Change 24730 by [EMAIL PROTECTED] on 2005/06/07 14:35:25
Subject: [PATCH] one more round of is_utf8_foo tuneup
From: Jarkko Hietaniemi <[EMAIL PROTECTED]>
Date: Sun, 05 Jun 2005 18:06:12 +0300
Message-ID: <[EMAIL PROTECTED]>
Affected files ...
... //depot/perl/embed.fnc#189 edit
... //depot/perl/embed.h#481 edit
... //depot/perl/global.sym#275 edit
... //depot/perl/mg.c#335 edit
... //depot/perl/pod/perlapi.pod#221 edit
... //depot/perl/proto.h#540 edit
... //depot/perl/utf8.c#236 edit
... //depot/perl/utf8.h#69 edit
Differences ...
==== //depot/perl/embed.fnc#189 (text) ====
Index: perl/embed.fnc
--- perl/embed.fnc#188~24706~ Sat Jun 4 09:55:27 2005
+++ perl/embed.fnc Tue Jun 7 07:35:25 2005
@@ -346,7 +346,8 @@
ApPR |bool |is_uni_xdigit_lc|UV c
Apd |STRLEN |is_utf8_char |NN const U8 *p
Apd |bool |is_utf8_string |NN const U8 *s|STRLEN len
-Apd |bool |is_utf8_string_loc|NN const U8 *s|STRLEN len|NN const U8 **p
+Apd |bool |is_utf8_string_loc|NN const U8 *s|STRLEN len|const U8 **p
+Apd |bool |is_utf8_string_loclen|NN const U8 *s|STRLEN len|const U8
**ep|STRLEN *el
ApR |bool |is_utf8_alnum |NN const U8 *p
ApR |bool |is_utf8_alnumc |NN const U8 *p
ApR |bool |is_utf8_idfirst|NN const U8 *p
==== //depot/perl/embed.h#481 (text+w) ====
Index: perl/embed.h
--- perl/embed.h#480~24706~ Sat Jun 4 09:55:27 2005
+++ perl/embed.h Tue Jun 7 07:35:25 2005
@@ -345,6 +345,7 @@
#define is_utf8_char Perl_is_utf8_char
#define is_utf8_string Perl_is_utf8_string
#define is_utf8_string_loc Perl_is_utf8_string_loc
+#define is_utf8_string_loclen Perl_is_utf8_string_loclen
#define is_utf8_alnum Perl_is_utf8_alnum
#define is_utf8_alnumc Perl_is_utf8_alnumc
#define is_utf8_idfirst Perl_is_utf8_idfirst
@@ -2344,6 +2345,7 @@
#define is_utf8_char(a) Perl_is_utf8_char(aTHX_ a)
#define is_utf8_string(a,b) Perl_is_utf8_string(aTHX_ a,b)
#define is_utf8_string_loc(a,b,c) Perl_is_utf8_string_loc(aTHX_ a,b,c)
+#define is_utf8_string_loclen(a,b,c,d) Perl_is_utf8_string_loclen(aTHX_
a,b,c,d)
#define is_utf8_alnum(a) Perl_is_utf8_alnum(aTHX_ a)
#define is_utf8_alnumc(a) Perl_is_utf8_alnumc(aTHX_ a)
#define is_utf8_idfirst(a) Perl_is_utf8_idfirst(aTHX_ a)
==== //depot/perl/global.sym#275 (text+w) ====
Index: perl/global.sym
--- perl/global.sym#274~24706~ Sat Jun 4 09:55:27 2005
+++ perl/global.sym Tue Jun 7 07:35:25 2005
@@ -207,6 +207,7 @@
Perl_is_utf8_char
Perl_is_utf8_string
Perl_is_utf8_string_loc
+Perl_is_utf8_string_loclen
Perl_is_utf8_alnum
Perl_is_utf8_alnumc
Perl_is_utf8_idfirst
==== //depot/perl/mg.c#335 (text) ====
Index: perl/mg.c
--- perl/mg.c#334~24580~ Thu May 26 01:43:38 2005
+++ perl/mg.c Tue Jun 7 07:35:25 2005
@@ -495,10 +495,12 @@
if (i > 0 && RX_MATCH_UTF8(rx)) {
char *s = rx->subbeg + s1;
char *send = rx->subbeg + t1;
+ const U8 *ep;
+ STRLEN el;
i = t1 - s1;
- if (is_utf8_string((U8*)s, i))
- i = Perl_utf8_length(aTHX_ (U8*)s, (U8*)send);
+ if (is_utf8_string_loclen((U8*)s, i, &ep, &el))
+ i = el;
}
if (i < 0)
Perl_croak(aTHX_ "panic: magic_len: %"IVdf, (IV)i);
==== //depot/perl/pod/perlapi.pod#221 (text+w) ====
Index: perl/pod/perlapi.pod
--- perl/pod/perlapi.pod#220~24728~ Tue Jun 7 06:15:45 2005
+++ perl/pod/perlapi.pod Tue Jun 7 07:35:25 2005
@@ -5325,6 +5325,8 @@
not mean 'a string that contains code points above 0x7F encoded in UTF-8'
because a valid ASCII string is a valid UTF-8 string.
