, &achar, sizeof(char));
- break;
- case 'c':
- aint = SvIV(fromstr);
- if ((aint < -128 || aint > 127) &&
- ckWARN(WARN_PACK))
- Perl_warner(aTHX_ packWARN(WARN_PACK),
- "Character in 'c' format wrapped in pack" );
- achar = aint & 255;
- sv_catpvn(cat, &achar, sizeof(char));
- break;
+ auv = SvUV(fromstr);
+ if (in_bytes) auv = auv % 0x100;
+ if (utf8) {
+ W_utf8:
+ if (cur > end) {
+ *cur = '\0';
+ SvCUR_set(cat, cur - start);
+
+ GROWING(0, cat, start, cur, len+UTF8_MAXLEN);
+ end = start+SvLEN(cat)-UTF8_MAXLEN;
+ }
+ cur = (char *) uvuni_to_utf8_flags((U8 *) cur,
+ NATIVE_TO_UNI(auv),
+ warn_utf8 ?
+ 0 : UNICODE_ALLOW_ANY);
+ } else {
+ if (auv >= 0x100) {
+ if (!SvUTF8(cat)) {
+ *cur = '\0';
+ SvCUR_set(cat, cur - start);
+ marked_upgrade(aTHX_ cat, symptr);
+ lookahead.flags |= FLAG_DO_UTF8;
+ lookahead.strbeg = symptr->strbeg;
+ utf8 = 1;
+ start = SvPVX(cat);
+ cur = start + SvCUR(cat);
+ end = start+SvLEN(cat)-UTF8_MAXLEN;
+ goto W_utf8;
+ }
+ Perl_ck_warner(aTHX_ packWARN(WARN_PACK),
+ "Character in 'W' format wrapped in pack");
+ auv &= 0xff;
+ }
+ if (cur >= end) {
+ *cur = '\0';
+ SvCUR_set(cat, cur - start);
+ GROWING(0, cat, start, cur, len+1);
+ end = start+SvLEN(cat)-1;
+ }
+ *(U8 *) cur++ = (U8)auv;
}
}
break;
- case 'U':
+ }
+ case 'U': {
+ char *end;
+
+ if (len == 0) {
+ if (!(symptr->flags & FLAG_DO_UTF8)) {
+ marked_upgrade(aTHX_ cat, symptr);
+ lookahead.flags |= FLAG_DO_UTF8;
+ lookahead.strbeg = symptr->strbeg;
+ }
+ utf8 = 0;
+ goto no_change;
+ }
+
+ end = start+SvLEN(cat);
+ if (!utf8) end -= UTF8_MAXLEN;
while (len-- > 0) {
+ UV auv;
fromstr = NEXTFROM;
- auint = UNI_TO_NATIVE(SvUV(fromstr));
- SvGROW(cat, SvCUR(cat) + UTF8_MAXBYTES + 1);
- SvCUR_set(cat,
- (char*)uvchr_to_utf8_flags((U8*)SvEND(cat),
- auint,
- ckWARN(WARN_UTF8) ?
- 0 : UNICODE_ALLOW_ANY)
- - SvPVX(cat));
+ auv = SvUV(fromstr);
+ if (utf8) {
+ U8 buffer[UTF8_MAXLEN], *endb;
+ endb = uvuni_to_utf8_flags(buffer, auv,
+ warn_utf8 ?
+ 0 : UNICODE_ALLOW_ANY);
+ if (cur+(endb-buffer)*UTF8_EXPAND >= end) {
+ *cur = '\0';
+ SvCUR_set(cat, cur - start);
+ GROWING(0, cat, start, cur,
+ len+(endb-buffer)*UTF8_EXPAND);
+ end = start+SvLEN(cat);
+ }
+ cur = bytes_to_uni(buffer, endb-buffer, cur);
+ } else {
+ if (cur >= end) {
+ *cur = '\0';
+ SvCUR_set(cat, cur - start);
+ GROWING(0, cat, start, cur, len+UTF8_MAXLEN);
+ end = start+SvLEN(cat)-UTF8_MAXLEN;
+ }
+ cur = (char *) uvuni_to_utf8_flags((U8 *) cur, auv,
+ warn_utf8 ?
