Change command line processing API to use wchar_t.

Fixes #2128.

Doc/c-api/init.rst

@@ -161,7 +161,7 @@ Initialization, Finalization, and Threads
    haven't been explicitly destroyed at that point.
 
 
-.. cfunction:: void Py_SetProgramName(char *name)
+.. cfunction:: void Py_SetProgramName(wchar_t *name)
 
    .. index::
       single: Py_Initialize()
@@ -170,11 +170,12 @@ Initialization, Finalization, and Threads
 
    This function should be called before :cfunc:`Py_Initialize` is called for
    the first time, if it is called at all.  It tells the interpreter the value
-   of the ``argv[0]`` argument to the :cfunc:`main` function of the program.
+   of the ``argv[0]`` argument to the :cfunc:`main` function of the program
+   (converted to wide characters).
    This is used by :cfunc:`Py_GetPath` and some other functions below to find
    the Python run-time libraries relative to the interpreter executable.  The
    default value is ``'python'``.  The argument should point to a
-   zero-terminated character string in static storage whose contents will not
+   zero-terminated wide character string in static storage whose contents will not
    change for the duration of the program's execution.  No code in the Python
    interpreter will change the contents of this storage.
 
@@ -188,7 +189,7 @@ Initialization, Finalization, and Threads
    value.
 
 
-.. cfunction:: char* Py_GetPrefix()
+.. cfunction:: wchar_t* Py_GetPrefix()
 
    Return the *prefix* for installed platform-independent files. This is derived
    through a number of complicated rules from the program name set with
@@ -201,7 +202,7 @@ Initialization, Finalization, and Threads
    It is only useful on Unix.  See also the next function.
 
 
-.. cfunction:: char* Py_GetExecPrefix()
+.. cfunction:: wchar_t* Py_GetExecPrefix()
 
    Return the *exec-prefix* for installed platform-*dependent* files.  This is
    derived through a number of complicated rules from the program name set with
@@ -236,7 +237,7 @@ Initialization, Finalization, and Threads
    platform.
 
 
-.. cfunction:: char* Py_GetProgramFullPath()
+.. cfunction:: wchar_t* Py_GetProgramFullPath()
 
    .. index::
       single: Py_SetProgramName()
@@ -249,7 +250,7 @@ Initialization, Finalization, and Threads
    to Python code as ``sys.executable``.
 
 
-.. cfunction:: char* Py_GetPath()
+.. cfunction:: wchar_t* Py_GetPath()
 
    .. index::
       triple: module; search; path
@@ -342,7 +343,7 @@ Initialization, Finalization, and Threads
    ``sys.version``.
 
 
-.. cfunction:: void PySys_SetArgv(int argc, char **argv)
+.. cfunction:: void PySys_SetArgv(int argc, wchar_t **argv)
 
    .. index::
       single: main()

Doc/c-api/sys.rst

@@ -84,11 +84,11 @@ accessible to C code.  They all work with the current interpreter thread's
 
    Reset :data:`sys.warnoptions` to an empty list.
 
-.. cfunction:: void PySys_AddWarnOption(char *s)
+.. cfunction:: void PySys_AddWarnOption(wchar_t *s)
 
    Append *s* to :data:`sys.warnoptions`.
 
-.. cfunction:: void PySys_SetPath(char *path)
+.. cfunction:: void PySys_SetPath(wchar_t *path)
 
    Set :data:`sys.path` to a list object of paths found in *path* which should
    be a list of paths separated with the platform's search path delimiter

Doc/c-api/unicode.rst

@@ -336,6 +336,8 @@ the system's :ctype:`wchar_t`.
 .. cfunction:: PyObject* PyUnicode_FromWideChar(const wchar_t *w, Py_ssize_t size)
 
    Create a Unicode object from the :ctype:`wchar_t` buffer *w* of the given size.
+   Passing -1 as the size indicates that the function must itself compute the length,
+   using wcslen.
    Return *NULL* on failure.
 
 

Doc/c-api/veryhigh.rst

@@ -25,16 +25,18 @@ are only passed to these functions if it is certain that they were created by
 the same library that the Python runtime is using.
 
 
-.. cfunction:: int Py_Main(int argc, char **argv)
+.. cfunction:: int Py_Main(int argc, wchar_t **argv)
 
-   The main program for the standard interpreter.  This is made available for
+   The main program for the standard interpreter.  This is made
-   programs which embed Python.  The *argc* and *argv* parameters should be
+   available for programs which embed Python.  The *argc* and *argv*
-   prepared exactly as those which are passed to a C program's :cfunc:`main`
+   parameters should be prepared exactly as those which are passed to
-   function.  It is important to note that the argument list may be modified (but
+   a C program's :cfunc:`main` function (converted to wchar_t
-   the contents of the strings pointed to by the argument list are not). The return
+   according to the user's locale).  It is important to note that the
-   value will be the integer passed to the :func:`sys.exit` function, ``1`` if the
+   argument list may be modified (but the contents of the strings
-   interpreter exits due to an exception, or ``2`` if the parameter list does not
+   pointed to by the argument list are not). The return value will be
-   represent a valid Python command line.
+   the integer passed to the :func:`sys.exit` function, ``1`` if the
+   interpreter exits due to an exception, or ``2`` if the parameter
+   list does not represent a valid Python command line.
 
 
 .. cfunction:: int PyRun_AnyFile(FILE *fp, const char *filename)

Include/osdefs.h

@@ -12,20 +12,20 @@ extern "C" {
 #if defined(MS_WINDOWS) || defined(__BORLANDC__) || defined(__WATCOMC__) || defined(__DJGPP__) || defined(PYOS_OS2)
 #if defined(PYOS_OS2) && defined(PYCC_GCC)
 #define MAXPATHLEN 260
-#define SEP '/'
+#define SEP L'/'
-#define ALTSEP '\\'
+#define ALTSEP L'\\'
 #else
-#define SEP '\\'
+#define SEP L'\\'
-#define ALTSEP '/'
+#define ALTSEP L'/'
 #define MAXPATHLEN 256
 #endif
-#define DELIM ';'
+#define DELIM L';'
 #endif
 #endif
 
 /* Filename separator */
 #ifndef SEP
-#define SEP '/'
+#define SEP L'/'
 #endif
 
 /* Max pathname length */
@@ -39,7 +39,7 @@ extern "C" {
 
 /* Search path entry delimiter */
 #ifndef DELIM
-#define DELIM ':'
+#define DELIM L':'
 #endif
 
 #ifdef __cplusplus

Include/pygetopt.h

@@ -7,9 +7,9 @@ extern "C" {
 
 PyAPI_DATA(int) _PyOS_opterr;
 PyAPI_DATA(int) _PyOS_optind;
-PyAPI_DATA(char *) _PyOS_optarg;
+PyAPI_DATA(wchar_t *) _PyOS_optarg;
 
-PyAPI_FUNC(int) _PyOS_GetOpt(int argc, char **argv, char *optstring);
+PyAPI_FUNC(int) _PyOS_GetOpt(int argc, wchar_t **argv, wchar_t *optstring);
 
 #ifdef __cplusplus
 }

Include/pythonrun.h

@@ -17,11 +17,11 @@ typedef struct {
 	int cf_flags;  /* bitmask of CO_xxx flags relevant to future */
 } PyCompilerFlags;
 
-PyAPI_FUNC(void) Py_SetProgramName(char *);
+PyAPI_FUNC(void) Py_SetProgramName(wchar_t *);
-PyAPI_FUNC(char *) Py_GetProgramName(void);
+PyAPI_FUNC(wchar_t *) Py_GetProgramName(void);
 
-PyAPI_FUNC(void) Py_SetPythonHome(char *);
+PyAPI_FUNC(void) Py_SetPythonHome(wchar_t *);
-PyAPI_FUNC(char *) Py_GetPythonHome(void);
+PyAPI_FUNC(wchar_t *) Py_GetPythonHome(void);
 
 PyAPI_FUNC(void) Py_Initialize(void);
 PyAPI_FUNC(void) Py_InitializeEx(int);
@@ -81,7 +81,7 @@ PyAPI_FUNC(void) Py_Exit(int);
 PyAPI_FUNC(int) Py_FdIsInteractive(FILE *, const char *);
 
 /* Bootstrap */
-PyAPI_FUNC(int) Py_Main(int argc, char **argv);
+PyAPI_FUNC(int) Py_Main(int argc, wchar_t **argv);
 
 /* Use macros for a bunch of old variants */
 #define PyRun_String(str, s, g, l) PyRun_StringFlags(str, s, g, l, NULL)
@@ -103,10 +103,10 @@ PyAPI_FUNC(int) Py_Main(int argc, char **argv);
         PyRun_FileExFlags(fp, p, s, g, l, 0, flags)
 
