diff options
| -rw-r--r-- | Misc/NEWS | 4 | ||||
| -rw-r--r-- | Modules/_localemodule.c | 127 | ||||
| -rwxr-xr-x | configure | 5 | ||||
| -rw-r--r-- | configure.in | 2 | ||||
| -rw-r--r-- | pyconfig.h.in | 3 |
5 files changed, 96 insertions, 45 deletions
@@ -9,6 +9,10 @@ What's New in Python 3.0a4? *Release date: XX-XXX-2008* +Extension Modules +----------------- + +- Use wchar_t functions in _locale module. What's New in Python 3.0a3? diff --git a/Modules/_localemodule.c b/Modules/_localemodule.c index 5677b90..8a32677 100644 --- a/Modules/_localemodule.c +++ b/Modules/_localemodule.c @@ -1,5 +1,5 @@ /*********************************************************** -Copyright (C) 1997, 2002, 2003, 2007 Martin von Loewis +Copyright (C) 1997, 2002, 2003, 2007, 2008 Martin von Loewis Permission to use, copy, modify, and distribute this software and its documentation for any purpose and without fee is hereby granted, @@ -45,6 +45,35 @@ PyDoc_STRVAR(locale__doc__, "Support for POSIX locales."); static PyObject *Error; +/* Convert a char* to a Unicode object according to the current locale */ +static PyObject* +str2uni(const char* s) +{ + size_t needed = mbstowcs(NULL, s, 0); + size_t res1; + wchar_t smallbuf[30]; + wchar_t *dest; + PyObject *res2; + if (needed == (size_t)-1) { + PyErr_SetString(PyExc_ValueError, "Cannot convert byte to string"); + return NULL; + } + if (needed < sizeof(smallbuf)) + dest = smallbuf; + else { + dest = PyMem_Malloc(needed+1); + if (!dest) + return PyErr_NoMemory(); + } + /* This shouldn't fail now */ + res1 = mbstowcs(dest, s, needed+1); + assert(res == needed); + res2 = PyUnicode_FromWideChar(dest, res1); + if (dest != smallbuf) + PyMem_Free(dest); + return res2; +} + /* support functions for formatting floating point numbers */ PyDoc_STRVAR(setlocale__doc__, @@ -107,7 +136,7 @@ PyLocale_setlocale(PyObject* self, PyObject* args) PyErr_SetString(Error, "unsupported locale setting"); return NULL; } - result_object = PyUnicode_FromString(result); + result_object = str2uni(result); if (!result_object) return NULL; } else { @@ -117,7 +146,7 @@ PyLocale_setlocale(PyObject* self, PyObject* args) PyErr_SetString(Error, "locale query failed"); return NULL; } - result_object = PyUnicode_FromString(result); + result_object = str2uni(result); } return result_object; } @@ -143,7 +172,7 @@ PyLocale_localeconv(PyObject* self) involved herein */ #define RESULT_STRING(s)\ - x = PyUnicode_DecodeUnicodeEscape(l->s, strlen(l->s), "strict");\ + x = str2uni(l->s); \ if (!x) goto failed;\ PyDict_SetItemString(result, #s, x);\ Py_XDECREF(x) @@ -191,29 +220,19 @@ PyLocale_localeconv(PyObject* self) return NULL; } +#if defined(HAVE_WCSCOLL) PyDoc_STRVAR(strcoll__doc__, "string,string -> int. Compares two strings according to the locale."); static PyObject* PyLocale_strcoll(PyObject* self, PyObject* args) { -#if !defined(HAVE_WCSCOLL) - char *s1,*s2; - - if (!PyArg_ParseTuple(args, "ss:strcoll", &s1, &s2)) - return NULL; - return PyLong_FromLong(strcoll(s1, s2)); -#else PyObject *os1, *os2, *result = NULL; wchar_t *ws1 = NULL, *ws2 = NULL; - int len1, len2; + Py_ssize_t len1, len2; - if (!PyArg_UnpackTuple(args, "strcoll", 2, 2, &os1, &os2)) + if (!PyArg_ParseTuple(args, "UU:strcoll", &os1, &os2)) return