+ 0 : UNICODE_ALLOW_ANY);
+ }
}
- *SvEND(cat) = '\0';
break;
+ }
/* Float and double added by gnb@melba.bby.oz.au 22/11/89 */
case 'f':
while (len-- > 0) {
+ float afloat;
+ NV anv;
fromstr = NEXTFROM;
+ anv = SvNV(fromstr);
#ifdef __VOS__
-/* VOS does not automatically map a floating-point overflow
- during conversion from double to float into infinity, so we
- do it by hand. This code should either be generalized for
- any OS that needs it, or removed if and when VOS implements
- posix-976 (suggestion to support mapping to infinity).
- Paul.Green@stratus.com 02-04-02. */
- if (SvNV(fromstr) > FLT_MAX)
- afloat = _float_constants[0]; /* single prec. inf. */
- else if (SvNV(fromstr) < -FLT_MAX)
- afloat = _float_constants[0]; /* single prec. inf. */
- else afloat = (float)SvNV(fromstr);
-#else
+ /* VOS does not automatically map a floating-point overflow
+ during conversion from double to float into infinity, so we
+ do it by hand. This code should either be generalized for
+ any OS that needs it, or removed if and when VOS implements
+ posix-976 (suggestion to support mapping to infinity).
+ Paul.Green@stratus.com 02-04-02. */
+{
+extern const float _float_constants[];
+ if (anv > FLT_MAX)
+ afloat = _float_constants[0]; /* single prec. inf. */
+ else if (anv < -FLT_MAX)
+ afloat = _float_constants[0]; /* single prec. inf. */
+ else afloat = (float) anv;
+}
+#else /* __VOS__ */
# if defined(VMS) && !defined(__IEEE_FP)
-/* IEEE fp overflow shenanigans are unavailable on VAX and optional
- * on Alpha; fake it if we don't have them.
- */
- if (SvNV(fromstr) > FLT_MAX)
- afloat = FLT_MAX;
- else if (SvNV(fromstr) < -FLT_MAX)
- afloat = -FLT_MAX;
- else afloat = (float)SvNV(fromstr);
+ /* IEEE fp overflow shenanigans are unavailable on VAX and optional
+ * on Alpha; fake it if we don't have them.
+ */
+ if (anv > FLT_MAX)
+ afloat = FLT_MAX;
+ else if (anv < -FLT_MAX)
+ afloat = -FLT_MAX;
+ else afloat = (float)anv;
# else
- afloat = (float)SvNV(fromstr);
+ afloat = (float)anv;
# endif
-#endif
+#endif /* __VOS__ */
DO_BO_PACK_N(afloat, float);
- sv_catpvn(cat, (char *)&afloat, sizeof (float));
+ PUSH_VAR(utf8, cur, afloat);
}
break;
case 'd':
while (len-- > 0) {
+ double adouble;
+ NV anv;
fromstr = NEXTFROM;
+ anv = SvNV(fromstr);
#ifdef __VOS__
-/* VOS does not automatically map a floating-point overflow
- during conversion from long double to double into infinity,
- so we do it by hand. This code should either be generalized
- for any OS that needs it, or removed if and when VOS
- implements posix-976 (suggestion to support mapping to
- infinity). Paul.Green@stratus.com 02-04-02. */
- if (SvNV(fromstr) > DBL_MAX)
- adouble = _double_constants[0]; /* double prec. inf. */
- else if (SvNV(fromstr) < -DBL_MAX)
- adouble = _double_constants[0]; /* double prec. inf. */
- else adouble = (double)SvNV(fromstr);
-#else
+ /* VOS does not automatically map a floating-point overflow
+ during conversion from long double to double into infinity,
+ so we do it by hand. This code should either be generalized
+ for any OS that needs it, or removed if and when VOS
+ implements posix-976 (suggestion to support mapping to
+ infinity). Paul.Green@stratus.com 02-04-02. */
+{
+extern const double _double_constants[];
+ if (anv > DBL_MAX)
+ adouble = _double_constants[0]; /* double prec. inf. */
+ else if (anv < -DBL_MAX)
+ adouble = _double_constants[0]; /* double prec. inf. */
+ else adouble = (double) anv;
+}
+#else /* __VOS__ */
# if defined(VMS) && !defined(__IEEE_FP)
-/* IEEE fp overflow shenanigans are unavailable on VAX and optional
- * on Alpha; fake it if we don't have them.