+See also is_utf8_string_loclen() and is_utf8_string_loc().
+
bool is_utf8_string(const U8 *s, STRLEN len)
=for hackers
@@ -5332,10 +5334,27 @@
=item is_utf8_string_loc
-Like is_ut8_string but store the location of the failure in
-the last argument.
+Like is_ut8_string() but stores the location of the failure (in the
+case of "utf8ness failure") or the location s+len (in the case of
+"utf8ness success") in the C<ep>.
+
+See also is_utf8_string_loclen() and is_utf8_string().
bool is_utf8_string_loc(const U8 *s, STRLEN len, const U8 **p)
+
+=for hackers
+Found in file utf8.c
+
+=item is_utf8_string_loclen
+
+Like is_ut8_string() but stores the location of the failure (in the
+case of "utf8ness failure") or the location s+len (in the case of
+"utf8ness success") in the C<ep>, and the number of UTF-8
+encoded characters in the C<el>.
+
+See also is_utf8_string_loc() and is_utf8_string().
+
+ bool is_utf8_string_loclen(const U8 *s, STRLEN len, const U8 **ep,
STRLEN *el)
=for hackers
Found in file utf8.c
==== //depot/perl/proto.h#540 (text+w) ====
Index: perl/proto.h
--- perl/proto.h#539~24706~ Sat Jun 4 09:55:27 2005
+++ perl/proto.h Tue Jun 7 07:35:25 2005
@@ -620,8 +620,10 @@
__attribute__nonnull__(pTHX_1);
PERL_CALLCONV bool Perl_is_utf8_string_loc(pTHX_ const U8 *s, STRLEN len,
const U8 **p)
- __attribute__nonnull__(pTHX_1)
- __attribute__nonnull__(pTHX_3);
+ __attribute__nonnull__(pTHX_1);
+
+PERL_CALLCONV bool Perl_is_utf8_string_loclen(pTHX_ const U8 *s, STRLEN
len, const U8 **ep, STRLEN *el)
+ __attribute__nonnull__(pTHX_1);
PERL_CALLCONV bool Perl_is_utf8_alnum(pTHX_ const U8 *p)
__attribute__warn_unused_result__
==== //depot/perl/utf8.c#236 (text) ====
Index: perl/utf8.c
--- perl/utf8.c#235~24707~ Sat Jun 4 20:55:11 2005
+++ perl/utf8.c Tue Jun 7 07:35:25 2005
@@ -240,10 +240,9 @@
STRLEN
Perl_is_utf8_char(pTHX_ const U8 *s)
{
- STRLEN len;
+ STRLEN len = UTF8SKIP(s);
#ifdef IS_UTF8_CHAR
- len = UTF8SKIP(s);
- if (len <= 4)
+ if (IS_UTF8_CHAR_FAST(len))
return IS_UTF8_CHAR(s, len) ? len : 0;
#endif /* #ifdef IS_UTF8_CHAR */
return is_utf8_char_slow(s, len);
@@ -257,6 +256,8 @@
not mean 'a string that contains code points above 0x7F encoded in UTF-8'
because a valid ASCII string is a valid UTF-8 string.
+See also is_utf8_string_loclen() and is_utf8_string_loc().
