Merged revisions 72040 via svnmerge from

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r72040 | eric.smith | 2009-04-27 15:04:37 -0400 (Mon, 27 Apr 2009) | 1 line

  Issue #5793: rationalize isdigit / isalpha / tolower, etc. Will port to py3k. Should fix Windows buildbot errors.
........

Include/Python.h

@@ -116,6 +116,7 @@
 #include "compile.h"
 #include "eval.h"
 
+#include "pyctype.h"
 #include "pystrtod.h"
 #include "pystrcmp.h"
 #include "dtoa.h"

Include/bytes_methods.h

@@ -38,23 +38,15 @@ extern const char _Py_capitalize__doc__[];
 extern const char _Py_swapcase__doc__[];
 extern const char _Py_maketrans__doc__[];
 
-#define FLAG_LOWER  0x01
-#define FLAG_UPPER  0x02
-#define FLAG_ALPHA  (FLAG_LOWER|FLAG_UPPER)
-#define FLAG_DIGIT  0x04
-#define FLAG_ALNUM  (FLAG_ALPHA|FLAG_DIGIT)
-#define FLAG_SPACE  0x08
-#define FLAG_XDIGIT 0x10
-
-extern const unsigned int _Py_ctype_table[256];
-
-#define ISLOWER(c) (_Py_ctype_table[Py_CHARMASK(c)] & FLAG_LOWER)
-#define ISUPPER(c) (_Py_ctype_table[Py_CHARMASK(c)] & FLAG_UPPER)
-#define ISALPHA(c) (_Py_ctype_table[Py_CHARMASK(c)] & FLAG_ALPHA)
-#define ISDIGIT(c) (_Py_ctype_table[Py_CHARMASK(c)] & FLAG_DIGIT)
-#define ISXDIGIT(c) (_Py_ctype_table[Py_CHARMASK(c)] & FLAG_XDIGIT)
-#define ISALNUM(c) (_Py_ctype_table[Py_CHARMASK(c)] & FLAG_ALNUM)
-#define ISSPACE(c) (_Py_ctype_table[Py_CHARMASK(c)] & FLAG_SPACE)
+/* These are left in for backward compatibility and will be removed
+   in 2.8/3.2 */
+#define ISLOWER(c)  Py_ISLOWER(c)
+#define ISUPPER(c)  Py_ISUPPER(c)
+#define ISALPHA(c)  Py_ISALPHA(c)
+#define ISDIGIT(c)  Py_ISDIGIT(c)
+#define ISXDIGIT(c) Py_ISXDIGIT(c)
+#define ISALNUM(c)  Py_ISALNUM(c)
+#define ISSPACE(c)  Py_ISSPACE(c)
 
 #undef islower
 #define islower(c) undefined_islower(c)
@@ -71,11 +63,10 @@ extern const unsigned int _Py_ctype_table[256];
 #undef isspace
 #define isspace(c) undefined_isspace(c)
 
-extern const unsigned char _Py_ctype_tolower[256];
-extern const unsigned char _Py_ctype_toupper[256];
-
-#define TOLOWER(c) (_Py_ctype_tolower[Py_CHARMASK(c)])
-#define TOUPPER(c) (_Py_ctype_toupper[Py_CHARMASK(c)])
+/* These are left in for backward compatibility and will be removed
+   in 2.8/3.2 */
+#define TOLOWER(c) Py_TOLOWER(c)
+#define TOUPPER(c) Py_TOUPPER(c)
 
 #undef tolower
 #define tolower(c) undefined_tolower(c)

Makefile.pre.in

@@ -294,6 +294,7 @@ PYTHON_OBJS=	\
 		Python/mysnprintf.o \
 		Python/peephole.o \
 		Python/pyarena.o \
+		Python/pyctype.o \
 		Python/pyfpe.o \
 		Python/pymath.o \
 		Python/pystate.o \
@@ -653,6 +654,7 @@ PYTHON_HEADERS= \
 		Include/pgen.h \
 		Include/pgenheaders.h \
 		Include/pyarena.h \
+		Include/pyctype.h \
 		Include/pydebug.h \
 		Include/pyerrors.h \
 		Include/pyfpe.h \

Misc/NEWS

@@ -12,6 +12,9 @@ What's New in Python 3.1 beta 1?
 Core and Builtins
 -----------------
 
+- Issue #5793: Rationalize isdigit / isalpha / tolower, etc. Includes
+  new Py_ISDIGIT / Py_ISALPHA / Py_TOLOWER, etc. in pctypes.h.
+
 - Issue #5835: Deprecate PyOS_ascii_formatd.
 