 /* In getpath.c */
-PyAPI_FUNC(char *) Py_GetProgramFullPath(void);
+PyAPI_FUNC(wchar_t *) Py_GetProgramFullPath(void);
-PyAPI_FUNC(char *) Py_GetPrefix(void);
+PyAPI_FUNC(wchar_t *) Py_GetPrefix(void);
-PyAPI_FUNC(char *) Py_GetExecPrefix(void);
+PyAPI_FUNC(wchar_t *) Py_GetExecPrefix(void);
-PyAPI_FUNC(char *) Py_GetPath(void);
+PyAPI_FUNC(wchar_t *) Py_GetPath(void);
 
 /* In their own files */
 PyAPI_FUNC(const char *) Py_GetVersion(void);

Include/sysmodule.h

@@ -9,8 +9,8 @@ extern "C" {
 
 PyAPI_FUNC(PyObject *) PySys_GetObject(const char *);
 PyAPI_FUNC(int) PySys_SetObject(const char *, PyObject *);
-PyAPI_FUNC(void) PySys_SetArgv(int, char **);
+PyAPI_FUNC(void) PySys_SetArgv(int, wchar_t **);
-PyAPI_FUNC(void) PySys_SetPath(const char *);
+PyAPI_FUNC(void) PySys_SetPath(const wchar_t *);
 
 PyAPI_FUNC(void) PySys_WriteStdout(const char *format, ...)
 			Py_GCC_ATTRIBUTE((format(printf, 1, 2)));
@@ -21,7 +21,7 @@ PyAPI_DATA(PyObject *) _PySys_TraceFunc, *_PySys_ProfileFunc;
 PyAPI_DATA(int) _PySys_CheckInterval;
 
 PyAPI_FUNC(void) PySys_ResetWarnOptions(void);
-PyAPI_FUNC(void) PySys_AddWarnOption(const char *);
+PyAPI_FUNC(void) PySys_AddWarnOption(const wchar_t *);
 
 #ifdef __cplusplus
 }

Misc/NEWS

@@ -12,6 +12,10 @@ What's New in Python 3.0a5?
 Core and Builtins
 -----------------
 
+- The command line processing was converted to pass Unicode strings
+  through as unmodified as possible; as a consequence, the C API
+  related to command line arguments was changed to use wchar_t.
+
 Extension Modules
 -----------------
 

Modules/_tkinter.c

@@ -3012,7 +3012,7 @@ ins_string(PyObject *d, char *name, char *val)
 PyMODINIT_FUNC
 init_tkinter(void)
 {
-	PyObject *m, *d;
+	PyObject *m, *d, *uexe, *cexe;
 
 	Py_TYPE(&Tkapp_Type) = &PyType_Type;
 
@@ -3065,7 +3065,16 @@ init_tkinter(void)
 
 	/* This helps the dynamic loader; in Unicode aware Tcl versions
 	   it also helps Tcl find its encodings. */
-	Tcl_FindExecutable(Py_GetProgramName());
+	uexe = PyUnicode_FromWideChar(Py_GetProgramName(), -1);
+	if (uexe) {
+		cexe = PyUnicode_AsEncodedString(uexe, 
+						 Py_FileSystemDefaultEncoding, 
+						 NULL);
+		if (cexe)
+			Tcl_FindExecutable(PyString_AsString(cexe));
+		Py_XDECREF(cexe);
+		Py_DECREF(uexe);
+	}
 
 	if (PyErr_Occurred())
 		return;

Modules/getpath.c

@@ -122,19 +122,81 @@
 #endif
 
 #ifndef LANDMARK
-#define LANDMARK "os.py"
+#define LANDMARK L"os.py"
 #endif
 
-static char prefix[MAXPATHLEN+1];
+static wchar_t prefix[MAXPATHLEN+1];
-static char exec_prefix[MAXPATHLEN+1];
+static wchar_t exec_prefix[MAXPATHLEN+1];
-static char progpath[MAXPATHLEN+1];
+static wchar_t progpath[MAXPATHLEN+1];
-static char *module_search_path = NULL;
+static wchar_t *module_search_path = NULL;
-static char lib_python[] = "lib/python" VERSION;
+static wchar_t lib_python[] = L"lib/python" VERSION;
+
+/* In principle, this should use HAVE__WSTAT, and _wstat
+   should be detected by autoconf. However, no current
+   POSIX system provides that function, so testing for
+   it is pointless. */
+#ifndef MS_WINDOWS
+static int
+_wstat(const wchar_t* path, struct stat *buf)
+{
+    char fname[PATH_MAX];
+    size_t res = wcstombs(fname, path, sizeof(fname));
+    if (res == (size_t)-1) {
+	errno = EINVAL;
+	return -1;
+    }
+    return stat(fname, buf);
+}
+#endif
+
+#ifndef MS_WINDOWS
+static wchar_t*
+_wgetcwd(wchar_t *buf, size_t size)
+{
+    char fname[PATH_MAX];
+    if (getcwd(fname, PATH_MAX) == NULL)
+	return NULL;
+    if (mbstowcs(buf, fname, size) >= size) {
+	errno = ERANGE;
+	return NULL;
+    }
+    return buf;
+}
+#endif
+
+#ifdef HAVE_READLINK
+int 
+_Py_wreadlink(const wchar_t *path, wchar_t *buf, size_t bufsiz)
+{
+    char cbuf[PATH_MAX];
+    char cpath[PATH_MAX];
+    int res;
+    size_t r1 = wcstombs(cpath, path, PATH_MAX);
+    if (r1 == (size_t)-1 || r1 >= PATH_MAX) {
+	errno = EINVAL;
+	return -1;
+    }
+    res = (int)readlink(cpath, cbuf, PATH_MAX);
+    if (res == -1)
+	return -1;
+    if (res == PATH_MAX) {
+	errno = EINVAL;
+	return -1;
+    }
+    r1 = mbstowcs(buf, cbuf, bufsiz);
+    if (r1 == -1) {
+	errno = EINVAL;
+	return -1;
+    }
+    return (int)r1;
+    
+}
+#endif
 
 static void
-reduce(char *dir)
+reduce(wchar_t *dir)
 {
-    size_t i = strlen(dir);
+    size_t i = wcslen(dir);
     while (i > 0 && dir[i] != SEP)
         --i;
     dir[i] = '\0';
@@ -142,10 +204,10 @@ reduce(char *dir)
 
 
 static int
-isfile(char *filename)          /* Is file, not directory */
+isfile(wchar_t *filename)          /* Is file, not directory */
 {
     struct stat buf;
-    if (stat(filename, &buf) != 0)
+    if (_wstat(filename, &buf) != 0)
         return 0;
     if (!S_ISREG(buf.st_mode))
         return 0;
@@ -154,14 +216,14 @@ isfile(char *filename)          /* Is file, not directory */
 
 
 static int
-ismodule(char *filename)        /* Is module -- check for .pyc/.pyo too */
+ismodule(wchar_t *filename)        /* Is module -- check for .pyc/.pyo too */
 {
     if (isfile(filename))
         return 1;
 
     /* Check for the compiled version of prefix. */
-    if (strlen(filename) < MAXPATHLEN) {
+    if (wcslen(filename) < MAXPATHLEN) {
-        strcat(filename, Py_OptimizeFlag ? "o" : "c");
+        wcscat(filename, Py_OptimizeFlag ? L"o" : L"c");
         if (isfile(filename))
             return 1;
     }
@@ -170,10 +232,10 @@ ismodule(char *filename)        /* Is module -- check for .pyc/.pyo too */
 
 
 static int
-isxfile(char *filename)         /* Is executable file */
+isxfile(wchar_t *filename)         /* Is executable file */
 {
     struct stat buf;
-    if (stat(filename, &buf) != 0)
+    if (_wstat(filename, &buf) != 0)
         return 0;
     if (!S_ISREG(buf.st_mode))
         return 0;
@@ -184,10 +246,10 @@ isxfile(char *filename)         /* Is executable file */
 
 
 static int
-isdir(char *filename)                   /* Is directory */
+isdir(wchar_t *filename)                   /* Is directory */
 {
     struct stat buf;
-    if (stat(filename, &buf) != 0)
+    if (_wstat(filename, &buf) != 0)
         return 0;
     if (!S_ISDIR(buf.st_mode))
         return 0;
@@ -205,34 +267,34 @@ isdir(char *filename)                   /* Is directory */
    stuff as fits will be appended.
 */
 static void
-joinpath(char *buffer, char *stuff)
+joinpath(wchar_t *buffer, wchar_t *stuff)
 {
     size_t n, k;
     if (stuff[0] == SEP)
         n = 0;
     else {
-        n = strlen(buffer);
+        n = wcslen(buffer);
         if (n > 0 && buffer[n-1] != SEP && n < MAXPATHLEN)
             buffer[n++] = SEP;
     }
     if (n > MAXPATHLEN)
     	Py_FatalError("buffer overflow in getpath.c's joinpath()");
-    k = strlen(stuff);
+    k = wcslen(stuff);
     if (n + k > MAXPATHLEN)
         k = MAXPATHLEN - n;
-    strncpy(buffer+n, stuff, k);
+    wcsncpy(buffer+n, stuff, k);
     buffer[n+k] = '\0';
 }
 