- */
- if (SvNV(fromstr) > DBL_MAX)
- adouble = DBL_MAX;
- else if (SvNV(fromstr) < -DBL_MAX)
- adouble = -DBL_MAX;
- else adouble = (double)SvNV(fromstr);
+ /* IEEE fp overflow shenanigans are unavailable on VAX and optional
+ * on Alpha; fake it if we don't have them.
+ */
+ if (anv > DBL_MAX)
+ adouble = DBL_MAX;
+ else if (anv < -DBL_MAX)
+ adouble = -DBL_MAX;
+ else adouble = (double)anv;
# else
- adouble = (double)SvNV(fromstr);
+ adouble = (double)anv;
# endif
-#endif
+#endif /* __VOS__ */
DO_BO_PACK_N(adouble, double);
- sv_catpvn(cat, (char *)&adouble, sizeof (double));
+ PUSH_VAR(utf8, cur, adouble);
}
break;
- case 'F':
+ case 'F': {
+ NV_bytes anv;
Zero(&anv, 1, NV); /* can be long double with unused bits */
while (len-- > 0) {
fromstr = NEXTFROM;
- anv = SvNV(fromstr);
+ anv.nv = SvNV(fromstr);
DO_BO_PACK_N(anv, NV);
- sv_catpvn(cat, (char *)&anv, NVSIZE);
+ PUSH_BYTES(utf8, cur, anv.bytes, sizeof(anv.bytes));
}
break;
+ }
#if defined(HAS_LONG_DOUBLE) && defined(USE_LONG_DOUBLE)
- case 'D':
+ case 'D': {
+ ld_bytes aldouble;
/* long doubles can have unused bits, which may be nonzero */
Zero(&aldouble, 1, long double);
while (len-- > 0) {
fromstr = NEXTFROM;
- aldouble = (long double)SvNV(fromstr);
+ aldouble.ld = (long double)SvNV(fromstr);
DO_BO_PACK_N(aldouble, long double);
- sv_catpvn(cat, (char *)&aldouble, LONG_DOUBLESIZE);
+ PUSH_BYTES(utf8, cur, aldouble.bytes, sizeof(aldouble.bytes));
}
break;
+ }
#endif
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
case 'n' | TYPE_IS_SHRIEKING:
+#endif
case 'n':
while (len-- > 0) {
+ I16 ai16;
fromstr = NEXTFROM;
ai16 = (I16)SvIV(fromstr);
#ifdef HAS_HTONS
ai16 = PerlSock_htons(ai16);
#endif
- CAT16(cat, &ai16);
+ PUSH16(utf8, cur, &ai16);
}
break;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
case 'v' | TYPE_IS_SHRIEKING:
+#endif
case 'v':
while (len-- > 0) {
+ I16 ai16;
fromstr = NEXTFROM;
ai16 = (I16)SvIV(fromstr);
#ifdef HAS_HTOVS
ai16 = htovs(ai16);
#endif
- CAT16(cat, &ai16);
+ PUSH16(utf8, cur, &ai16);
}
break;
case 'S' | TYPE_IS_SHRIEKING:
#if SHORTSIZE != SIZE16
- {
- while (len-- > 0) {
- fromstr = NEXTFROM;
- aushort = SvUV(fromstr);
- DO_BO_PACK(aushort, s);
- sv_catpvn(cat, (char *)&aushort, sizeof(unsigned short));
- }
- }
+ while (len-- > 0) {
+ unsigned short aushort;
+ fromstr = NEXTFROM;
+ aushort = SvUV(fromstr);
+ DO_BO_PACK(aushort, s);
+ PUSH_VAR(utf8, cur, aushort);
+ }
break;
#else
/* Fall through! */
#endif
case 'S':
- {
- while (len-- > 0) {
- fromstr = NEXTFROM;
- au16 = (U16)SvUV(fromstr);
- DO_BO_PACK(au16, 16);
- CAT16(cat, &au16);
- }
-
+ while (len-- > 0) {
+ U16 au16;
+ fromstr = NEXTFROM;
+ au16 = (U16)SvUV(fromstr);
+ DO_BO_PACK(au16, 16);
+ PUSH16(utf8, cur, &au16);
}
break;
case 's' | TYPE_IS_SHRIEKING:
#if SHORTSIZE != SIZE16
- {
- while (len-- > 0) {
- fromstr = NEXTFROM;
- ashort = SvIV(fromstr);
- DO_BO_PACK(ashort, s);
- sv_catpvn(cat, (char *)&ashort, sizeof(short));
- }
