This function is for code that wants any warning and/or error messages to be
returned to the caller rather than be displayed. All messages that would have
-been displayed if all lexcial warnings are enabled will be returned.
+been displayed if all lexical warnings are enabled will be returned.
It is just like C<L</uvchr_to_utf8_flags>> but it takes an extra parameter
placed after all the others, C<msgs>. If this parameter is 0, this function
STRLEN len = OFFUNISKIP(uv);
U8 *p = d+len-1;
while (p > d) {
- *p-- = I8_TO_NATIVE_UTF8((uv & UTF_CONTINUATION_MASK) | UTF_CONTINUATION_MARK);
- uv >>= UTF_ACCUMULATION_SHIFT;
+ *p-- = I8_TO_NATIVE_UTF8((uv & MASK) | MARK);
+ uv >>= SHIFT;
}
*p = I8_TO_NATIVE_UTF8((uv & UTF_START_MASK(len)) | UTF_START_MARK(len));
return d+len;
PERL_STATIC_INLINE char *
S_unexpected_non_continuation_text(pTHX_ const U8 * const s,
- /* How many bytes to print */
+ /* Max number of bytes to print */
STRLEN print_len,
/* Which one is the non-continuation */
? "immediately"
: Perl_form(aTHX_ "%d bytes",
(int) non_cont_byte_pos);
+ const U8 * x = s + non_cont_byte_pos;
+ const U8 * e = s + print_len;
PERL_ARGS_ASSERT_UNEXPECTED_NON_CONTINUATION_TEXT;
* calculated, it's likely faster to pass it; verify under DEBUGGING */
assert(expect_len == UTF8SKIP(s));
+ /* As a defensive coding measure, don't output anything past a NUL. Such
+ * bytes shouldn't be in the middle of a malformation, and could mark the
+ * end of the allocated string, and what comes after is undefined */
+ for (; x < e; x++) {
+ if (*x == '\0') {
+ x++; /* Output this particular NUL */
+ break;
+ }
+ }
+
return Perl_form(aTHX_ "%s: %s (unexpected non-continuation byte 0x%02x,"
" %s after start byte 0x%02x; need %d bytes, got %d)",
malformed_text,
- _byte_dump_string(s, print_len, 0),
+ _byte_dump_string(s, x - s, 0),
*(s + non_cont_byte_pos),
where,
*s,
*/
UV
-Perl_utf8n_to_uvchr(pTHX_ const U8 *s,
- STRLEN curlen,
- STRLEN *retlen,
- const U32 flags)
+Perl_utf8n_to_uvchr(const U8 *s,
+ STRLEN curlen,
+ STRLEN *retlen,
+ const U32 flags)
{
PERL_ARGS_ASSERT_UTF8N_TO_UVCHR;
return utf8n_to_uvchr_error(s, curlen, retlen, flags, NULL);
}
-/* The tables below come from http://bjoern.hoehrmann.de/utf-8/decoder/dfa/,
- * which requires this copyright notice */
-
-/* Copyright (c) 2008-2009 Bjoern Hoehrmann <bjoern@hoehrmann.de>
-
-Permission is hereby granted, free of charge, to any person obtaining a copy of
-this software and associated documentation files (the "Software"), to deal in
-the Software without restriction, including without limitation the rights to
-use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies
-of the Software, and to permit persons to whom the Software is furnished to do
-so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
-
-*/
-
-#if 0
-static U8 utf8d_C9[] = {
- /* The first part of the table maps bytes to character classes that
- * to reduce the size of the transition table and create bitmasks. */
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, /*-1F*/
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, /*-3F*/
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, /*-5F*/
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, /*-7F*/
- 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, /*-9F*/
- 7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7, 7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7, /*-BF*/
- 8,8,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /*-DF*/
- 10,3,3,3,3,3,3,3,3,3,3,3,3,4,3,3, 11,6,6,6,5,8,8,8,8,8,8,8,8,8,8,8, /*-FF*/
-
- /* The second part is a transition table that maps a combination