NULL; - /* Both arguments must be unicode, or it's an error. */ - if (!PyUnicode_Check(os1) || !PyUnicode_Check(os2)) { - PyErr_SetString(PyExc_ValueError, "strcoll arguments must be strings"); - } /* Convert the unicode strings to wchar[]. */ len1 = PyUnicode_GET_SIZE(os1) + 1; ws1 = PyMem_MALLOC(len1 * sizeof(wchar_t)); @@ -240,40 +259,62 @@ PyLocale_strcoll(PyObject* self, PyObject* args) if (ws1) PyMem_FREE(ws1); if (ws2) PyMem_FREE(ws2); return result; -#endif } +#endif - +#ifdef HAVE_WCSXFRM PyDoc_STRVAR(strxfrm__doc__, "string -> string. Returns a string that behaves for cmp locale-aware."); static PyObject* PyLocale_strxfrm(PyObject* self, PyObject* args) { - char *s, *buf; + Py_UNICODE *s0; + Py_ssize_t n0; + wchar_t *s, *buf = NULL; size_t n1, n2; - PyObject *result; + PyObject *result = NULL; + Py_ssize_t i; - if (!PyArg_ParseTuple(args, "s:strxfrm", &s)) + if (!PyArg_ParseTuple(args, "u#:strxfrm", &s0, &n0)) return NULL; +#ifdef HAVE_USABLE_WCHAR_T + s = s0; +#else + s = PyMem_Malloc(n0+1); + if (!s) + return PyErr_NoMemory(); + for (i=0; i<=n0; i++) + s[i] = s0[i]; +#endif + /* assume no change in size, first */ - n1 = strlen(s) + 1; + n1 = wcslen(s) + 1; buf = PyMem_Malloc(n1); - if (!buf) - return PyErr_NoMemory(); - n2 = strxfrm(buf, s, n1) + 1; - if (n2 > n1) { + if (!buf) { + PyErr_NoMemory(); + goto exit; + } + n2 = wcsxfrm(buf, s, n1); + if (n2 >= n1) { /* more space needed */ - buf = PyMem_Realloc(buf, n2); - if (!buf) - return PyErr_NoMemory(); - strxfrm(buf, s, n2); + buf = PyMem_Realloc(buf, n2+1); + if (!buf) { + PyErr_NoMemory(); + goto exit; + } + n2 = wcsxfrm(buf, s, n2); } - result = PyUnicode_FromString(buf); - PyMem_Free(buf); + result = PyUnicode_FromWideChar(buf, n2); + exit: + if (buf) PyMem_Free(buf); +#ifdef HAVE_USABLE_WCHAR_T + PyMem_Free(s); +#endif return result; } +#endif #if defined(MS_WINDOWS) static PyObject* @@ -472,9 +513,7 @@ PyLocale_nl_langinfo(PyObject* self, PyObject* args) instead of an empty string for nl_langinfo(ERA). */ const char *result = nl_langinfo(item); result = result != NULL ? result : ""; - /* XXX may have to convert this to wcs first. */ - return PyUnicode_DecodeUnicodeEscape(result, strlen(result), - "strict"); + return str2uni(result); } PyErr_SetString(PyExc_ValueError, "unsupported langinfo constant"); return NULL; @@ -493,7 +532,7 @@ PyIntl_gettext(PyObject* self, PyObject *args) char *in; if (!PyArg_ParseTuple(args, "z", &in)) return 0; - return PyUnicode_FromString(gettext(in)); + return str2uni(gettext(in)); } PyDoc_STRVAR(dgettext__doc__, @@ -506,7 +545,7 @@ PyIntl_dgettext(PyObject* self, PyObject *args) char *domain, *in; if (!PyArg_ParseTuple(args, "zz", &domain, &in)) return 0; - return PyUnicode_FromString(dgettext(domain, in)); + return str2uni(dgettext(domain, in)); } PyDoc_STRVAR(dcgettext__doc__, @@ -520,7 +559,7 @@ PyIntl_dcgettext(PyObject *self, PyObject *args) int category; if (!PyArg_ParseTuple(args, "zzi", &domain, &msgid, &category)) return 0; - return PyUnicode_FromString(dcgettext(domain,msgid,category)); + return str2uni(dcgettext(domain,msgid,category)); } PyDoc_STRVAR(textdomain__doc__, @@ -538,7 +577,7 @@ PyIntl_textdomain(PyObject* self, PyObject* args) PyErr_SetFromErrno(PyExc_OSError); return