+ while (len-- > 0) {
+ short ashort;
+ fromstr = NEXTFROM;
+ ashort = SvIV(fromstr);
+ DO_BO_PACK(ashort, s);
+ PUSH_VAR(utf8, cur, ashort);
}
break;
#else
#endif
case 's':
while (len-- > 0) {
+ I16 ai16;
fromstr = NEXTFROM;
ai16 = (I16)SvIV(fromstr);
DO_BO_PACK(ai16, 16);
- CAT16(cat, &ai16);
+ PUSH16(utf8, cur, &ai16);
}
break;
case 'I':
case 'I' | TYPE_IS_SHRIEKING:
while (len-- > 0) {
+ unsigned int auint;
fromstr = NEXTFROM;
auint = SvUV(fromstr);
DO_BO_PACK(auint, i);
- sv_catpvn(cat, (char*)&auint, sizeof(unsigned int));
+ PUSH_VAR(utf8, cur, auint);
}
break;
case 'j':
while (len-- > 0) {
+ IV aiv;
fromstr = NEXTFROM;
aiv = SvIV(fromstr);
#if IVSIZE == INTSIZE
DO_BO_PACK(aiv, l);
#elif defined(HAS_QUAD) && IVSIZE == U64SIZE
DO_BO_PACK(aiv, 64);
+#else
+ Perl_croak(aTHX_ "'j' not supported on this platform");
#endif
- sv_catpvn(cat, (char*)&aiv, IVSIZE);
+ PUSH_VAR(utf8, cur, aiv);
}
break;
case 'J':
while (len-- > 0) {
+ UV auv;
fromstr = NEXTFROM;
auv = SvUV(fromstr);
#if UVSIZE == INTSIZE
DO_BO_PACK(auv, l);
#elif defined(HAS_QUAD) && UVSIZE == U64SIZE
DO_BO_PACK(auv, 64);
+#else
+ Perl_croak(aTHX_ "'J' not supported on this platform");
#endif
- sv_catpvn(cat, (char*)&auv, UVSIZE);
+ PUSH_VAR(utf8, cur, auv);
}
break;
case 'w':
while (len-- > 0) {
+ NV anv;
fromstr = NEXTFROM;
anv = SvNV(fromstr);
- if (anv < 0)
+ if (anv < 0) {
+ *cur = '\0';
+ SvCUR_set(cat, cur - start);
Perl_croak(aTHX_ "Cannot compress negative numbers in pack");
+ }
/* 0xFFFFFFFFFFFFFFFF may cast to 18446744073709551616.0,
which is == UV_MAX_P1. IOK is fine (instead of UV_only), as
any negative IVs will have already been got by the croak()
above. IOK is untrue for fractions, so we test them
against UV_MAX_P1. */
- if (SvIOK(fromstr) || anv < UV_MAX_P1)
- {
- char buf[(sizeof(UV)*8)/7+1];
+ if (SvIOK(fromstr) || anv < UV_MAX_P1) {
+ char buf[(sizeof(UV)*CHAR_BIT)/7+1];
char *in = buf + sizeof(buf);
UV auv = SvUV(fromstr);
auv >>= 7;
} while (auv);
buf[sizeof(buf) - 1] &= 0x7f; /* clear continue bit */
- sv_catpvn(cat, in, (buf + sizeof(buf)) - in);
- }
- else if (SvPOKp(fromstr)) { /* decimal string arithmetics */
- char *from, *result, *in;
- SV *norm;
- STRLEN len;
- bool done;
-
- /* Copy string and check for compliance */
- from = SvPV(fromstr, len);
- if ((norm = is_an_int(from, len)) == NULL)
- Perl_croak(aTHX_ "Can only compress unsigned integers in pack");
-
- New('w', result, len, char);
- in = result + len;
- done = FALSE;
- while (!done)
- *--in = div128(norm, &done) | 0x80;
- result[len - 1] &= 0x7F; /* clear continue bit */
- sv_catpvn(cat, in, (result + len) - in);
- Safefree(result);
- SvREFCNT_dec(norm); /* free norm */
- }
+ PUSH_GROWING_BYTES(utf8, cat, start, cur,
+ in, (buf + sizeof(buf)) - in);
+ } else if (SvPOKp(fromstr))
+ goto w_string;
else if (SvNOKp(fromstr)) {
/* 10**NV_MAX_10_EXP is the largest power of 10
so 10**(NV_MAX_10_EXP+1) is definately unrepresentable
floating-point value.
