static const char unees[] =
"Malformed UTF-8 character (unexpected end of string)";
static const char cp_above_legal_max[] =
- "Use of code point 0x%" UVXf " is deprecated; the permissible max is 0x%" UVXf ". This will be fatal in Perl 5.28";
+ "Use of code point 0x%" UVXf " is not allowed; the"
+ " permissible max is 0x%" UVXf;
-#define MAX_NON_DEPRECATED_CP ((UV) (IV_MAX))
+#define MAX_EXTERNALLY_LEGAL_CP ((UV) (IV_MAX))
/*
=head1 Unicode Support
* performance hit on these high EBCDIC code points. */
if (UNLIKELY(UNICODE_IS_SUPER(uv))) {
- if ( UNLIKELY(uv > MAX_NON_DEPRECATED_CP)
- && ckWARN_d(WARN_DEPRECATED))
- {
- Perl_warner(aTHX_ packWARN(WARN_DEPRECATED),
- cp_above_legal_max, uv, MAX_NON_DEPRECATED_CP);
+ if (UNLIKELY(uv > MAX_EXTERNALLY_LEGAL_CP)) {
+ Perl_croak(aTHX_ cp_above_legal_max, uv, MAX_EXTERNALLY_LEGAL_CP);
}
if ( (flags & UNICODE_WARN_SUPER)
|| ( (flags & UNICODE_WARN_PERL_EXTENDED)
*(d++) = uv;
-This function accepts any UV as input, but very high code points (above
-C<IV_MAX> on the platform) will raise a deprecation warning. This is
-typically 0x7FFF_FFFF in a 32-bit word.
+This function accepts any code point from 0..C<IV_MAX> as input.
+C<IV_MAX> is typically 0x7FFF_FFFF in a 32-bit word.
It is possible to forbid or warn on non-Unicode code points, or those that may
be problematic by using L</uvchr_to_utf8_flags>.
*(d++) = uv;
-If C<flags> is 0, this function accepts any UV as input, but very high code
-points (above C<IV_MAX> for the platform) will raise a deprecation warning.
-This is typically 0x7FFF_FFFF in a 32-bit word.
+If C<flags> is 0, this function accepts any code point from 0..C<IV_MAX> as
+input. C<IV_MAX> is typically 0x7FFF_FFFF in a 32-bit word.
Specifying C<flags> can further restrict what is allowed and not warned on, as
follows:
return -1;
}
-/* Anything larger than this will overflow the word if it were converted into a UV */
-#if defined(UV_IS_QUAD)
+#if defined(UV_IS_QUAD) /* These assume IV_MAX is 2**63-1 */
# ifdef EBCDIC /* Actually is I8 */
# define HIGHEST_REPRESENTABLE_UTF8 \
- "\xFF\xAF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF"
+ "\xFF\xA7\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF"
# else
# define HIGHEST_REPRESENTABLE_UTF8 \
- "\xFF\x80\x8F\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF"
-# endif
-#else /* 32-bit */
-# ifdef EBCDIC
-# define HIGHEST_REPRESENTABLE_UTF8 \
- "\xFF\xA0\xA0\xA0\xA0\xA0\xA0\xA3\xBF\xBF\xBF\xBF\xBF\xBF"
-# else
-# define HIGHEST_REPRESENTABLE_UTF8 "\xFE\x83\xBF\xBF\xBF\xBF\xBF"
+ "\xFF\x80\x87\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF"
# endif
#endif
const bool consider_overlongs)
{
/* Returns an int indicating whether or not the UTF-8 sequence from 's' to
- * 'e' - 1 would overflow a UV on this platform; that is if it represents a
- * code point larger than the highest representable code point. It returns
- * 1 if it does overflow; 0 if it doesn't, and -1 if there isn't enough
- * information to tell. This last return value can happen if the sequence
- * is incomplete, missing some trailing bytes that would form a complete
- * character. If there are enough bytes to make a definitive decision,
- * this function does so.
+ * 'e' - 1 would overflow an IV on this platform; that is if it represents
+ * a code point larger than the highest representable code point. It
+ * returns 1 if it does overflow; 0 if it doesn't, and -1 if there isn't
+ * enough information to tell. This last return value can happen if the
+ * sequence is incomplete, missing some trailing bytes that would form a
+ * complete character. If there are enough bytes to make a definitive
+ * decision, this function does so.