- * of a state of the automaton and a character class to a state. */
- 0,12,24,36,60,96,84,12,12,12,48,72, 12,12,12,12,12,12,12,12,12,12,12,12,
- 12, 0,12,12,12,12,12, 0,12, 0,12,12, 12,24,12,12,12,12,12,24,12,24,12,12,
- 12,12,12,12,12,12,12,24,12,12,12,12, 12,24,12,12,12,12,12,12,12,24,12,12,
- 12,12,12,12,12,12,12,36,12,36,12,12, 12,36,12,12,12,12,12,36,12,36,12,12,
- 12,36,12,12,12,12,12,12,12,12,12,12
-};
-
-#endif
-
-#ifndef EBCDIC
-
-/* This is a version of the above table customized for Perl that doesn't
- * exclude surrogates and accepts start bytes up through F7 (representing
- * 2**21 - 1). */
-static U8 dfa_tab_for_perl[] = {
- /* The first part of the table maps bytes to character classes to reduce
- * the size of the transition table and create bitmasks. */
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, /*-1F*/
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, /*-3F*/
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, /*-5F*/
- 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, /*-7F*/
- 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9, /*-9F*/
- 7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7, 7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7, /*-BF*/
- 8,8,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, /*-DF*/
- 10,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, 11,4,4,4,4,4,4,4,8,8,8,8,8,8,8,8, /*-FF*/
-
- /* The second part is a transition table that maps a combination
- * of a state of the automaton and a character class to a state. */
- 0,12,24,36,96,12,12,12,12,12,48,72, 12,12,12,12,12,12,12,12,12,12,12,12,/*23*/
- 12, 0,12,12,12,12,12, 0,12, 0,12,12, 12,24,12,12,12,12,12,24,12,24,12,12,/*47*/
- 12,12,12,12,12,12,12,24,12,12,12,12, 12,24,12,12,12,12,12,12,12,24,12,12,/*71*/
- 12,12,12,12,12,12,12,36,12,36,12,12, 12,36,12,12,12,12,12,36,12,36,12,12,/*95*/
- 12,36,12,12,12,12,12,36,12,36,12,12 /* 96- 107 */
-
- /* The customization was to repurpose the surrogates type '4' to instead be
- * for start bytes F1-F7. Types 5 and 6 are now unused, and their entries in
- * the transition part of the table are set to 12, so are illegal.
- *
- * To do higher code points would require expansion and some rearrangement of
- * the table. The type '1' entries for continuation bytes 80-8f would have to
- * be split into several types, because they aren't treated uniformly for
- * higher start bytes, since overlongs for F8 are 80-87; FC: 80-83; and FE:
- * 80-81. We start needing to worry about overflow if FE is included.
- * Ignoring, FE and FF, we could use type 5 for F9-FB, and 6 for FD (remember
- * from the web site that these are used to right shift). FE would
- * necessarily be type 7; and FF, type 8. And new states would have to be
- * created for F8 and FC (and FE and FF if used), so quite a bit of work would
- * be involved.
- *
- * XXX Better would be to customize the table so that the noncharacters are
- * excluded. This again is non trivial, but doing so would simplify the code
- * that uses this, and might make it small enough to make it inlinable */
-};
-
-#endif
-
/*
=for apidoc utf8n_to_uvchr_error
*/
UV
-Perl_utf8n_to_uvchr_error(pTHX_ const U8 *s,
+Perl_utf8n_to_uvchr_error(const U8 *s,
STRLEN curlen,
STRLEN *retlen,
const U32 flags,
*/
UV
-Perl_utf8n_to_uvchr_msgs(pTHX_ const U8 *s,
+Perl__utf8n_to_uvchr_msgs_helper(const U8 *s,
STRLEN curlen,
STRLEN *retlen,
const U32 flags,
{
const U8 * const s0 = s;
const U8 * send = s0 + curlen;
- U32 possible_problems = 0; /* A bit is set here for each potential problem
- found as we go along */
- UV uv = (UV) -1;
- STRLEN expectlen = 0; /* How long should this sequence be?