 - Issue #4971: Fix titlecase for characters that are their own

Objects/bytearrayobject.c

@@ -2176,16 +2176,16 @@ split_whitespace(const char *s, Py_ssize_t len, Py_ssize_t maxcount)
 
     for (i = j = 0; i < len; ) {
         /* find a token */
-        while (i < len && ISSPACE(s[i]))
+        while (i < len && Py_ISSPACE(s[i]))
             i++;
         j = i;
-        while (i < len && !ISSPACE(s[i]))
+        while (i < len && !Py_ISSPACE(s[i]))
             i++;
         if (j < i) {
             if (maxcount-- <= 0)
                 break;
             SPLIT_ADD(s, j, i);
-            while (i < len && ISSPACE(s[i]))
+            while (i < len && Py_ISSPACE(s[i]))
                 i++;
             j = i;
         }
@@ -2410,16 +2410,16 @@ rsplit_whitespace(const char *s, Py_ssize_t len, Py_ssize_t maxcount)
 
     for (i = j = len - 1; i >= 0; ) {
         /* find a token */
-        while (i >= 0 && ISSPACE(s[i]))
+        while (i >= 0 && Py_ISSPACE(s[i]))
             i--;
         j = i;
-        while (i >= 0 && !ISSPACE(s[i]))
+        while (i >= 0 && !Py_ISSPACE(s[i]))
             i--;
         if (j > i) {
             if (maxcount-- <= 0)
                 break;
             SPLIT_ADD(s, i + 1, j + 1);
-            while (i >= 0 && ISSPACE(s[i]))
+            while (i >= 0 && Py_ISSPACE(s[i]))
                 i--;
             j = i;
         }
@@ -2986,11 +2986,11 @@ hex_digit_to_int(Py_UNICODE c)
 {
     if (c >= 128)
         return -1;
-    if (ISDIGIT(c))
+    if (Py_ISDIGIT(c))
         return c - '0';
     else {
-        if (ISUPPER(c))
-            c = TOLOWER(c);
+        if (Py_ISUPPER(c))
+            c = Py_TOLOWER(c);
         if (c >= 'a' && c <= 'f')
             return c - 'a' + 10;
     }

Objects/stringlib/stringdefs.h

@@ -13,8 +13,8 @@
 #define STRINGLIB_EMPTY          nullstring
 #define STRINGLIB_ISDECIMAL(x)   ((x >= '0') && (x <= '9'))
 #define STRINGLIB_TODECIMAL(x)   (STRINGLIB_ISDECIMAL(x) ? (x - '0') : -1)
-#define STRINGLIB_TOUPPER        toupper
-#define STRINGLIB_TOLOWER        tolower
+#define STRINGLIB_TOUPPER        Py_TOUPPER
+#define STRINGLIB_TOLOWER        Py_TOLOWER
 #define STRINGLIB_FILL           memset
 #define STRINGLIB_STR            PyBytes_AS_STRING
 #define STRINGLIB_LEN            PyBytes_GET_SIZE

PC/VC6/pythoncore.dsp

@@ -603,6 +603,10 @@ SOURCE=..\..\Python\pyarena.c
 # End Source File
 # Begin Source File
 
+SOURCE=..\..\Python\pyctype.c
+# End Source File
+# Begin Source File
+
 SOURCE=..\..\Python\pyfpe.c
 # End Source File
 # Begin Source File

PC/VS7.1/pythoncore.vcproj

@@ -720,6 +720,9 @@
 		<File
 			RelativePath="..\..\Python\pyarena.c">
 		</File>
+		<File
+			RelativePath="..\..\Python\pyctype.c">
+		</File>
 		<File
 			RelativePath="..\..\Python\pyfpe.c">
 		</File>

PC/VS8.0/pythoncore.vcproj

@@ -850,6 +850,10 @@
 				RelativePath="..\..\Include\pyarena.h"
 				>
 			</File>
+			<File
+				RelativePath="..\..\Include\pyctype.h"
+				>
+			</File>
 			<File
 				RelativePath="..\..\Include\pydebug.h"
 				>
@@ -1730,6 +1734,10 @@
 				RelativePath="..\..\Python\pyarena.c"
 				>
 			</File>
+			<File
+				RelativePath="..\..\Python\pyctype.c"
+				>
+			</File>
 			<File
 				RelativePath="..\..\Python\pyfpe.c"
 				>