*
* If 'consider_overlongs' is TRUE, the function checks for the possibility
* that the sequence is an overlong that doesn't overflow. Otherwise, it
*
* 'e' - 1 must not be beyond a full character. */
- const STRLEN len = e - s;
- const U8 *x;
- const U8 * y = (const U8 *) HIGHEST_REPRESENTABLE_UTF8;
PERL_ARGS_ASSERT_DOES_UTF8_OVERFLOW;
assert(s <= e && s + UTF8SKIP(s) >= e);
+#if ! defined(UV_IS_QUAD)
+
+ return is_utf8_cp_above_31_bits(s, e, consider_overlongs);
+
+#else
+
+ PERL_UNUSED_ARG(consider_overlongs);
+
+ {
+ const STRLEN len = e - s;
+ const U8 *x;
+ const U8 * y = (const U8 *) HIGHEST_REPRESENTABLE_UTF8;
+
+ for (x = s; x < e; x++, y++) {
+
+ if (UNLIKELY(NATIVE_UTF8_TO_I8(*x) == *y)) {
+ continue;
+ }
+
+ /* If this byte is larger than the corresponding highest UTF-8
+ * byte, the sequence overflow; otherwise the byte is less than,
+ * and so the sequence doesn't overflow */
+ return NATIVE_UTF8_TO_I8(*x) > *y;
+
+ }
+
+ /* Got to the end and all bytes are the same. If the input is a whole
+ * character, it doesn't overflow. And if it is a partial character,
+ * there's not enough information to tell */
+ if (len < sizeof(HIGHEST_REPRESENTABLE_UTF8) - 1) {
+ return -1;
+ }
+
+ return 0;
+ }
+
+#endif
+
+}
+
+#if 0
+
+/* This is the portions of the above function that deal with UV_MAX instead of
+ * IV_MAX. They are left here in case we want to combine them so that internal
+ * uses can have larger code points. The only logic difference is that the
+ * 32-bit EBCDIC platform is treate like the 64-bit, and the 32-bit ASCII has
+ * different logic.
+ */
+
+/* Anything larger than this will overflow the word if it were converted into a UV */
+#if defined(UV_IS_QUAD)
+# ifdef EBCDIC /* Actually is I8 */
+# define HIGHEST_REPRESENTABLE_UTF8 \
+ "\xFF\xAF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF"
+# else
+# define HIGHEST_REPRESENTABLE_UTF8 \
+ "\xFF\x80\x8F\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF\xBF"
+# endif
+#else /* 32-bit */
+# ifdef EBCDIC
+# define HIGHEST_REPRESENTABLE_UTF8 \
+ "\xFF\xA0\xA0\xA0\xA0\xA0\xA0\xA3\xBF\xBF\xBF\xBF\xBF\xBF"
+# else
+# define HIGHEST_REPRESENTABLE_UTF8 "\xFE\x83\xBF\xBF\xBF\xBF\xBF"
+# endif
+#endif
+
#if ! defined(UV_IS_QUAD) && ! defined(EBCDIC)
/* On 32 bit ASCII machines, many overlongs that start with FF don't
return s[sizeof(FF_OVERLONG_PREFIX) - 1] > 0x83;
}
-#else
-
- PERL_UNUSED_ARG(consider_overlongs);
-
+/* Starting with the #else, the rest of the function is identical except
+ * 1. we need to move the 'len' declaration to be global to the function
+ * 2. the endif move to just after the UNUSED_ARG.
+ * An empty endif is given just below to satisfy the preprocessor
+ */
#endif
- for (x = s; x < e; x++, y++) {
-
- if (UNLIKELY(NATIVE_UTF8_TO_I8(*x) == *y)) {
- continue;
- }
-
- /* If this byte is larger than the corresponding highest UTF-8 byte,
- * the sequence overflow; otherwise the byte is less than, and so the
- * sequence doesn't overflow */
- return NATIVE_UTF8_TO_I8(*x) > *y;
-
- }
-
- /* Got to the end and all bytes are the same. If the input is a whole
- * character, it doesn't overflow. And if it is a partial character,
- * there's not enough information to tell */
- if (len < sizeof(HIGHEST_REPRESENTABLE_UTF8) - 1) {
- return -1;
- }
-
- return 0;
-}
+#endif
#undef F0_ABOVE_OVERLONG
#undef F8_ABOVE_OVERLONG
}
char *
-Perl__byte_dump_string(pTHX_ const U8 * s, const STRLEN len, const bool format)
+Perl__byte_dump_string(pTHX_ const U8 * const start, const STRLEN len, const bool format)
{
/* Returns a mortalized C string that is a displayable copy of the 'len'
- * bytes starting at 's'. 'format' gives how to display each byte.