 /* copy_absolute requires that path be allocated at least
    MAXPATHLEN + 1 bytes and that p be no more than MAXPATHLEN bytes. */
 static void
-copy_absolute(char *path, char *p)
+copy_absolute(wchar_t *path, wchar_t *p)
 {
     if (p[0] == SEP)
-        strcpy(path, p);
+        wcscpy(path, p);
     else {
-        getcwd(path, MAXPATHLEN);
+        _wgetcwd(path, MAXPATHLEN);
         if (p[0] == '.' && p[1] == SEP)
             p += 2;
         joinpath(path, p);
@@ -241,46 +303,46 @@ copy_absolute(char *path, char *p)
 
 /* absolutize() requires that path be allocated at least MAXPATHLEN+1 bytes. */
 static void
-absolutize(char *path)
+absolutize(wchar_t *path)
 {
-    char buffer[MAXPATHLEN + 1];
+    wchar_t buffer[MAXPATHLEN + 1];
 
     if (path[0] == SEP)
         return;
     copy_absolute(buffer, path);
-    strcpy(path, buffer);
+    wcscpy(path, buffer);
 }
 
 /* search_for_prefix requires that argv0_path be no more than MAXPATHLEN
    bytes long.
 */
 static int
-search_for_prefix(char *argv0_path, char *home)
+search_for_prefix(wchar_t *argv0_path, wchar_t *home)
 {
     size_t n;
-    char *vpath;
+    wchar_t *vpath;
 
     /* If PYTHONHOME is set, we believe it unconditionally */
     if (home) {
-        char *delim;
+        wchar_t *delim;
-        strncpy(prefix, home, MAXPATHLEN);
+        wcsncpy(prefix, home, MAXPATHLEN);
-        delim = strchr(prefix, DELIM);
+        delim = wcschr(prefix, DELIM);
         if (delim)
-            *delim = '\0';
+            *delim = L'\0';
         joinpath(prefix, lib_python);
         joinpath(prefix, LANDMARK);
         return 1;
     }
 
     /* Check to see if argv[0] is in the build directory */
-    strcpy(prefix, argv0_path);
+    wcscpy(prefix, argv0_path);
-    joinpath(prefix, "Modules/Setup");
+    joinpath(prefix, L"Modules/Setup");
     if (isfile(prefix)) {
         /* Check VPATH to see if argv0_path is in the build directory. */
-        vpath = VPATH;
+        vpath = L"" VPATH;
-        strcpy(prefix, argv0_path);
+        wcscpy(prefix, argv0_path);
         joinpath(prefix, vpath);
-        joinpath(prefix, "Lib");
+        joinpath(prefix, L"Lib");
         joinpath(prefix, LANDMARK);
         if (ismodule(prefix))
             return -1;
@@ -289,17 +351,17 @@ search_for_prefix(char *argv0_path, char *home)
     /* Search from argv0_path, until root is found */
     copy_absolute(prefix, argv0_path);
     do {
-        n = strlen(prefix);
+        n = wcslen(prefix);
         joinpath(prefix, lib_python);
         joinpath(prefix, LANDMARK);
         if (ismodule(prefix))
             return 1;
-        prefix[n] = '\0';
+        prefix[n] = L'\0';
         reduce(prefix);
     } while (prefix[0]);
 
     /* Look at configure's PREFIX */
-    strncpy(prefix, PREFIX, MAXPATHLEN);
+    wcsncpy(prefix, L"" PREFIX, MAXPATHLEN);
     joinpath(prefix, lib_python);
     joinpath(prefix, LANDMARK);
     if (ismodule(prefix))
@@ -314,26 +376,26 @@ search_for_prefix(char *argv0_path, char *home)
    MAXPATHLEN bytes long.
 */
 static int
-search_for_exec_prefix(char *argv0_path, char *home)
+search_for_exec_prefix(wchar_t *argv0_path, wchar_t *home)
 {
     size_t n;
 
     /* If PYTHONHOME is set, we believe it unconditionally */
     if (home) {
-        char *delim;
+        wchar_t *delim;
-        delim = strchr(home, DELIM);
+        delim = wcschr(home, DELIM);
         if (delim)
-            strncpy(exec_prefix, delim+1, MAXPATHLEN);
+            wcsncpy(exec_prefix, delim+1, MAXPATHLEN);
         else
-            strncpy(exec_prefix, home, MAXPATHLEN);
+            wcsncpy(exec_prefix, home, MAXPATHLEN);
         joinpath(exec_prefix, lib_python);
-        joinpath(exec_prefix, "lib-dynload");
+        joinpath(exec_prefix, L"lib-dynload");
         return 1;
     }
 
     /* Check to see if argv[0] is in the build directory */
-    strcpy(exec_prefix, argv0_path);
+    wcscpy(exec_prefix, argv0_path);
-    joinpath(exec_prefix, "Modules/Setup");
+    joinpath(exec_prefix, L"Modules/Setup");
     if (isfile(exec_prefix)) {
         reduce(exec_prefix);
         return -1;
@@ -342,19 +404,19 @@ search_for_exec_prefix(char *argv0_path, char *home)
     /* Search from argv0_path, until root is found */
     copy_absolute(exec_prefix, argv0_path);
     do {
-        n = strlen(exec_prefix);
+        n = wcslen(exec_prefix);
         joinpath(exec_prefix, lib_python);
-        joinpath(exec_prefix, "lib-dynload");
+        joinpath(exec_prefix, L"lib-dynload");
         if (isdir(exec_prefix))
             return 1;
-        exec_prefix[n] = '\0';
+        exec_prefix[n] = L'\0';
         reduce(exec_prefix);
     } while (exec_prefix[0]);
 
     /* Look at configure's EXEC_PREFIX */
-    strncpy(exec_prefix, EXEC_PREFIX, MAXPATHLEN);
+    wcsncpy(exec_prefix, L"" EXEC_PREFIX, MAXPATHLEN);
     joinpath(exec_prefix, lib_python);
-    joinpath(exec_prefix, "lib-dynload");
+    joinpath(exec_prefix, L"lib-dynload");
     if (isdir(exec_prefix))
         return 1;
 
@@ -366,22 +428,25 @@ search_for_exec_prefix(char *argv0_path, char *home)
 static void
 calculate_path(void)
 {
-    extern char *Py_GetProgramName(void);
+    extern wchar_t *Py_GetProgramName(void);
 
-    static char delimiter[2] = {DELIM, '\0'};
+    static wchar_t delimiter[2] = {DELIM, '\0'};
-    static char separator[2] = {SEP, '\0'};
+    static wchar_t separator[2] = {SEP, '\0'};
-    char *pythonpath = PYTHONPATH;
+    wchar_t *pythonpath = L"" PYTHONPATH;
-    char *rtpypath = Py_GETENV("PYTHONPATH");
+    char *_rtpypath = Py_GETENV("PYTHONPATH"); /* XXX use wide version on Windows */
-    char *home = Py_GetPythonHome();
+    wchar_t rtpypath[MAXPATHLEN+1];
-    char *path = getenv("PATH");
+    wchar_t *home = Py_GetPythonHome();
-    char *prog = Py_GetProgramName();
+    char *_path = getenv("PATH");
-    char argv0_path[MAXPATHLEN+1];
+    wchar_t wpath[MAXPATHLEN+1];
-    char zip_path[MAXPATHLEN+1];
+    wchar_t *path = NULL;
+    wchar_t *prog = Py_GetProgramName();
+    wchar_t argv0_path[MAXPATHLEN+1];
+    wchar_t zip_path[MAXPATHLEN+1];
     int pfound, efound; /* 1 if found; -1 if found build directory */
-    char *buf;
+    wchar_t *buf;
     size_t bufsz;
     size_t prefixsz;
-    char *defpath = pythonpath;
+    wchar_t *defpath = pythonpath;
 #ifdef WITH_NEXT_FRAMEWORK
     NSModule pythonModule;
 #endif
@@ -393,13 +458,22 @@ calculate_path(void)
 #endif
 #endif
 