*/
#ifdef NV_MAX_10_EXP
-/* char buf[1 + (int)((NV_MAX_10_EXP + 1) * 0.47456)]; -- invalid C */
+ /* char buf[1 + (int)((NV_MAX_10_EXP + 1) * 0.47456)]; -- invalid C */
char buf[1 + (int)((NV_MAX_10_EXP + 1) / 2)]; /* valid C */
#else
-/* char buf[1 + (int)((308 + 1) * 0.47456)]; -- invalid C */
+ /* char buf[1 + (int)((308 + 1) * 0.47456)]; -- invalid C */
char buf[1 + (int)((308 + 1) / 2)]; /* valid C */
#endif
char *in = buf + sizeof(buf);
- anv = Perl_floor(anv);
+ anv = Perl_floor(anv);
do {
- NV next = Perl_floor(anv / 128);
+ const NV next = Perl_floor(anv / 128);
if (in <= buf) /* this cannot happen ;-) */
Perl_croak(aTHX_ "Cannot compress integer in pack");
*--in = (unsigned char)(anv - (next * 128)) | 0x80;
anv = next;
} while (anv > 0);
buf[sizeof(buf) - 1] &= 0x7f; /* clear continue bit */
- sv_catpvn(cat, in, (buf + sizeof(buf)) - in);
- }
- else {
- char *from, *result, *in;
+ PUSH_GROWING_BYTES(utf8, cat, start, cur,
+ in, (buf + sizeof(buf)) - in);
+ } else {
+ const char *from;
+ char *result, *in;
SV *norm;
STRLEN len;
bool done;
+ w_string:
/* Copy string and check for compliance */
- from = SvPV(fromstr, len);
+ from = SvPV_const(fromstr, len);
if ((norm = is_an_int(from, len)) == NULL)
Perl_croak(aTHX_ "Can only compress unsigned integers in pack");
- New('w', result, len, char);
+ Newx(result, len, char);
in = result + len;
done = FALSE;
- while (!done)
- *--in = div128(norm, &done) | 0x80;
+ while (!done) *--in = div128(norm, &done) | 0x80;
result[len - 1] &= 0x7F; /* clear continue bit */
- sv_catpvn(cat, in, (result + len) - in);
+ PUSH_GROWING_BYTES(utf8, cat, start, cur,
+ in, (result + len) - in);
Safefree(result);
SvREFCNT_dec(norm); /* free norm */
- }
+ }
}
break;
case 'i':
case 'i' | TYPE_IS_SHRIEKING:
while (len-- > 0) {
+ int aint;
fromstr = NEXTFROM;
aint = SvIV(fromstr);
DO_BO_PACK(aint, i);
- sv_catpvn(cat, (char*)&aint, sizeof(int));
+ PUSH_VAR(utf8, cur, aint);
}
break;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
case 'N' | TYPE_IS_SHRIEKING:
+#endif
case 'N':
while (len-- > 0) {
+ U32 au32;
fromstr = NEXTFROM;
au32 = SvUV(fromstr);
#ifdef HAS_HTONL
au32 = PerlSock_htonl(au32);
#endif
- CAT32(cat, &au32);
+ PUSH32(utf8, cur, &au32);
}
break;
+#ifdef PERL_PACK_CAN_SHRIEKSIGN
case 'V' | TYPE_IS_SHRIEKING:
+#endif
case 'V':
while (len-- > 0) {
+ U32 au32;