+ * bytes starting at 'start'. 'format' gives how to display each byte.
* Currently, there are only two formats, so it is currently a bool:
* 0 \xab
* 1 ab (that is a space between two hex digit bytes)
const STRLEN output_len = 4 * len + 1; /* 4 bytes per each input, plus a
trailing NUL */
- const U8 * const e = s + len;
+ const U8 * s = start;
+ const U8 * const e = start + len;
char * output;
char * d;
SAVEFREEPV(output);
d = output;
- for (; s < e; s++) {
+ for (s = start; s < e; s++) {
const unsigned high_nibble = (*s & 0xF0) >> 4;
const unsigned low_nibble = (*s & 0x0F);
if (format) {
- *d++ = ' ';
+ if (s > start) {
+ *d++ = ' ';
+ }
}
else {
*d++ = '\\';
malformation|/C<UTF8_GOT_LONG>> is also present. The new names accurately
describe the situation in all cases.
-It is now deprecated to have very high code points (above C<IV_MAX> on the
-platforms) and this function will raise a deprecation warning for these (unless
-such warnings are turned off). This value is typically 0x7FFF_FFFF (2**31 -1)
-in a 32-bit word.
All other code points corresponding to Unicode characters, including private
use and those yet to be assigned, are never considered malformed and never
=item C<UTF8_GOT_OVERFLOW>
The input sequence was malformed in that it is for a code point that is not
-representable in the number of bits available in a UV on the current platform.
+representable in the number of bits available in an IV on the current platform.
=item C<UTF8_GOT_SHORT>
|UTF8_WARN_NONCHAR
|UTF8_WARN_SURROGATE
|UTF8_WARN_SUPER
- |UTF8_WARN_PERL_EXTENDED))
- /* In case of a malformation, 'uv' is not valid, and has
- * been changed to something in the Unicode range.
- * Currently we don't output a deprecation message if there
- * is already a malformation, so we don't have to special
- * case the test immediately below */
- || ( UNLIKELY(uv > MAX_NON_DEPRECATED_CP)
- && ckWARN_d(WARN_DEPRECATED))))
+ |UTF8_WARN_PERL_EXTENDED))))
{
/* If there were no malformations, or the only malformation is an
* overlong, 'uv' is valid */
disallowed = TRUE;
}
- /* Likewise, warn if any say to, plus if deprecation warnings
- * are on, because this code point is above IV_MAX */
- if ( ckWARN_d(WARN_DEPRECATED)
- || ! (flags & UTF8_ALLOW_OVERFLOW)
- || (flags & (UTF8_WARN_SUPER|UTF8_WARN_PERL_EXTENDED)))
+ /* Likewise, warn if any say to */
+ if ( ! (flags & UTF8_ALLOW_OVERFLOW)
+ || (flags & (UTF8_WARN_SUPER|UTF8_WARN_PERL_EXTENDED)))
{
/* The warnings code explicitly says it doesn't handle the
*errors |= UTF8_GOT_SUPER;
disallowed = TRUE;
}
-
- /* The deprecated warning overrides any non-deprecated one. If
- * there are other problems, a deprecation message is not
- * really helpful, so don't bother to raise it in that case.
- * This also keeps the code from having to handle the case
- * where 'uv' is not valid. */
- if ( ! (orig_problems
- & (UTF8_GOT_TOO_SHORT|UTF8_GOT_OVERFLOW))
- && UNLIKELY(uv > MAX_NON_DEPRECATED_CP)
- && ckWARN_d(WARN_DEPRECATED))
- {
- message = Perl_form(aTHX_ cp_above_legal_max,
- uv, MAX_NON_DEPRECATED_CP);
- pack_warn = packWARN(WARN_DEPRECATED);
- }
}
else if (possible_problems & UTF8_GOT_NONCHAR) {
possible_problems &= ~UTF8_GOT_NONCHAR;
U8 tmpbuf[UTF8_MAXBYTES+1];
const U8 * const e = uvoffuni_to_utf8_flags(tmpbuf,
uv, 0);
- const char * preface = (uv <= PERL_UNICODE_MAX)
- ? "U+"
- : "0x";
+ /* Don't use U+ for non-Unicode code points, which
+ * includes those in the Latin1 range */
+ const char * preface = ( uv > PERL_UNICODE_MAX
+#ifdef EBCDIC
+ || uv <= 0xFF
+#endif
+ )
+ ? "0x"
+ : "U+";
message = Perl_form(aTHX_
"%s: %s (overlong; instead use %s to represent"
" %s%0*" UVXf ")",
preface,
((uv < 256) ? 2 : 4), /* Field width of 2 for
small code points */
- uv);
+ UNI_TO_NATIVE(uv));
}
}
}
See L</utf8n_to_uvchr> for details on when the REPLACEMENT CHARACTER is
returned.