+    if (_path) {
+	    size_t r = mbstowcs(wpath, _path, MAXPATHLEN+1);
+	    path = wpath;
+	    if (r == (size_t)-1 || r > MAXPATHLEN) {
+		    /* Could not convert PATH, or it's too long. */
+		    path = NULL;
+	    }
+    }
+
 	/* If there is no slash in the argv0 path, then we have to
 	 * assume python is on the user's $PATH, since there's no
 	 * other way to find a directory to start the search from.  If
 	 * $PATH isn't exported, you lose.
 	 */
-	if (strchr(prog, SEP))
+	if (wcschr(prog, SEP))
-		strncpy(progpath, prog, MAXPATHLEN);
+		wcsncpy(progpath, prog, MAXPATHLEN);
 #ifdef __APPLE__
      /* On Mac OS X, if a script uses an interpreter of the form
       * "#!/opt/python2.3/bin/python", the kernel only passes "python"
@@ -416,24 +490,24 @@ calculate_path(void)
 #endif /* __APPLE__ */
 	else if (path) {
 		while (1) {
-			char *delim = strchr(path, DELIM);
+			wchar_t *delim = wcschr(path, DELIM);
 
 			if (delim) {
 				size_t len = delim - path;
 				if (len > MAXPATHLEN)
 					len = MAXPATHLEN;
-				strncpy(progpath, path, len);
+				wcsncpy(progpath, path, len);
 				*(progpath + len) = '\0';
 			}
 			else
-				strncpy(progpath, path, MAXPATHLEN);
+				wcsncpy(progpath, path, MAXPATHLEN);
 
 			joinpath(progpath, prog);
 			if (isxfile(progpath))
 				break;
 
 			if (!delim) {
-				progpath[0] = '\0';
+				progpath[0] = L'\0';
 				break;
 			}
 			path = delim + 1;
@@ -443,7 +517,7 @@ calculate_path(void)
 		progpath[0] = '\0';
 	if (progpath[0] != SEP)
 		absolutize(progpath);
-	strncpy(argv0_path, progpath, MAXPATHLEN);
+	wcsncpy(argv0_path, progpath, MAXPATHLEN);
 	argv0_path[MAXPATHLEN] = '\0';
 
 #ifdef WITH_NEXT_FRAMEWORK
@@ -454,7 +528,7 @@ calculate_path(void)
 	*/
     pythonModule = NSModuleForSymbol(NSLookupAndBindSymbol("_Py_Initialize"));
     /* Use dylib functions to find out where the framework was loaded from */
-    buf = (char *)NSLibraryNameForModule(pythonModule);
+    buf = (wchar_t *)NSLibraryNameForModule(pythonModule);
     if (buf != NULL) {
         /* We're in a framework. */
         /* See if we might be in the build directory. The framework in the
@@ -464,39 +538,39 @@ calculate_path(void)
         ** be running the interpreter in the build directory, so we use the
         ** build-directory-specific logic to find Lib and such.
         */
-        strncpy(argv0_path, buf, MAXPATHLEN);
+        wcsncpy(argv0_path, buf, MAXPATHLEN);
         reduce(argv0_path);
         joinpath(argv0_path, lib_python);
         joinpath(argv0_path, LANDMARK);
         if (!ismodule(argv0_path)) {
                 /* We are in the build directory so use the name of the
                    executable - we know that the absolute path is passed */
-                strncpy(argv0_path, prog, MAXPATHLEN);
+                wcsncpy(argv0_path, prog, MAXPATHLEN);
         }
         else {
                 /* Use the location of the library as the progpath */
-                strncpy(argv0_path, buf, MAXPATHLEN);
+                wcsncpy(argv0_path, buf, MAXPATHLEN);
         }
     }
 #endif
 
 #if HAVE_READLINK
     {
-        char tmpbuffer[MAXPATHLEN+1];
+        wchar_t tmpbuffer[MAXPATHLEN+1];
-        int linklen = readlink(progpath, tmpbuffer, MAXPATHLEN);
+        int linklen = _Py_wreadlink(progpath, tmpbuffer, MAXPATHLEN);
         while (linklen != -1) {
             /* It's not null terminated! */
             tmpbuffer[linklen] = '\0';
             if (tmpbuffer[0] == SEP)
                 /* tmpbuffer should never be longer than MAXPATHLEN,
                    but extra check does not hurt */
-                strncpy(argv0_path, tmpbuffer, MAXPATHLEN);
+                wcsncpy(argv0_path, tmpbuffer, MAXPATHLEN);
             else {
                 /* Interpret relative to progpath */
                 reduce(argv0_path);
                 joinpath(argv0_path, tmpbuffer);
             }
-            linklen = readlink(argv0_path, tmpbuffer, MAXPATHLEN);
+            linklen = _Py_wreadlink(argv0_path, tmpbuffer, MAXPATHLEN);
         }
     }
 #endif /* HAVE_READLINK */
@@ -510,22 +584,22 @@ calculate_path(void)
         if (!Py_FrozenFlag)
             fprintf(stderr,
                 "Could not find platform independent libraries <prefix>\n");
-        strncpy(prefix, PREFIX, MAXPATHLEN);
+        wcsncpy(prefix, L"" PREFIX, MAXPATHLEN);
         joinpath(prefix, lib_python);
     }
     else
         reduce(prefix);
 
-    strncpy(zip_path, prefix, MAXPATHLEN);
+    wcsncpy(zip_path, prefix, MAXPATHLEN);
-    zip_path[MAXPATHLEN] = '\0';
+    zip_path[MAXPATHLEN] = L'\0';
     if (pfound > 0) { /* Use the reduced prefix returned by Py_GetPrefix() */
         reduce(zip_path);
         reduce(zip_path);
     }
     else
-        strncpy(zip_path, PREFIX, MAXPATHLEN);
+        wcsncpy(zip_path, L"" PREFIX, MAXPATHLEN);
-    joinpath(zip_path, "lib/python00.zip");
+    joinpath(zip_path, L"lib/python00.zip");
-    bufsz = strlen(zip_path);	/* Replace "00" with version */
+    bufsz = wcslen(zip_path);	/* Replace "00" with version */
     zip_path[bufsz - 6] = VERSION[0];
     zip_path[bufsz - 5] = VERSION[2];
 
@@ -533,8 +607,8 @@ calculate_path(void)
         if (!Py_FrozenFlag)
             fprintf(stderr,
                 "Could not find platform dependent libraries <exec_prefix>\n");
-        strncpy(exec_prefix, EXEC_PREFIX, MAXPATHLEN);
+        wcsncpy(exec_prefix, L"" EXEC_PREFIX, MAXPATHLEN);
-        joinpath(exec_prefix, "lib/lib-dynload");
+        joinpath(exec_prefix, L"lib/lib-dynload");
     }
     /* If we found EXEC_PREFIX do *not* reduce it!  (Yet.) */
 
@@ -546,13 +620,19 @@ calculate_path(void)
      */
     bufsz = 0;
 
-    if (rtpypath)
+    if (_rtpypath) {
-        bufsz += strlen(rtpypath) + 1;
+	size_t s = mbstowcs(rtpypath, _rtpypath, sizeof(rtpypath)/sizeof(wchar_t));
+	if (s == (size_t)-1 || s >=sizeof(rtpypath))
+	    /* XXX deal with errors more gracefully */
+	    _rtpypath = NULL;
+	if (_rtpypath)
+	    bufsz += wcslen(rtpypath) + 1;
+    }
 
-    prefixsz = strlen(prefix) + 1;
+    prefixsz = wcslen(prefix) + 1;
 
     while (1) {
-        char *delim = strchr(defpath, DELIM);
+        wchar_t *delim = wcschr(defpath, DELIM);
 
         if (defpath[0] != SEP)
             /* Paths are relative to prefix */
@@ -561,65 +641,65 @@ calculate_path(void)
         if (delim)
             bufsz += delim - defpath + 1;
         else {
-            bufsz += strlen(defpath) + 1;
+            bufsz += wcslen(defpath) + 1;
             break;
         }
         defpath = delim + 1;
     }
 
-    bufsz += strlen(zip_path) + 1;
+    bufsz += wcslen(zip_path) + 1;
-    bufsz += strlen(exec_prefix) + 1;
+    bufsz += wcslen(exec_prefix) + 1;
 
     /* This is the only malloc call in this file */
-    buf = (char *)PyMem_Malloc(bufsz);
+    buf = (wchar_t *)PyMem_Malloc(bufsz*sizeof(wchar_t));
 
     if (buf == NULL) {
         /* We can't exit, so print a warning and limp along */
         fprintf(stderr, "Not enough memory for dynamic PYTHONPATH.\n");
         fprintf(stderr, "Using default static PYTHONPATH.\n");
-        module_search_path = PYTHONPATH;
+        module_search_path = L"" PYTHONPATH;
     }
     else {
         /* Run-time value of $PYTHONPATH goes first */
-        if (rtpypath) {
+        if (_rtpypath) {
-            strcpy(buf, rtpypath);
+            wcscpy(buf, rtpypath);
-            strcat(buf, delimiter);
+            wcscat(buf, delimiter);
         }
         else
             buf[0] = '\0';
 
         /* Next is the default zip path */
-        strcat(buf, zip_path);
+        wcscat(buf, zip_path);
-        strcat(buf, delimiter);
+        wcscat(buf, delimiter);
 
         /* Next goes merge of compile-time $PYTHONPATH with
          * dynamically located prefix.
          */
         defpath = pythonpath;
         while (1) {
-            char *delim = strchr(defpath, DELIM);
+            wchar_t *delim = wcschr(defpath, DELIM);
 
             if (defpath[0] != SEP) {
-                strcat(buf, prefix);
+                wcscat(buf, prefix);
-                strcat(buf, separator);
+                wcscat(buf, separator);
             }
 
             if (delim) {
                 size_t len = delim - defpath + 1;
-                size_t end = strlen(buf) + len;
+                size_t end = wcslen(buf) + len;
-                strncat(buf, defpath, len);
+                wcsncat(buf, defpath, len);
                 *(buf + end) = '\0';
             }
             else {
-                strcat(buf, defpath);
+                wcscat(buf, defpath);
                 break;
             }
             defpath = delim + 1;
         }
-        strcat(buf, delimiter);
+        wcscat(buf, delimiter);
 