fromstr = NEXTFROM;
au32 = SvUV(fromstr);
#ifdef HAS_HTOVL
au32 = htovl(au32);
#endif
- CAT32(cat, &au32);
+ PUSH32(utf8, cur, &au32);
}
break;
case 'L' | TYPE_IS_SHRIEKING:
#if LONGSIZE != SIZE32
- {
- while (len-- > 0) {
- fromstr = NEXTFROM;
- aulong = SvUV(fromstr);
- DO_BO_PACK(aulong, l);
- sv_catpvn(cat, (char *)&aulong, sizeof(unsigned long));
- }
+ while (len-- > 0) {
+ unsigned long aulong;
+ fromstr = NEXTFROM;
+ aulong = SvUV(fromstr);
+ DO_BO_PACK(aulong, l);
+ PUSH_VAR(utf8, cur, aulong);
}
break;
#else
/* Fall though! */
#endif
case 'L':
- {
- while (len-- > 0) {
- fromstr = NEXTFROM;
- au32 = SvUV(fromstr);
- DO_BO_PACK(au32, 32);
- CAT32(cat, &au32);
- }
+ while (len-- > 0) {
+ U32 au32;
+ fromstr = NEXTFROM;
+ au32 = SvUV(fromstr);
+ DO_BO_PACK(au32, 32);
+ PUSH32(utf8, cur, &au32);
}
break;
case 'l' | TYPE_IS_SHRIEKING:
#if LONGSIZE != SIZE32
- {
- while (len-- > 0) {
- fromstr = NEXTFROM;
- along = SvIV(fromstr);
- DO_BO_PACK(along, l);
- sv_catpvn(cat, (char *)&along, sizeof(long));
- }
+ while (len-- > 0) {
+ long along;
+ fromstr = NEXTFROM;
+ along = SvIV(fromstr);
+ DO_BO_PACK(along, l);
+ PUSH_VAR(utf8, cur, along);
}
break;
#else
#endif
case 'l':
while (len-- > 0) {
+ I32 ai32;
fromstr = NEXTFROM;
ai32 = SvIV(fromstr);
DO_BO_PACK(ai32, 32);
- CAT32(cat, &ai32);
+ PUSH32(utf8, cur, &ai32);
}
break;
#ifdef HAS_QUAD
case 'Q':
while (len-- > 0) {
+ Uquad_t auquad;
fromstr = NEXTFROM;
- auquad = (Uquad_t)SvUV(fromstr);
+ auquad = (Uquad_t) SvUV(fromstr);
DO_BO_PACK(auquad, 64);
- sv_catpvn(cat, (char*)&auquad, sizeof(Uquad_t));
+ PUSH_VAR(utf8, cur, auquad);
}
break;
case 'q':
while (len-- > 0) {
+ Quad_t aquad;
fromstr = NEXTFROM;
aquad = (Quad_t)SvIV(fromstr);
DO_BO_PACK(aquad, 64);
- sv_catpvn(cat, (char*)&aquad, sizeof(Quad_t));
+ PUSH_VAR(utf8, cur, aquad);
}
break;
-#endif
+#endif /* HAS_QUAD */
case 'P':
len = 1; /* assume SV is correct length */
+ GROWING(utf8, cat, start, cur, sizeof(char *));
/* Fall through! */
case 'p':
while (len-- > 0) {
+ const char *aptr;
+
fromstr = NEXTFROM;
- if (fromstr == &PL_sv_undef)
- aptr = NULL;
+ SvGETMAGIC(fromstr);
+ if (!SvOK(fromstr)) aptr = NULL;
else {
- STRLEN n_a;
/* XXX better yet, could spirit away the string to
* a safe spot and hang on to it until the result
* of pack() (and all copies of the result) are
* gone.