PC/os2emx/Makefile

@@ -349,6 +349,7 @@ SRC.PYTHON=	$(addprefix $(TOP), \
 		Python/mysnprintf.c \
 		Python/mystrtoul.c \
 		Python/pyarena.c \
+		Python/pyctype.c \
 		Python/pyfpe.c \
 		Python/pystate.c \
 		Python/pystrtod.c \

PCbuild/pythoncore.vcproj

@@ -846,6 +846,10 @@
 				RelativePath="..\Include\pyarena.h"
 				>
 			</File>
+			<File
+				RelativePath="..\Include\pyctype.h"
+				>
+			</File>
 			<File
 				RelativePath="..\Include\pydebug.h"
 				>
@@ -1730,6 +1734,10 @@
 				RelativePath="..\Python\pyarena.c"
 				>
 			</File>
+			<File
+				RelativePath="..\Python\pyctype.c"
+				>
+			</File>
 			<File
 				RelativePath="..\Python\pyfpe.c"
 				>

Python/pystrtod.c

@@ -3,12 +3,6 @@
 #include <Python.h>
 #include <locale.h>
 
-/* ascii character tests (as opposed to locale tests) */
-#define ISSPACE(c)  ((c) == ' ' || (c) == '\f' || (c) == '\n' || \
-                     (c) == '\r' || (c) == '\t' || (c) == '\v')
-#define ISDIGIT(c)  ((c) >= '0' && (c) <= '9')
-
-
 /**
  * PyOS_ascii_strtod:
  * @nptr:    the string to convert to a numeric value.
@@ -104,7 +98,7 @@ PyOS_ascii_strtod(const char *nptr, char **endptr)
 
 	p = nptr;
 	/* Skip leading space */
-	while (ISSPACE(*p))
+	while (Py_ISSPACE(*p))
 		p++;
 
 	/* Process leading sign, if present */
@@ -147,7 +141,7 @@ PyOS_ascii_strtod(const char *nptr, char **endptr)
 		goto invalid_string;
 
 	/* Check that what's left begins with a digit or decimal point */
-	if (!ISDIGIT(*p) && *p != '.')
+	if (!Py_ISDIGIT(*p) && *p != '.')
 		goto invalid_string;
 
 	digits_pos = p;
@@ -158,7 +152,7 @@ PyOS_ascii_strtod(const char *nptr, char **endptr)
 		   swapped for the current locale's decimal point before we
 		   call strtod.  On the other hand, if we find the current
 		   locale's decimal point then the input is invalid. */
-		while (ISDIGIT(*p))
+		while (Py_ISDIGIT(*p))
 			p++;
 
 		if (*p == '.')
@@ -166,14 +160,14 @@ PyOS_ascii_strtod(const char *nptr, char **endptr)
 			decimal_point_pos = p++;
 
 			/* locate end of number */
-			while (ISDIGIT(*p))
+			while (Py_ISDIGIT(*p))
 				p++;
 
 			if (*p == 'e' || *p == 'E')
 				p++;
 			if (*p == '+' || *p == '-')
 				p++;
-			while (ISDIGIT(*p))
+			while (Py_ISDIGIT(*p))
 				p++;
 			end = p;
 		}
@@ -269,7 +263,7 @@ change_decimal_from_locale_to_dot(char* buffer)
 
 		if (*buffer == '+' || *buffer == '-')
 			buffer++;
-		while (isdigit(Py_CHARMASK(*buffer)))
+		while (Py_ISDIGIT(*buffer))
 			buffer++;
 		if (strncmp(buffer, decimal_point, decimal_point_len) == 0) {
 			*buffer = '.';
@@ -312,7 +306,7 @@ ensure_minimum_exponent_length(char* buffer, size_t buf_size)
 
 		/* Find the end of the exponent, keeping track of leading
 		   zeros. */
-		while (*p && isdigit(Py_CHARMASK(*p))) {
+		while (*p && Py_ISDIGIT(*p)) {
 			if (in_leading_zeros && *p == '0')
 				++leading_zero_cnt;
 			if (*p != '0')
@@ -375,11 +369,11 @@ ensure_decimal_point(char* buffer, size_t buf_size)
 		/* Skip leading sign, if present.  I think this could only
 		   ever be '-', but it can't hurt to check for both. */
 		++p;
-	while (*p && isdigit(Py_CHARMASK(*p)))
+	while (*p && Py_ISDIGIT(*p))
 		++p;
 
 	if (*p == '.') {
-		if (isdigit(Py_CHARMASK(*(p+1)))) {
+		if (Py_ISDIGIT(*(p+1))) {
 			/* Nothing to do, we already have a decimal
 			   point and a digit after it */
 		}