-Code points above the platform's C<IV_MAX> will raise a deprecation warning,
-unless those are turned off.
-
=cut
Also implemented as a macro in utf8.h
next possible position in C<s> that could begin a non-malformed character.
See L</utf8n_to_uvchr> for details on when the REPLACEMENT CHARACTER is returned.
-Code points above the platform's C<IV_MAX> will raise a deprecation warning,
-unless those are turned off.
-
=cut
*/
assert(send > s);
- /* Call the low level routine, asking for checks */
return NATIVE_TO_UNI(utf8_to_uvchr_buf(s, send, retlen));
}
/*
=for apidoc utf8_length
-Return the length of the UTF-8 char encoded string C<s> in characters.
-Stops at C<e> (inclusive). If C<e E<lt> s> or if the scan would end
-up past C<e>, croaks.
+Returns the number of characters in the sequence of UTF-8-encoded bytes starting
+at C<s> and ending at the byte just before C<e>. If <s> and <e> point to the
+same place, it returns 0 with no warning raised.
+
+If C<e E<lt> s> or if the scan would end up past C<e>, it raises a UTF8 warning
+and returns the number of valid characters.
=cut
*/
}
finish_and_return:
- *d = '\0';
- *lenp = d - converted_start;
+ *d = '\0';
+ *lenp = d - converted_start;
/* Trim unused space */
Renew(converted_start, *lenp + 1, U8);
append_utf8_from_native_byte(*s, &d);
s++;
}
+
*d = '\0';
*lenp = d-dst;
+
+ /* Trim unused space */
+ Renew(dst, *lenp + 1, U8);
+
return dst;
}
/*
- * Convert native (big-endian) or reversed (little-endian) UTF-16 to UTF-8.
+ * Convert native (big-endian) UTF-16 to UTF-8. For reversed (little-endian),
+ * use utf16_to_utf8_reversed().
*
- * Destination must be pre-extended to 3/2 source. Do not use in-place.
- * We optimize for native, for obvious reasons. */
+ * UTF-16 requires 2 bytes for every code point below 0x10000; otherwise 4 bytes.
+ * UTF-8 requires 1-3 bytes for every code point below 0x1000; otherwise 4 bytes.
+ * UTF-EBCDIC requires 1-4 bytes for every code point below 0x1000; otherwise 4-5 bytes.
+ *
+ * These functions don't check for overflow. The worst case is every code
+ * point in the input is 2 bytes, and requires 4 bytes on output. (If the code
+ * is never going to run in EBCDIC, it is 2 bytes requiring 3 on output.) Therefore the
+ * destination must be pre-extended to 2 times the source length.