         /* Finally, on goes the directory for dynamic-load modules */
-        strcat(buf, exec_prefix);
+        wcscat(buf, exec_prefix);
 
         /* And publish the results */
         module_search_path = buf;
@@ -636,26 +716,26 @@ calculate_path(void)
 	/* The prefix is the root directory, but reduce() chopped
 	 * off the "/". */
 	if (!prefix[0])
-		strcpy(prefix, separator);
+		wcscpy(prefix, separator);
     }
     else
-        strncpy(prefix, PREFIX, MAXPATHLEN);
+        wcsncpy(prefix, L"" PREFIX, MAXPATHLEN);
 
     if (efound > 0) {
         reduce(exec_prefix);
         reduce(exec_prefix);
         reduce(exec_prefix);
 	if (!exec_prefix[0])
-		strcpy(exec_prefix, separator);
+		wcscpy(exec_prefix, separator);
     }
     else
-        strncpy(exec_prefix, EXEC_PREFIX, MAXPATHLEN);
+        wcsncpy(exec_prefix, L"" EXEC_PREFIX, MAXPATHLEN);
 }
 
 
 /* External interface */
 
-char *
+wchar_t *
 Py_GetPath(void)
 {
     if (!module_search_path)
@@ -663,7 +743,7 @@ Py_GetPath(void)
     return module_search_path;
 }
 
-char *
+wchar_t *
 Py_GetPrefix(void)
 {
     if (!module_search_path)
@@ -671,7 +751,7 @@ Py_GetPrefix(void)
     return prefix;
 }
 
-char *
+wchar_t *
 Py_GetExecPrefix(void)
 {
     if (!module_search_path)
@@ -679,7 +759,7 @@ Py_GetExecPrefix(void)
     return exec_prefix;
 }
 
-char *
+wchar_t *
 Py_GetProgramFullPath(void)
 {
     if (!module_search_path)

Modules/main.c

@@ -12,6 +12,7 @@
 #include <windows.h>
 #ifdef HAVE_FCNTL_H
 #include <fcntl.h>
+#define PATH_MAX MAXPATHLEN
 #endif
 #endif
 
@@ -40,17 +41,17 @@ extern "C" {
 #endif
 
 /* For Py_GetArgcArgv(); set by main() */
-static char **orig_argv;
+static wchar_t **orig_argv;
 static int  orig_argc;
 
 /* command line options */
-#define BASE_OPTS "bBc:dEhim:OStuvVW:xX?"
+#define BASE_OPTS L"bBc:dEhim:OStuvVW:xX?"
 
 #define PROGRAM_OPTS BASE_OPTS
 
 /* Short usage message (with %s for argv0) */
 static char *usage_line =
-"usage: %s [option] ... [-c cmd | -m mod | file | -] [arg] ...\n";
+"usage: %ls [option] ... [-c cmd | -m mod | file | -] [arg] ...\n";
 
 /* Long usage message, split into parts < 512 bytes */
 static char *usage_1 = "\
@@ -96,9 +97,30 @@ PYTHONHOME   : alternate <prefix> directory (or <prefix>%c<exec_prefix>).\n\
 PYTHONCASEOK : ignore case in 'import' statements (Windows).\n\
 ";
 
+#ifndef MS_WINDOWS
+static FILE*
+_wfopen(const wchar_t *path, const wchar_t *mode)
+{
+	char cpath[PATH_MAX];
+	char cmode[10];
+	size_t r;
+	r = wcstombs(cpath, path, PATH_MAX);
+	if (r == (size_t)-1 || r >= PATH_MAX) {
+		errno = EINVAL;
+		return NULL;
+	}
+	r = wcstombs(cmode, mode, 10);
+	if (r == (size_t)-1 || r >= 10) {
+		errno = EINVAL;
+		return NULL;
+	}
+	return fopen(cpath, cmode);
+}
+#endif
+
 
 static int
-usage(int exitcode, char* program)
+usage(int exitcode, wchar_t* program)
 {
 	FILE *f = exitcode ? stderr : stdout;
 
@@ -187,11 +209,11 @@ static int RunModule(char *module, int set_argv0)
 	return 0;
 }
 
-static int RunMainFromImporter(char *filename)
+static int RunMainFromImporter(wchar_t *filename)
 {
 	PyObject *argv0 = NULL, *importer = NULL;
 
-	if ((argv0 = PyUnicode_DecodeFSDefault(filename)) &&
+	if ((argv0 = PyUnicode_FromWideChar(filename,wcslen(filename))) &&
 	    (importer = PyImport_GetImporter(argv0)) &&
 	    (importer->ob_type != &PyNullImporter_Type))
 	{
@@ -249,12 +271,12 @@ WaitForThreadShutdown(void)
 /* Main program */
 
 int
-Py_Main(int argc, char **argv)
+Py_Main(int argc, wchar_t **argv)
 {
 	int c;
 	int sts;
 	char *command = NULL;
-	char *filename = NULL;
+	wchar_t *filename = NULL;
 	char *module = NULL;
 	FILE *fp = stdin;
 	char *p;
@@ -275,14 +297,19 @@ Py_Main(int argc, char **argv)
 
 	while ((c = _PyOS_GetOpt(argc, argv, PROGRAM_OPTS)) != EOF) {
 		if (c == 'c') {
+			size_t r1 = wcslen(_PyOS_optarg) + 2;
+			size_t r2;
 			/* -c is the last option; following arguments
 			   that look like options are left for the
 			   command to interpret. */
-			command = (char *)malloc(strlen(_PyOS_optarg) + 2);
+			command = (char *)malloc(r1);
 			if (command == NULL)
 				Py_FatalError(
 				   "not enough memory to copy -c argument");
-			strcpy(command, _PyOS_optarg);
+			r2 = wcstombs(command, _PyOS_optarg, r1);
+			if (r2 > r1-2)
+				Py_FatalError(
+				    "not enough memory to copy -c argument");
 			strcat(command, "\n");
 			break;
 		}
@@ -291,11 +318,16 @@ Py_Main(int argc, char **argv)
 			/* -m is the last option; following arguments
 			   that look like options are left for the
 			   module to interpret. */
-			module = (char *)malloc(strlen(_PyOS_optarg) + 2);
+			size_t r1 = wcslen(_PyOS_optarg) + 1;
+			size_t r2;
+			module = (char *)malloc(r1);
 			if (module == NULL)
 				Py_FatalError(
 				   "not enough memory to copy -m argument");
-			strcpy(module, _PyOS_optarg);
+			r2 = wcstombs(module, _PyOS_optarg, r1);
+			if (r2 >= r1)
+				Py_FatalError(
+				   "not enough memory to copy -m argument");
 			break;
 		}
 
@@ -355,7 +387,7 @@ Py_Main(int argc, char **argv)
 			version++;
 			break;
 
-		case 'W':
+		case 'W': 
 			PySys_AddWarnOption(_PyOS_optarg);
 			break;
 
@@ -384,7 +416,7 @@ Py_Main(int argc, char **argv)
 		unbuffered = 1;
 
 	if (command == NULL && module == NULL && _PyOS_optind < argc &&
-	    strcmp(argv[_PyOS_optind], "-") != 0)
+	    wcscmp(argv[_PyOS_optind], L"-") != 0)
 	{
 #ifdef __VMS
 		filename = decc$translate_vms(argv[_PyOS_optind]);
@@ -462,14 +494,14 @@ Py_Main(int argc, char **argv)
 	if (command != NULL) {
 		/* Backup _PyOS_optind and force sys.argv[0] = '-c' */
 		_PyOS_optind--;
-		argv[_PyOS_optind] = "-c";
+		argv[_PyOS_optind] = L"-c";
 	}
 
 	if (module != NULL) {
 		/* Backup _PyOS_optind and force sys.argv[0] = '-c'
 		   so that PySys_SetArgv correctly sets sys.path[0] to ''*/
 		_PyOS_optind--;
-		argv[_PyOS_optind] = "-c";
+		argv[_PyOS_optind] = L"-c";
 	}
 