- (initialized to silence compilers' wrong
- warning) */
- STRLEN avail_len = 0; /* When input is too short, gives what that is */
- U32 discard_errors = 0; /* Used to save branches when 'errors' is NULL;
- this gets set and discarded */
+ U32 possible_problems; /* A bit is set here for each potential problem
+ found as we go along */
+ UV uv;
+ STRLEN expectlen; /* How long should this sequence be? */
+ STRLEN avail_len; /* When input is too short, gives what that is */
+ U32 discard_errors; /* Used to save branches when 'errors' is NULL; this
+ gets set and discarded */
/* The below are used only if there is both an overlong malformation and a
* too short one. Otherwise the first two are set to 's0' and 'send', and
* the third not used at all */
- U8 * adjusted_s0 = (U8 *) s0;
+ U8 * adjusted_s0;
U8 temp_char_buf[UTF8_MAXBYTES + 1]; /* Used to avoid a Newx in this
routine; see [perl #130921] */
- UV uv_so_far = 0; /* (Initialized to silence compilers' wrong warning) */
+ UV uv_so_far;
+ dTHX;
+
+ PERL_ARGS_ASSERT__UTF8N_TO_UVCHR_MSGS_HELPER;
+
+ /* Here, is one of: a) malformed; b) a problematic code point (surrogate,
+ * non-unicode, or nonchar); or c) on ASCII platforms, one of the Hangul
+ * syllables that the dfa doesn't properly handle. Quickly dispose of the
+ * final case. */
+
+#ifndef EBCDIC
+
+ /* Each of the affected Hanguls starts with \xED */
+
+ if (is_HANGUL_ED_utf8_safe(s0, send)) {
+ if (retlen) {
+ *retlen = 3;
+ }
+ if (errors) {
+ *errors = 0;
+ }
+ if (msgs) {
+ *msgs = NULL;
+ }
- UV state = 0;
+ return ((0xED & UTF_START_MASK(3)) << (2 * UTF_ACCUMULATION_SHIFT))
+ | ((s0[1] & UTF_CONTINUATION_MASK) << UTF_ACCUMULATION_SHIFT)
+ | (s0[2] & UTF_CONTINUATION_MASK);
+ }
+
+#endif
- PERL_ARGS_ASSERT_UTF8N_TO_UVCHR_MSGS;
+ /* In conjunction with the exhaustive tests that can be enabled in
+ * APItest/t/utf8_warn_base.pl, this can make sure the dfa does precisely
+ * what it is intended to do, and that no flaws in it are masked by
+ * dropping down and executing the code below
+ assert(! isUTF8_CHAR(s0, send)
+ || UTF8_IS_SURROGATE(s0, send)
+ || UTF8_IS_SUPER(s0, send)
+ || UTF8_IS_NONCHAR(s0,send));
+ */
+
+ s = s0;
+ uv = *s0;
+ possible_problems = 0;
+ expectlen = 0;
+ avail_len = 0;
+ discard_errors = 0;
+ adjusted_s0 = (U8 *) s0;
+ uv_so_far = 0;
if (errors) {
*errors = 0;
*retlen = expectlen;
}
- /* An invariant is trivially well-formed */
- if (UTF8_IS_INVARIANT(*s0)) {
- return *s0;
- }
-
-#ifndef EBCDIC
-
- /* Measurements show that this dfa is somewhat faster than the regular code
- * below, so use it first, dropping down for the non-normal cases. */
-
-# define PERL_UTF8_DECODE_REJECT 12
-
- while (s < send && LIKELY(state != PERL_UTF8_DECODE_REJECT)) {
- UV type = dfa_tab_for_perl[*s];
-
- if (state != 0) {
- uv = (*s & 0x3fu) | (uv << UTF_ACCUMULATION_SHIFT);
- state = dfa_tab_for_perl[256 + state + type];
- }
- else {
- uv = (0xff >> type) & (*s);
- state = dfa_tab_for_perl[256 + type];