+
=cut
*/
@@ -276,26 +277,26 @@
if (UTF8_IS_INVARIANT(*x))
c = 1;
else if (!UTF8_IS_START(*x))
- return FALSE;
+ goto out;
else {
/* ... and call is_utf8_char() only if really needed. */
#ifdef IS_UTF8_CHAR
c = UTF8SKIP(x);
- if (c <= 4) {
- if (!IS_UTF8_CHAR(x, c))
- return FALSE;
- } else {
- if (!is_utf8_char_slow(x, c))
- return FALSE;
- }
+ if (IS_UTF8_CHAR_FAST(c)) {
+ if (!IS_UTF8_CHAR(x, c))
+ goto out;
+ } else if (!is_utf8_char_slow(x, c))
+ goto out;
#else
c = is_utf8_char(x);
#endif /* #ifdef IS_UTF8_CHAR */
if (!c)
- return FALSE;
+ goto out;
}
x += c;
}
+
+ out:
if (x != send)
return FALSE;
@@ -303,16 +304,20 @@
}
/*
-=for apidoc A|bool|is_utf8_string_loc|const U8 *s|STRLEN len|const U8 **p
+=for apidoc A|bool|is_utf8_string_loclen|const U8 *s|STRLEN len|const U8
**ep|const STRLEN *el
-Like is_ut8_string but store the location of the failure in
-the last argument.
+Like is_ut8_string() but stores the location of the failure (in the
+case of "utf8ness failure") or the location s+len (in the case of
+"utf8ness success") in the C<ep>, and the number of UTF-8
+encoded characters in the C<el>.
+
+See also is_utf8_string_loc() and is_utf8_string().
=cut
*/
bool
-Perl_is_utf8_string_loc(pTHX_ const U8 *s, STRLEN len, const U8 **p)
+Perl_is_utf8_string_loclen(pTHX_ const U8 *s, STRLEN len, const U8 **ep,
STRLEN *el)
{
const U8* x = s;
const U8* send;
@@ -321,34 +326,60 @@
if (!len && s)
len = strlen((const char *)s);
send = s + len;
+ if (el)
+ *el = 0;
while (x < send) {
/* Inline the easy bits of is_utf8_char() here for speed... */
if (UTF8_IS_INVARIANT(*x))
- c = 1;
- else if (!UTF8_IS_START(*x)) {
- if (p)
- *p = x;
- return FALSE;
- }
+ c = 1;
+ else if (!UTF8_IS_START(*x))
+ goto out;
else {
- /* ... and call is_utf8_char() only if really needed. */
- c = is_utf8_char(x);
- if (!c) {
- if (p)
- *p = x;
- return FALSE;
- }
+ /* ... and call is_utf8_char() only if really needed. */
+#ifdef IS_UTF8_CHAR
+ c = UTF8SKIP(x);
+ if (IS_UTF8_CHAR_FAST(c)) {
+ if (!IS_UTF8_CHAR(x, c))
+ c = 0;
+ } else
+ c = is_utf8_char_slow(x, c);
+#else
+ c = is_utf8_char(x);
+#endif /* #ifdef IS_UTF8_CHAR */
+ if (!c)
+ goto out;
}
- x += c;
+ x += c;
+ if (el)
+ (*el)++;
}
- if (x != send) {
- if (p)
- *p = x;
+
+ out:
+ if (ep)
+ *ep = x;
+ if (x != send)
return FALSE;
- }
return TRUE;
+}
+
+/*
+=for apidoc A|bool|is_utf8_string_loc|const U8 *s|STRLEN len|const U8
**ep|const STRLEN *el
+
+Like is_ut8_string() but stores the location of the failure (in the
+case of "utf8ness failure") or the location s+len (in the case of
+"utf8ness success") in the C<ep>.
+
+See also is_utf8_string_loclen() and is_utf8_string().
+
+=cut
+*/
+
+bool
+Perl_is_utf8_string_loc(pTHX_ const U8 *s, STRLEN len, const U8 **ep)
+{
+ return is_utf8_string_loclen(s, len, ep, 0);
}
/*
==== //depot/perl/utf8.h#69 (text) ====
Index: perl/utf8.h
--- perl/utf8.h#68~24687~ Fri Jun 3 01:08:25 2005
+++ perl/utf8.h Tue Jun 7 07:35:25 2005
@@ -327,3 +327,5 @@
(n) == 3 ? IS_UTF8_CHAR_3(p) : \
(n) == 4 ? IS_UTF8_CHAR_4(p) : 0)
+#define IS_UTF8_CHAR_FAST(n) ((n) <= 4)
+
End of Patch.