 	PySys_SetArgv(argc-_PyOS_optind, argv+_PyOS_optind);
@@ -506,8 +538,8 @@ Py_Main(int argc, char **argv)
 		}
 
 		if (sts==-1 && filename!=NULL) {
-			if ((fp = fopen(filename, "r")) == NULL) {
+			if ((fp = _wfopen(filename, L"r")) == NULL) {
-				fprintf(stderr, "%s: can't open file '%s': [Errno %d] %s\n",
+				fprintf(stderr, "%s: can't open file '%ls': [Errno %d] %s\n",
 					argv[0], filename, errno, strerror(errno));
 
 				return 2;
@@ -528,7 +560,7 @@ Py_Main(int argc, char **argv)
 				struct stat sb;
 				if (fstat(fileno(fp), &sb) == 0 &&
 				    S_ISDIR(sb.st_mode)) {
-					fprintf(stderr, "%s: '%s' is a directory, cannot continue\n", argv[0], filename);
+					fprintf(stderr, "%ls: '%ls' is a directory, cannot continue\n", argv[0], filename);
 					fclose(fp);
 					return 1;
 				}
@@ -536,9 +568,17 @@ Py_Main(int argc, char **argv)
 		}
 
 		if (sts==-1) {
+			char cfilename[PATH_MAX];
+			char *p_cfilename = "<stdin>";
+			if (filename) {
+				size_t r = wcstombs(cfilename, filename, PATH_MAX);
+				p_cfilename = cfilename;
+				if (r == (size_t)-1 || r >= PATH_MAX)
+					p_cfilename = "<decoding error>";
+			}
 			sts = PyRun_AnyFileExFlags(
 				fp,
-				filename == NULL ? "<stdin>" : filename,
+				p_cfilename,
 				filename != NULL, &cf) != 0;
 		}
 		
@@ -589,7 +629,7 @@ Py_Main(int argc, char **argv)
    This is rare, but it is needed by the secureware extension. */
 
 void
-Py_GetArgcArgv(int *argc, char ***argv)
+Py_GetArgcArgv(int *argc, wchar_t ***argv)
 {
 	*argc = orig_argc;
 	*argv = orig_argv;

Modules/python.c

@@ -1,14 +1,27 @@
 /* Minimal main program -- everything is loaded from the library */
 
 #include "Python.h"
+#include <locale.h>
 
 #ifdef __FreeBSD__
 #include <floatingpoint.h>
 #endif
 
+#ifdef MS_WINDOWS
+int
+wmain(int argc, wchar_t **argv)
+{
+	return Py_Main(argc, argv);
+}
+#else
 int
 main(int argc, char **argv)
 {
+	wchar_t **argv_copy = PyMem_Malloc(sizeof(wchar_t*)*argc);
+	/* We need a second copies, as Python might modify the first one. */
+	wchar_t **argv_copy2 = PyMem_Malloc(sizeof(wchar_t*)*argc);
+	int i, res;
+	char *oldloc;
 	/* 754 requires that FP exceptions run in "no stop" mode by default,
 	 * and until C vendors implement C99's ways to control FP exceptions,
 	 * Python requires non-stop mode.  Alas, some platforms enable FP
@@ -20,5 +33,33 @@ main(int argc, char **argv)
 	m = fpgetmask();
 	fpsetmask(m & ~FP_X_OFL);
 #endif
-	return Py_Main(argc, argv);
+	if (!argv_copy || !argv_copy2) {
+		fprintf(stderr, "out of memory");
+		return 1;
+	}
+	oldloc = setlocale(LC_ALL, NULL);
+	setlocale(LC_ALL, "");
+	for (i = 0; i < argc; i++) {
+		size_t argsize = mbstowcs(NULL, argv[i], 0);
+		if (argsize == (size_t)-1) {
+			fprintf(stderr, "Could not convert argument %d to string", i);
+			return 1;
+		}
+		argv_copy[i] = PyMem_Malloc((argsize+1)*sizeof(wchar_t));
+		argv_copy2[i] = argv_copy[i];
+		if (!argv_copy[i]) {
+			fprintf(stderr, "out of memory");
+			return 1;
+		}
+		mbstowcs(argv_copy[i], argv[i], argsize+1);
+	}
+	setlocale(LC_ALL, oldloc);
+	res = Py_Main(argc, argv_copy);
+	for (i = 0; i < argc; i++) {
+		PyMem_Free(argv_copy2[i]);
+	}
+	PyMem_Free(argv_copy);
+	PyMem_Free(argv_copy2);
+	return res;
 }
+#endif

Objects/unicodeobject.c

@@ -553,10 +553,16 @@ PyObject *PyUnicode_FromWideChar(register const wchar_t *w,
     PyUnicodeObject *unicode;
 
     if (w == NULL) {
+        if (size == 0)
+            return PyUnicode_FromStringAndSize(NULL, 0);
 	PyErr_BadInternalCall();
 	return NULL;
     }
 
+    if (size == -1) {
+        size = wcslen(w);
+    }
+
     unicode = _PyUnicode_New(size);
     if (!unicode)
         return NULL;

PC/WinMain.c

@@ -12,5 +12,5 @@ int WINAPI WinMain(
     int nCmdShow              /* show state of window */
 )
 {
-    return Py_Main(__argc, __argv);
+    return Py_Main(__argc, __wargv);
 }

PC/getpathp.c

@@ -56,10 +56,10 @@
 
 #include "Python.h"
 #include "osdefs.h"
+#include <wchar.h>
 
 #ifdef MS_WINDOWS
 #include <windows.h>
-#include <tchar.h>
 #endif
 
 #ifdef HAVE_SYS_TYPES_H
@@ -82,17 +82,17 @@
  */
 
 #ifndef LANDMARK
-#define LANDMARK "lib\\os.py"
+#define LANDMARK L"lib\\os.py"
 #endif
 
-static char prefix[MAXPATHLEN+1];
+static wchar_t prefix[MAXPATHLEN+1];
-static char progpath[MAXPATHLEN+1];
+static wchar_t progpath[MAXPATHLEN+1];
-static char dllpath[MAXPATHLEN+1];
+static wchar_t dllpath[MAXPATHLEN+1];
-static char *module_search_path = NULL;
+static wchar_t *module_search_path = NULL;
 
 
 static int
-is_sep(char ch)	/* determine if "ch" is a separator character */
+is_sep(wchar_t ch)	/* determine if "ch" is a separator character */
 {
 #ifdef ALTSEP
 	return ch == SEP || ch == ALTSEP;
@@ -105,9 +105,9 @@ is_sep(char ch)	/* determine if "ch" is a separator character */
    beyond existing terminator.
 */
 static void
-reduce(char *dir)
+reduce(wchar_t *dir)
 {
-	size_t i = strlen(dir);
+	size_t i = wcslen(dir);
 	while (i > 0 && !is_sep(dir[i]))
 		--i;
 	dir[i] = '\0';
@@ -115,24 +115,24 @@ reduce(char *dir)
 	
 
 static int
-exists(char *filename)
+exists(wchar_t *filename)
 {
-	struct stat buf;
+	struct _stat64 buf;
-	return stat(filename, &buf) == 0;
+	return _wstat64(filename, &buf) == 0;
 }
 
 /* Assumes 'filename' MAXPATHLEN+1 bytes long - 
    may extend 'filename' by one character.
 */
 static int
-ismodule(char *filename)	/* Is module -- check for .pyc/.pyo too */
+ismodule(wchar_t *filename)	/* Is module -- check for .pyc/.pyo too */
 {
 	if (exists(filename))
 		return 1;
 
 	/* Check for the compiled version of prefix. */
-	if (strlen(filename) < MAXPATHLEN) {
+	if (wcslen(filename) < MAXPATHLEN) {
-		strcat(filename, Py_OptimizeFlag ? "o" : "c");
+		wcscat(filename, Py_OptimizeFlag ? L"o" : L"c");
 		if (exists(filename))
 			return 1;
 	}
@@ -149,22 +149,22 @@ ismodule(char *filename)	/* Is module -- check for .pyc/.pyo too */
    stuff as fits will be appended.
 */
 static void
-join(char *buffer, char *stuff)
+join(wchar_t *buffer, wchar_t *stuff)
 {
 	size_t n, k;
 	if (is_sep(stuff[0]))
 		n = 0;
 	else {
-		n = strlen(buffer);
+		n = wcslen(buffer);
 		if (n > 0 && !is_sep(buffer[n-1]) && n < MAXPATHLEN)
 			buffer[n++] = SEP;
 	}
 	if (n > MAXPATHLEN)
 		Py_FatalError("buffer overflow in getpathp.c's joinpath()");
-	k = strlen(stuff);
+	k = wcslen(stuff);
 	if (n + k > MAXPATHLEN)
 		k = MAXPATHLEN - n;
-	strncpy(buffer+n, stuff, k);
+	wcsncpy(buffer+n, stuff, k);
 	buffer[n+k] = '\0';
 }
 
@@ -173,12 +173,12 @@ join(char *buffer, char *stuff)
    'landmark' can not overflow prefix if too long.
 */
 static int
-gotlandmark(char *landmark)
+gotlandmark(wchar_t *landmark)
 {
 	int ok;
 	Py_ssize_t n;
 