- }
-
- if (state == 0) {
-
- /* If this could be a code point that the flags don't allow (the first
- * surrogate is the first such possible one), delve further, but we already
- * have calculated 'uv' */
- if ( (flags & (UTF8_DISALLOW_ILLEGAL_INTERCHANGE
- |UTF8_WARN_ILLEGAL_INTERCHANGE))
- && uv >= UNICODE_SURROGATE_FIRST)
- {
- curlen = s + 1 - s0;
- goto got_uv;
- }
-
- return uv;
- }
-
- s++;
- }
-
- /* Here, is some sort of failure. Use the full mechanism */
-
- uv = *s0;
-
-#endif
-
/* A continuation character can't start a valid sequence */
if (UNLIKELY(UTF8_IS_CONTINUATION(uv))) {
possible_problems |= UTF8_GOT_CONTINUATION;
}
}
- got_uv:
-
/* Here, we have found all the possible problems, except for when the input
* is for a problematic code point not allowed by the input parameters. */
return _to_upper_title_latin1((U8) c, p, lenp, 'S');
}
- uvchr_to_utf8(p, c);
- return CALL_UPPER_CASE(c, p, p, lenp);
+ return CALL_UPPER_CASE(c, NULL, p, lenp);
}
UV
return _to_upper_title_latin1((U8) c, p, lenp, 's');
}
- uvchr_to_utf8(p, c);
- return CALL_TITLE_CASE(c, p, p, lenp);
+ return CALL_TITLE_CASE(c, NULL, p, lenp);
}
STATIC U8
return to_lower_latin1((U8) c, p, lenp, 0 /* 0 is a dummy arg */ );
}
- uvchr_to_utf8(p, c);
- return CALL_LOWER_CASE(c, p, p, lenp);
+ return CALL_LOWER_CASE(c, NULL, p, lenp);
}
UV
/* Here, above 255. If no special needs, just use the macro */
if ( ! (flags & (FOLD_FLAGS_LOCALE|FOLD_FLAGS_NOMIX_ASCII))) {
- uvchr_to_utf8(p, c);
- return CALL_FOLD_CASE(c, p, p, lenp, flags & FOLD_FLAGS_FULL);
+ return CALL_FOLD_CASE(c, NULL, p, lenp, flags & FOLD_FLAGS_FULL);
}
else { /* Otherwise, _toFOLD_utf8_flags has the intelligence to deal with
the special flags. */
if (*p == '_')
return TRUE;
- return is_utf8_common(p, &PL_utf8_xidstart, "XIdStart", NULL);
+
+ return is_utf8_common(p, NULL,
+ "This is buggy if this gets used",
+ PL_utf8_xidstart);
}
bool
{
PERL_ARGS_ASSERT__IS_UTF8_IDCONT;
- return is_utf8_common(p, &PL_utf8_idcont, "IdContinue", NULL);
+ return is_utf8_common(p, NULL,
+ "This is buggy if this gets used",
+ PL_utf8_idcont);
}
bool
{
PERL_ARGS_ASSERT__IS_UTF8_XIDCONT;
- return is_utf8_common(p, &PL_utf8_xidcont, "XIdContinue", NULL);
+ return is_utf8_common(p, NULL,
+ "This is buggy if this gets used",
+ PL_utf8_xidcont);
}
bool
{
PERL_ARGS_ASSERT__IS_UTF8_MARK;
- return is_utf8_common(p, &PL_utf8_mark, "IsM", NULL);
+ return is_utf8_common(p, NULL, "IsM", PL_utf8_mark);
}
STATIC UV
unsigned int i;
const unsigned int * cp_list;
U8 * d;
+
+ /* 'index' is guaranteed to be non-negative, as this is an inversion
+ * map that covers all possible inputs. See [perl #133365] */
SSize_t index = _invlist_search(invlist, uv1);
IV base = invmap[index];
/* Here, there was no mapping defined, which means that the code point maps
* to itself. Return the inputs */
cases_to_self:
- len = UTF8SKIP(p);
- if (p != ustrp) { /* Don't copy onto itself */
- Copy(p, ustrp, len, U8);
+ if (p) {