NULL; } - return PyUnicode_FromString(domain); + return str2uni(domain); } PyDoc_STRVAR(bindtextdomain__doc__, @@ -556,7 +595,7 @@ PyIntl_bindtextdomain(PyObject* self,PyObject*args) PyErr_SetFromErrno(PyExc_OSError); return NULL; } - return PyUnicode_FromString(dirname); + return str2uni(dirname); } #ifdef HAVE_BIND_TEXTDOMAIN_CODESET @@ -572,7 +611,7 @@ PyIntl_bind_textdomain_codeset(PyObject* self,PyObject*args) return NULL; codeset = bind_textdomain_codeset(domain, codeset); if (codeset) - return PyUnicode_FromString(codeset); + return str2uni(codeset); Py_RETURN_NONE; } #endif @@ -584,10 +623,14 @@ static struct PyMethodDef PyLocale_Methods[] = { METH_VARARGS, setlocale__doc__}, {"localeconv", (PyCFunction) PyLocale_localeconv, METH_NOARGS, localeconv__doc__}, +#ifdef HAVE_WCSCOLL {"strcoll", (PyCFunction) PyLocale_strcoll, METH_VARARGS, strcoll__doc__}, +#endif +#ifdef HAVE_WCSXFRM {"strxfrm", (PyCFunction) PyLocale_strxfrm, METH_VARARGS, strxfrm__doc__}, +#endif #if defined(MS_WINDOWS) || defined(__APPLE__) {"_getdefaultlocale", (PyCFunction) PyLocale_getdefaultlocale, METH_NOARGS}, #endif @@ -1,5 +1,5 @@ #! /bin/sh -# From configure.in Revision: 60787 . +# From configure.in Revision: 61238 . # Guess values for system-dependent variables and create Makefiles. # Generated by GNU Autoconf 2.61 for python 3.0. # @@ -15715,6 +15715,7 @@ echo "${ECHO_T}MACHDEP_OBJS" >&6; } + for ac_func in alarm bind_textdomain_codeset chown clock confstr \ ctermid execv fchmod fchown fork fpathconf ftime ftruncate \ gai_strerror getgroups getlogin getloadavg getpeername getpgid getpid \ @@ -15726,7 +15727,7 @@ for ac_func in alarm bind_textdomain_codeset chown clock confstr \ setlocale setregid setreuid setsid setpgid setpgrp setuid setvbuf snprintf \ sigaction siginterrupt sigrelse strftime strlcpy \ sysconf tcgetpgrp tcsetpgrp tempnam timegm times tmpfile tmpnam tmpnam_r \ - truncate uname unsetenv utimes waitpid wait3 wait4 wcscoll _getpty + truncate uname unsetenv utimes waitpid wait3 wait4 wcscoll wcsxfrm _getpty do as_ac_var=`echo "ac_cv_func_$ac_func" | $as_tr_sh` { echo "$as_me:$LINENO: checking for $ac_func" >&5 diff --git a/configure.in b/configure.in index 0959682..129b688 100644 --- a/configure.in +++ b/configure.in @@ -2284,7 +2284,7 @@ AC_CHECK_FUNCS(alarm bind_textdomain_codeset chown clock confstr \ setlocale setregid setreuid setsid setpgid setpgrp setuid setvbuf snprintf \ sigaction siginterrupt sigrelse strftime strlcpy \ sysconf tcgetpgrp tcsetpgrp tempnam timegm times tmpfile tmpnam tmpnam_r \ - truncate uname unsetenv utimes waitpid wait3 wait4 wcscoll _getpty) + truncate uname unsetenv utimes waitpid wait3 wait4 wcscoll wcsxfrm _getpty) # For some functions, having a definition is not sufficient, since # we want to take their address. diff --git a/pyconfig.h.in b/pyconfig.h.in index df0a6d9..1a0460d 100644 --- a/pyconfig.h.in +++ b/pyconfig.h.in @@ -776,6 +776,9 @@ /* Define to 1 if you have the `wcscoll' function. */ #undef HAVE_WCSCOLL +/* Define to 1 if you have the `wcsxfrm' function. */ +#undef HAVE_WCSXFRM + /* Define if tzset() actually switches the local timezone in a meaningful way. */ #undef HAVE_WORKING_TZSET |