-	n = strlen(prefix);
+	n = wcslen(prefix);
 	join(prefix, landmark);
 	ok = ismodule(prefix);
 	prefix[n] = '\0';
@@ -188,10 +188,10 @@ gotlandmark(char *landmark)
 /* assumes argv0_path is MAXPATHLEN+1 bytes long, already \0 term'd. 
    assumption provided by only caller, calculate_path() */
 static int
-search_for_prefix(char *argv0_path, char *landmark)
+search_for_prefix(wchar_t *argv0_path, wchar_t *landmark)
 {
 	/* Search from argv0_path, until landmark is found */
-	strcpy(prefix, argv0_path);
+	wcscpy(prefix, argv0_path);
 	do {
 		if (gotlandmark(landmark))
 			return 1;
@@ -219,39 +219,39 @@ extern const char *PyWin_DLLVersionString;
    in advance.  It could be simplied now Win16/Win32s is dead!
 */
 
-static char *
+static wchar_t *
 getpythonregpath(HKEY keyBase, int skipcore)
 {
 	HKEY newKey = 0;
 	DWORD dataSize = 0;
 	DWORD numKeys = 0;
 	LONG rc;
-	char *retval = NULL;
+	wchar_t *retval = NULL;
-	TCHAR *dataBuf = NULL;
+	WCHAR *dataBuf = NULL;
-	static const TCHAR keyPrefix[] = _T("Software\\Python\\PythonCore\\");
+	static const WCHAR keyPrefix[] = L"Software\\Python\\PythonCore\\";
-	static const TCHAR keySuffix[] = _T("\\PythonPath");
+	static const WCHAR keySuffix[] = L"\\PythonPath";
 	size_t versionLen;
 	DWORD index;
-	TCHAR *keyBuf = NULL;
+	WCHAR *keyBuf = NULL;
-	TCHAR *keyBufPtr;
+	WCHAR *keyBufPtr;
-	TCHAR **ppPaths = NULL;
+	WCHAR **ppPaths = NULL;
-
+	
 	/* Tried to use sysget("winver") but here is too early :-( */
-	versionLen = _tcslen(PyWin_DLLVersionString);
+	versionLen = strlen(PyWin_DLLVersionString);
 	/* Space for all the chars, plus one \0 */
 	keyBuf = keyBufPtr = malloc(sizeof(keyPrefix) + 
-		                    sizeof(TCHAR)*(versionLen-1) + 
+		                    sizeof(WCHAR)*(versionLen-1) + 
 				    sizeof(keySuffix));
 	if (keyBuf==NULL) goto done;
 
-	memcpy(keyBufPtr, keyPrefix, sizeof(keyPrefix)-sizeof(TCHAR));
+	memcpy(keyBufPtr, keyPrefix, sizeof(keyPrefix)-sizeof(WCHAR));
-	keyBufPtr += sizeof(keyPrefix)/sizeof(TCHAR) - 1;
+	keyBufPtr += sizeof(keyPrefix)/sizeof(WCHAR) - 1;
-	memcpy(keyBufPtr, PyWin_DLLVersionString, versionLen * sizeof(TCHAR));
+	mbstowcs(keyBufPtr, PyWin_DLLVersionString, versionLen);
 	keyBufPtr += versionLen;
 	/* NULL comes with this one! */
 	memcpy(keyBufPtr, keySuffix, sizeof(keySuffix));
 	/* Open the root Python key */
-	rc=RegOpenKeyEx(keyBase,
+	rc=RegOpenKeyExW(keyBase,
 	                keyBuf, /* subkey */
 	                0, /* reserved */
 	                KEY_READ,
@@ -265,31 +265,31 @@ getpythonregpath(HKEY keyBase, int skipcore)
 	/* Allocate a temp array of char buffers, so we only need to loop 
 	   reading the registry once
 	*/
-	ppPaths = malloc( sizeof(TCHAR *) * numKeys );
+	ppPaths = malloc( sizeof(WCHAR *) * numKeys );
 	if (ppPaths==NULL) goto done;
-	memset(ppPaths, 0, sizeof(TCHAR *) * numKeys);
+	memset(ppPaths, 0, sizeof(WCHAR *) * numKeys);
 	/* Loop over all subkeys, allocating a temp sub-buffer. */
 	for(index=0;index<numKeys;index++) {
-		TCHAR keyBuf[MAX_PATH+1];
+		WCHAR keyBuf[MAX_PATH+1];
 		HKEY subKey = 0;
 		DWORD reqdSize = MAX_PATH+1;
 		/* Get the sub-key name */
-		DWORD rc = RegEnumKeyEx(newKey, index, keyBuf, &reqdSize,
+		DWORD rc = RegEnumKeyExW(newKey, index, keyBuf, &reqdSize,
-		                        NULL, NULL, NULL, NULL );
+					 NULL, NULL, NULL, NULL );
 		if (rc!=ERROR_SUCCESS) goto done;
 		/* Open the sub-key */
-		rc=RegOpenKeyEx(newKey,
+		rc=RegOpenKeyExW(newKey,
 						keyBuf, /* subkey */
 						0, /* reserved */
 						KEY_READ,
 						&subKey);
 		if (rc!=ERROR_SUCCESS) goto done;
 		/* Find the value of the buffer size, malloc, then read it */
-		RegQueryValueEx(subKey, NULL, 0, NULL, NULL, &reqdSize);
+		RegQueryValueExW(subKey, NULL, 0, NULL, NULL, &reqdSize);
 		if (reqdSize) {
 			ppPaths[index] = malloc(reqdSize);
 			if (ppPaths[index]) {
-				RegQueryValueEx(subKey, NULL, 0, NULL, 
+				RegQueryValueExW(subKey, NULL, 0, NULL, 
 				                (LPBYTE)ppPaths[index], 
 				                &reqdSize);
 				dataSize += reqdSize + 1; /* 1 for the ";" */
@@ -302,19 +302,19 @@ getpythonregpath(HKEY keyBase, int skipcore)
 	if (dataSize == 0) goto done;
 
 	/* original datasize from RegQueryInfo doesn't include the \0 */
-	dataBuf = malloc((dataSize+1) * sizeof(TCHAR));
+	dataBuf = malloc((dataSize+1) * sizeof(WCHAR));
 	if (dataBuf) {
-		TCHAR *szCur = dataBuf;
+		WCHAR *szCur = dataBuf;
 		DWORD reqdSize = dataSize;
 		/* Copy our collected strings */
 		for (index=0;index<numKeys;index++) {
 			if (index > 0) {
-				*(szCur++) = _T(';');
+				*(szCur++) = L';';
 				dataSize--;
 			}
 			if (ppPaths[index]) {
-				Py_ssize_t len = _tcslen(ppPaths[index]);
+				Py_ssize_t len = wcslen(ppPaths[index]);
-				_tcsncpy(szCur, ppPaths[index], len);
+				wcsncpy(szCur, ppPaths[index], len);
 				szCur += len;
 				assert(dataSize > (DWORD)len);
 				dataSize -= (DWORD)len;
@@ -325,30 +325,17 @@ getpythonregpath(HKEY keyBase, int skipcore)
 		else {
 			/* If we have no values, we dont need a ';' */
 			if (numKeys) {
-				*(szCur++) = _T(';');
+				*(szCur++) = L';';
 				dataSize--;
 			}
 			/* Now append the core path entries - 
 			   this will include the NULL 
 			*/
-			rc = RegQueryValueEx(newKey, NULL, 0, NULL, 
+			rc = RegQueryValueExW(newKey, NULL, 0, NULL, 
-			                     (LPBYTE)szCur, &dataSize);
+					      (LPBYTE)szCur, &dataSize);
 		}
-		/* And set the result - caller must free 
+		/* And set the result - caller must free */
-		   If MBCS, it is fine as is.  If Unicode, allocate new
-		   buffer and convert.
-		*/
-#ifdef UNICODE
-		retval = (char *)malloc(reqdSize+1);
-		if (retval)
-			WideCharToMultiByte(CP_ACP, 0, 
-					dataBuf, -1, /* source */ 
-					retval, reqdSize+1, /* dest */
-					NULL, NULL);
-		free(dataBuf);
-#else
 		retval = dataBuf;
-#endif
 	}
 done:
 	/* Loop freeing my temp buffers */
@@ -368,45 +355,21 @@ done:
 static void
 get_progpath(void)
 {
-	extern char *Py_GetProgramName(void);
+	extern wchar_t *Py_GetProgramName(void);
-	char *path = getenv("PATH");
+	wchar_t *path = _wgetenv(L"PATH");
-	char *prog = Py_GetProgramName();
+	wchar_t *prog = Py_GetProgramName();
 