+ len = UTF8SKIP(p);
+ if (p != ustrp) { /* Don't copy onto itself */
+ Copy(p, ustrp, len, U8);
+ }
+ *lenp = len;
+ }
+ else {
+ *lenp = uvchr_to_utf8(ustrp, uv1) - ustrp;
}
-
- if (lenp)
- *lenp = len;
return uv1;
* the return can point to them, but single code points aren't, so would
* need to be constructed if we didn't employ something like this API */
+ /* 'index' is guaranteed to be non-negative, as this is an inversion map
+ * that covers all possible inputs. See [perl #133365] */
SSize_t index = _invlist_search(PL_utf8_foldclosures, cp);
int base = _Perl_IVCF_invmap[index];
/* Get the 0th element, which is needed to setup the inversion list
* */
while (isSPACE(*l)) l++;
+ after_atou = (char *) lend;
if (!grok_atoUV((const char *)l, &element0, &after_atou)) {
Perl_croak(aTHX_ "panic: Expecting a valid 0th element for"
" inversion list");
" elements than available", elements);
}
while (isSPACE(*l)) l++;
+ after_atou = (char *) lend;
if (!grok_atoUV((const char *)l, other_elements_ptr++,
&after_atou))
{
}
/*
+=for apidoc utf8_to_uvchr
+
+Returns the native code point of the first character in the string C<s>
+which is assumed to be in UTF-8 encoding; C<retlen> will be set to the
+length, in bytes, of that character.
+
+Some, but not all, UTF-8 malformations are detected, and in fact, some
+malformed input could cause reading beyond the end of the input buffer, which
+is why this function is deprecated. Use L</utf8_to_uvchr_buf> instead.
+
+If C<s> points to one of the detected malformations, and UTF8 warnings are
+enabled, zero is returned and C<*retlen> is set (if C<retlen> isn't
+C<NULL>) to -1. If those warnings are off, the computed value if well-defined (or
+the Unicode REPLACEMENT CHARACTER, if not) is silently returned, and C<*retlen>
+is set (if C<retlen> isn't NULL) so that (S<C<s> + C<*retlen>>) is the
+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.
+
+=cut
+*/
+
+UV
+Perl_utf8_to_uvchr(pTHX_ const U8 *s, STRLEN *retlen)
+{
+ PERL_ARGS_ASSERT_UTF8_TO_UVCHR;
+
+ /* This function is unsafe if malformed UTF-8 input is given it, which is
+ * why the function is deprecated. If the first byte of the input
+ * indicates that there are more bytes remaining in the sequence that forms
+ * the character than there are in the input buffer, it can read past the
+ * end. But we can make it safe if the input string happens to be
+ * NUL-terminated, as many strings in Perl are, by refusing to read past a
+ * NUL. A NUL indicates the start of the next character anyway. If the
+ * input isn't NUL-terminated, the function remains unsafe, as it always
+ * has been.
+ *
+ * An initial NUL has to be handled separately, but all ASCIIs can be
+ * handled the same way, speeding up this common case */
+
+ if (UTF8_IS_INVARIANT(*s)) { /* Assumes 's' contains at least 1 byte */
+ return (UV) *s;
+ }
+
+ return utf8_to_uvchr_buf(s,
+ s + my_strnlen((char *) s, UTF8_MAXBYTES),
+ retlen);
+}
+
+/*
* ex: set ts=8 sts=4 sw=4 et:
*/