Improve tolower()-related functions on Windows and VC2005 by caching locale and using

tolower_l function.
# Gives 10-18% speedup on bench.php
# Does not change behaviour for non-windows

5 files changed, 38 insertions, 6 deletions

diff --git a/Zend/zend_operators.c b/Zend/zend_operators.c
index e83fa22e46..d8853ebb7a 100644
--- a/Zend/zend_operators.c
+++ b/Zend/zend_operators.c
@@ -32,6 +32,15 @@
 
 #define LONG_SIGN_MASK (1L << (8*sizeof(long)-1))
 
+#if ZEND_USE_TOLOWER_L
+#include <locale.h>
+static _locale_t current_locale = NULL;
+/* this is true global! may lead to strange effects on ZTS, but so is setlocale() */
+#define zend_tolower(c) _tolower_l(c, current_locale)
+#else
+#define zend_tolower(c) tolower(c)
+#endif
+
 ZEND_API int zend_atoi(const char *str, int str_len)
 {
 	int retval;
@@ -1818,6 +1827,13 @@ ZEND_API int zval_is_true(zval *op)
 	return (op->value.lval ? 1 : 0);
 }
 
+#ifdef ZEND_USE_TOLOWER_L
+ZEND_API void zend_update_current_locale()
+{
+	current_locale = _get_current_locale();
+}
+#endif
+
 ZEND_API char *zend_str_tolower_copy(char *dest, const char *source, unsigned int length)
 {
 	register unsigned char *str = (unsigned char*)source;
@@ -1825,7 +1841,7 @@ ZEND_API char *zend_str_tolower_copy(char *dest, const char *source, unsigned in
 	register unsigned char *end = str + length;
 
 	while (str < end) {
-		*result++ = tolower((int)*str++);
+		*result++ = zend_tolower((int)*str++);
 	}
 	*result = '\0';
 
@@ -1838,7 +1854,7 @@ ZEND_API void zend_str_tolower(char *str, unsigned int length)
 	register unsigned char *end = p + length;
 
 	while (p < end) {
-		*p = tolower((int)*p);
+		*p = zend_tolower((int)*p);
 		p++;
 	}
 }
@@ -1876,8 +1892,8 @@ ZEND_API int zend_binary_strcasecmp(char *s1, uint len1, char *s2, uint len2)
 	len = MIN(len1, len2);
 
 	while (len--) {
-		c1 = tolower((int)*(unsigned char *)s1++);
-		c2 = tolower((int)*(unsigned char *)s2++);
+		c1 = zend_tolower((int)*(unsigned char *)s1++);
+		c2 = zend_tolower((int)*(unsigned char *)s2++);
 		if (c1 != c2) {
 			return c1 - c2;
 		}
@@ -1895,8 +1911,8 @@ ZEND_API int zend_binary_strncasecmp(char *s1, uint len1, char *s2, uint len2, u
 	len = MIN(length, MIN(len1, len2));
 
 	while (len--) {
-		c1 = tolower((int)*(unsigned char *)s1++);
-		c2 = tolower((int)*(unsigned char *)s2++);
+		c1 = zend_tolower((int)*(unsigned char *)s1++);
+		c2 = zend_tolower((int)*(unsigned char *)s2++);
 		if (c1 != c2) {
 			return c1 - c2;
 		}
diff --git a/Zend/zend_operators.h b/Zend/zend_operators.h
index c7c3991d57..5763179971 100644
--- a/Zend/zend_operators.h
+++ b/Zend/zend_operators.h
@@ -346,6 +346,19 @@ END_EXTERN_C()
 #define Z_TYPE_P(zval_p)	Z_TYPE(*zval_p)
 #define Z_TYPE_PP(zval_pp)	Z_TYPE(**zval_pp)
 
+#if HAVE_SETLOCALE && defined(ZEND_WIN32) && !defined(ZTS) && defined(_MSC_VER) && (_MSC_VER >= 1400)
+/* This is performance improvement of tolower() on Windows and VC2005
+ * GIves 10-18% on bench.php
+ */
+#define ZEND_USE_TOLOWER_L 1
+#endif
+
+#ifdef ZEND_USE_TOLOWER_L
+ZEND_API void zend_update_current_locale();
+#else
+#define zend_update_current_locale()
+#endif
+
 #endif
 
 /*
diff --git a/ext/standard/basic_functions.c b/ext/standard/basic_functions.c
index 49754349a8..69b816d226 100644
--- a/ext/standard/basic_functions.c
+++ b/ext/standard/basic_functions.c
@@ -4157,6 +4157,7 @@ PHP_RSHUTDOWN_FUNCTION(basic)
 	if (BG(locale_string) != NULL) {
 		setlocale(LC_ALL, "C");
 		setlocale(LC_CTYPE, "");
+		zend_update_current_locale();
 	}
 	STR_FREE(BG(locale_string));
 	BG(locale_string) = NULL;
diff --git a/ext/standard/string.c b/ext/standard/string.c
index 34c874e1c4..d7f4b62324 100644
--- a/ext/standard/string.c
+++ b/ext/standard/string.c
@@ -3971,6 +3971,7 @@ PHP_FUNCTION(setlocale)
 		}
 		
 		retval = setlocale (cat, loc);
+		zend_update_current_locale();
 		if (retval) {
 			/* Remember if locale was changed */
 			if (loc) {
diff --git a/main/main.c b/main/main.c
index 667bca3663..8279c04037 100644
--- a/main/main.c
+++ b/main/main.c
@@ -1502,6 +1502,7 @@ int php_module_startup(sapi_module_struct *sf, zend_module_entry *additional_mod
 
 #if HAVE_SETLOCALE
 	setlocale(LC_CTYPE, "");
+	zend_update_current_locale();
 #endif
 
 #if HAVE_TZSET