+ *
+ * Do not use in-place. We optimize for native, for obvious reasons. */
U8*
Perl_utf16_to_utf8(pTHX_ U8* p, U8* d, I32 bytelen, I32 *newlen)
*d++ = UTF8_TWO_BYTE_LO(UNI_TO_NATIVE(uv));
continue;
}
+
#define FIRST_HIGH_SURROGATE UNICODE_SURROGATE_FIRST
#define LAST_HIGH_SURROGATE 0xDBFF
#define FIRST_LOW_SURROGATE 0xDC00
#define LAST_LOW_SURROGATE UNICODE_SURROGATE_LAST
+#define FIRST_IN_PLANE1 0x10000
/* This assumes that most uses will be in the first Unicode plane, not
* needing surrogates */
}
p += 2;
uv = ((uv - FIRST_HIGH_SURROGATE) << 10)
- + (low - FIRST_LOW_SURROGATE) + 0x10000;
+ + (low - FIRST_LOW_SURROGATE) + FIRST_IN_PLANE1;
}
}
#ifdef EBCDIC
d = uvoffuni_to_utf8_flags(d, uv, 0);
#else
- if (uv < 0x10000) {
+ if (uv < FIRST_IN_PLANE1) {
*d++ = (U8)(( uv >> 12) | 0xe0);
*d++ = (U8)(((uv >> 6) & 0x3f) | 0x80);
*d++ = (U8)(( uv & 0x3f) | 0x80);
}
if (UNLIKELY(UNICODE_IS_SUPER(uv1))) {
- if ( UNLIKELY(uv1 > MAX_NON_DEPRECATED_CP)
- && ckWARN_d(WARN_DEPRECATED))
- {
- Perl_warner(aTHX_ packWARN(WARN_DEPRECATED),
- cp_above_legal_max, uv1, MAX_NON_DEPRECATED_CP);
+ if (UNLIKELY(uv1 > MAX_EXTERNALLY_LEGAL_CP)) {
+ Perl_croak(aTHX_ cp_above_legal_max, uv1,
+ MAX_EXTERNALLY_LEGAL_CP);
}
if (ckWARN_d(WARN_NON_UNICODE)) {
const char* desc = (PL_op) ? OP_DESC(PL_op) : normal;
if (flags & FOLD_FLAGS_LOCALE) {
# define LONG_S_T LATIN_SMALL_LIGATURE_LONG_S_T_UTF8
- const unsigned int long_s_t_len = sizeof(LONG_S_T) - 1;
-
# ifdef LATIN_CAPITAL_LETTER_SHARP_S_UTF8
# define CAP_SHARP_S LATIN_CAPITAL_LETTER_SHARP_S_UTF8
- const unsigned int cap_sharp_s_len = sizeof(CAP_SHARP_S) - 1;
-
/* Special case these two characters, as what normally gets
* returned under locale doesn't work */
- if (UTF8SKIP(p) == cap_sharp_s_len
- && memEQ((char *) p, CAP_SHARP_S, cap_sharp_s_len))
+ if (memEQs((char *) p, UTF8SKIP(p), CAP_SHARP_S))
{
/* diag_listed_as: Can't do %s("%s") on non-UTF-8 locale; resolved to "%s". */
Perl_ck_warner(aTHX_ packWARN(WARN_LOCALE),
}
else
#endif
- if (UTF8SKIP(p) == long_s_t_len
- && memEQ((char *) p, LONG_S_T, long_s_t_len))
+ if (memEQs((char *) p, UTF8SKIP(p), LONG_S_T))
{
/* diag_listed_as: Can't do %s("%s") on non-UTF-8 locale; resolved to "%s". */
Perl_ck_warner(aTHX_ packWARN(WARN_LOCALE),
* 255/256 boundary which is forbidden under /l, and so the code
* wouldn't catch that they are equivalent (which they are only in
* this release) */
- else if (UTF8SKIP(p) == sizeof(DOTTED_I) - 1
- && memEQ((char *) p, DOTTED_I, sizeof(DOTTED_I) - 1))
- {
+ else if (memEQs((char *) p, UTF8SKIP(p), DOTTED_I)) {
/* diag_listed_as: Can't do %s("%s") on non-UTF-8 locale; resolved to "%s". */
Perl_ck_warner(aTHX_ packWARN(WARN_LOCALE),
"Can't do fc(\"\\x{0130}\") on non-UTF-8 locale; "
* size based on worst possible case, which is each line in the input
* creates 2 elements in the inversion list: 1) the beginning of a
* range in the list; 2) the beginning of a range not in the list. */
- while ((loc = (strchr(loc, '\n'))) != NULL) {
+ while ((loc = (char *) memchr(loc, '\n', lend - (U8 *) loc)) != NULL) {
elements += 2;
loc++;
}
/* May change: warns if surrogates, non-character code points, or
* non-Unicode code points are in 's' which has length 'len' bytes.
* Returns TRUE if none found; FALSE otherwise. The only other validity
- * check is to make sure that this won't exceed the string's length.
- *
- * Code points above the platform's C<IV_MAX> will raise a deprecation
- * warning, unless those are turned off. */
+ * check is to make sure that this won't exceed the string's length nor
+ * overflow */
const U8* const e = s + len;
bool ok = TRUE;