*/
- if (ckWARN(WARN_PACK) && (SvTEMP(fromstr)
- || (SvPADTMP(fromstr)
- && !SvREADONLY(fromstr))))
- {
- Perl_warner(aTHX_ packWARN(WARN_PACK),
- "Attempt to pack pointer to temporary value");
+ if ((SvTEMP(fromstr) || (SvPADTMP(fromstr) &&
+ !SvREADONLY(fromstr)))) {
+ Perl_ck_warner(aTHX_ packWARN(WARN_PACK),
+ "Attempt to pack pointer to temporary value");
}
if (SvPOK(fromstr) || SvNIOK(fromstr))
- aptr = SvPV(fromstr,n_a);
+ aptr = SvPV_nomg_const_nolen(fromstr);
else
- aptr = SvPV_force(fromstr,n_a);
+ aptr = SvPV_force_flags_nolen(fromstr, 0);
}
- DO_BO_PACK_P(aptr);
- sv_catpvn(cat, (char*)&aptr, sizeof(char*));
+ DO_BO_PACK_PC(aptr);
+ PUSH_VAR(utf8, cur, aptr);
}
break;
- case 'u':
+ case 'u': {
+ const char *aptr, *aend;
+ bool from_utf8;
+
fromstr = NEXTFROM;
- aptr = SvPV(fromstr, fromlen);
- SvGROW(cat, fromlen * 4 / 3);
- if (len <= 2)
- len = 45;
- else
- len = len / 3 * 3;
+ if (len <= 2) len = 45;
+ else len = len / 3 * 3;
+ if (len >= 64) {
+ Perl_ck_warner(aTHX_ packWARN(WARN_PACK),
+ "Field too wide in 'u' format in pack");
+ len = 63;
+ }
+ aptr = SvPV_const(fromstr, fromlen);
+ from_utf8 = DO_UTF8(fromstr);
+ if (from_utf8) {
+ aend = aptr + fromlen;
+ fromlen = sv_len_utf8(fromstr);
+ } else aend = NULL; /* Unused, but keep compilers happy */
+ GROWING(utf8, cat, start, cur, (fromlen+2) / 3 * 4 + (fromlen+len-1)/len * 2);
while (fromlen > 0) {
+ U8 *end;
I32 todo;
+ U8 hunk[1+63/3*4+1];
if ((I32)fromlen > len)
todo = len;
else
todo = fromlen;
- doencodes(cat, aptr, todo);
+ if (from_utf8) {
+ char buffer[64];
+ if (!uni_to_bytes(aTHX_ &aptr, aend, buffer, todo,
+ 'u' | TYPE_IS_PACK)) {
+ *cur = '\0';
+ SvCUR_set(cat, cur - start);
+ Perl_croak(aTHX_ "panic: string is shorter than advertised");
+ }
+ end = doencodes(hunk, buffer, todo);
+ } else {
+ end = doencodes(hunk, aptr, todo);
+ aptr += todo;
+ }
+ PUSH_BYTES(utf8, cur, hunk, end-hunk);
fromlen -= todo;
- aptr += todo;
}
break;
}
+ }
+ *cur = '\0';
+ SvCUR_set(cat, cur - start);
+ no_change:
*symptr = lookahead;
}
return beglist;
PP(pp_pack)
{
- dSP; dMARK; dORIGMARK; dTARGET;
+ dVAR; dSP; dMARK; dORIGMARK; dTARGET;
register SV *cat = TARG;
STRLEN fromlen;
- register char *pat = SvPVx(*++MARK, fromlen);
- register char *patend = pat + fromlen;
+ SV *pat_sv = *++MARK;
+ register const char *pat = SvPV_const(pat_sv, fromlen);
+ register const char *patend = pat + fromlen;
MARK++;
- sv_setpvn(cat, "", 0);
+ sv_setpvs(cat, "");
+ SvUTF8_off(cat);
packlist(cat, pat, patend, MARK, SP + 1);
* indent-tabs-mode: t
* End:
*
- * vim: shiftwidth=4:
-*/
+ * ex: set ts=8 sts=4 sw=4 noet:
+ */