 #ifdef MS_WINDOWS
 	extern HANDLE PyWin_DLLhModule;
-#ifdef UNICODE
-	WCHAR wprogpath[MAXPATHLEN+1];
-	/* Windows documents that GetModuleFileName() will "truncate",
-	   but makes no mention of the null terminator.  Play it safe.
-	   PLUS Windows itself defines MAX_PATH as the same, but anyway...
-	*/
-	wprogpath[MAXPATHLEN]=_T('\0');
-	if (PyWin_DLLhModule &&
-	    GetModuleFileName(PyWin_DLLhModule, wprogpath, MAXPATHLEN)) {
-		WideCharToMultiByte(CP_ACP, 0, 
-		                    wprogpath, -1, 
-		                    dllpath, MAXPATHLEN+1, 
-		                    NULL, NULL);
-	}
-	wprogpath[MAXPATHLEN]=_T('\0');
-	if (GetModuleFileName(NULL, wprogpath, MAXPATHLEN)) {
-		WideCharToMultiByte(CP_ACP, 0, 
-		                    wprogpath, -1, 
-		                    progpath, MAXPATHLEN+1, 
-		                    NULL, NULL);
-		return;
-	}
-#else
 	/* static init of progpath ensures final char remains \0 */
 	if (PyWin_DLLhModule)
-		if (!GetModuleFileName(PyWin_DLLhModule, dllpath, MAXPATHLEN))
+		if (!GetModuleFileNameW(PyWin_DLLhModule, dllpath, MAXPATHLEN))
 			dllpath[0] = 0;
-	if (GetModuleFileName(NULL, progpath, MAXPATHLEN))
+	if (GetModuleFileNameW(NULL, progpath, MAXPATHLEN))
 		return;
-#endif
 #endif
 	if (prog == NULL || *prog == '\0')
-		prog = "python";
+		prog = L"python";
 
 	/* If there is no slash in the argv0 path, then we have to
 	 * assume python is on the user's $PATH, since there's no
@@ -414,24 +377,24 @@ get_progpath(void)
 	 * $PATH isn't exported, you lose.
 	 */
 #ifdef ALTSEP
-	if (strchr(prog, SEP) || strchr(prog, ALTSEP))
+	if (wcschr(prog, SEP) || wcschr(prog, ALTSEP))
 #else
-	if (strchr(prog, SEP))
+	if (wcschr(prog, SEP))
 #endif
-		strncpy(progpath, prog, MAXPATHLEN);
+		wcsncpy(progpath, prog, MAXPATHLEN);
 	else if (path) {
 		while (1) {
-			char *delim = strchr(path, DELIM);
+			wchar_t *delim = wcschr(path, DELIM);
 
 			if (delim) {
 				size_t len = delim - path;
 				/* ensure we can't overwrite buffer */
 				len = min(MAXPATHLEN,len);
-				strncpy(progpath, path, len);
+				wcsncpy(progpath, path, len);
 				*(progpath + len) = '\0';
 			}
 			else
-				strncpy(progpath, path, MAXPATHLEN);
+				wcsncpy(progpath, path, MAXPATHLEN);
 
 			/* join() is safe for MAXPATHLEN+1 size buffer */
 			join(progpath, prog);
@@ -452,23 +415,31 @@ get_progpath(void)
 static void
 calculate_path(void)
 {
-	char argv0_path[MAXPATHLEN+1];
+	wchar_t argv0_path[MAXPATHLEN+1];
-	char *buf;
+	wchar_t *buf;
 	size_t bufsz;
-	char *pythonhome = Py_GetPythonHome();
+	wchar_t *pythonhome = Py_GetPythonHome();
-	char *envpath = Py_GETENV("PYTHONPATH");
+	char *_envpath = Py_GETENV("PYTHONPATH");
+	wchar_t wenvpath[MAXPATHLEN+1];
+	wchar_t *envpath = NULL;
 
 #ifdef MS_WINDOWS
 	int skiphome, skipdefault;
-	char *machinepath = NULL;
+	wchar_t *machinepath = NULL;
-	char *userpath = NULL;
+	wchar_t *userpath = NULL;
-	char zip_path[MAXPATHLEN+1];
+	wchar_t zip_path[MAXPATHLEN+1];
 	size_t len;
 #endif
+	if (_envpath) {
+		size_t r = mbstowcs(wenvpath, _envpath, MAXPATHLEN+1);
+		envpath = wenvpath;
+		if (r == (size_t)-1 || r >= MAXPATHLEN)
+			envpath = NULL;
+	}
 
 	get_progpath();
 	/* progpath guaranteed \0 terminated in MAXPATH+1 bytes. */
-	strcpy(argv0_path, progpath);
+	wcscpy(argv0_path, progpath);
 	reduce(argv0_path);
 	if (pythonhome == NULL || *pythonhome == '\0') {
 		if (search_for_prefix(argv0_path, LANDMARK))
@@ -477,7 +448,7 @@ calculate_path(void)
 			pythonhome = NULL;
 	}
 	else
-		strncpy(prefix, pythonhome, MAXPATHLEN);
+		wcsncpy(prefix, pythonhome, MAXPATHLEN);
 
 	if (envpath && *envpath == '\0')
 		envpath = NULL;
@@ -486,11 +457,11 @@ calculate_path(void)
 #ifdef MS_WINDOWS
 	/* Calculate zip archive path */
 	if (dllpath[0])		/* use name of python DLL */
-		strncpy(zip_path, dllpath, MAXPATHLEN);
+		wcsncpy(zip_path, dllpath, MAXPATHLEN);
 	else			/* use name of executable program */
-		strncpy(zip_path, progpath, MAXPATHLEN);
+		wcsncpy(zip_path, progpath, MAXPATHLEN);
 	zip_path[MAXPATHLEN] = '\0';
-	len = strlen(zip_path);
+	len = wcslen(zip_path);
 	if (len > 4) {
 		zip_path[len-3] = 'z';	/* change ending to "zip" */
 		zip_path[len-2] = 'i';
@@ -524,29 +495,29 @@ calculate_path(void)
 
 	/* Calculate size of return buffer */
 	if (pythonhome != NULL) {
-		char *p;
+		wchar_t *p;
 		bufsz = 1;	
 		for (p = PYTHONPATH; *p; p++) {
 			if (*p == DELIM)
 				bufsz++; /* number of DELIM plus one */
 		}
-		bufsz *= strlen(pythonhome);
+		bufsz *= wcslen(pythonhome);
 	}
 	else
 		bufsz = 0;
-	bufsz += strlen(PYTHONPATH) + 1;
+	bufsz += wcslen(PYTHONPATH) + 1;
-	bufsz += strlen(argv0_path) + 1;
+	bufsz += wcslen(argv0_path) + 1;
 #ifdef MS_WINDOWS
 	if (userpath)
-		bufsz += strlen(userpath) + 1;
+		bufsz += wcslen(userpath) + 1;
 	if (machinepath)
-		bufsz += strlen(machinepath) + 1;
+		bufsz += wcslen(machinepath) + 1;
-	bufsz += strlen(zip_path) + 1;
+	bufsz += wcslen(zip_path) + 1;
 #endif
 	if (envpath != NULL)
-		bufsz += strlen(envpath) + 1;
+		bufsz += wcslen(envpath) + 1;
 
-	module_search_path = buf = malloc(bufsz);
+	module_search_path = buf = malloc(bufsz*sizeof(wchar_t));
 	if (buf == NULL) {
 		/* We can't exit, so print a warning and limp along */
 		fprintf(stderr, "Can't malloc dynamic PYTHONPATH.\n");
@@ -568,57 +539,57 @@ calculate_path(void)
 	}
 
 	if (envpath) {
-		strcpy(buf, envpath);
+		wcscpy(buf, envpath);
-		buf = strchr(buf, '\0');
+		buf = wcschr(buf, L'\0');
 		*buf++ = DELIM;
 	}
 #ifdef MS_WINDOWS
 	if (zip_path[0]) {
-		strcpy(buf, zip_path);
+		wcscpy(buf, zip_path);
-		buf = strchr(buf, '\0');
+		buf = wcschr(buf, L'\0');
 		*buf++ = DELIM;
 	}
 	if (userpath) {
-		strcpy(buf, userpath);
+		wcscpy(buf, userpath);
-		buf = strchr(buf, '\0');
+		buf = wcschr(buf, L'\0');
 		*buf++ = DELIM;
 		free(userpath);
 	}
 	if (machinepath) {
-		strcpy(buf, machinepath);
+		wcscpy(buf, machinepath);
-		buf = strchr(buf, '\0');
+		buf = wcschr(buf, L'\0');
 		*buf++ = DELIM;
 		free(machinepath);
 	}
 	if (pythonhome == NULL) {
 		if (!skipdefault) {
-			strcpy(buf, PYTHONPATH);
+			wcscpy(buf, PYTHONPATH);
-			buf = strchr(buf, '\0');
+			buf = wcschr(buf, L'\0');
 		}
 	}
 #else
 	if (pythonhome == NULL) {
-		strcpy(buf, PYTHONPATH);
+		wcscpy(buf, PYTHONPATH);
-		buf = strchr(buf, '\0');
+		buf = wcschr(buf, L'\0');
 	}
 #endif /* MS_WINDOWS */
 	else {
-		char *p = PYTHONPATH;
+		wchar_t *p = PYTHONPATH;
-		char *q;
+		wchar_t *q;
 		size_t n;
 		for (;;) {
-			q = strchr(p, DELIM);
+			q = wcschr(p, DELIM);
 			if (q == NULL)
-				n = strlen(p);
+				n = wcslen(p);
 			else
 				n = q-p;
 			if (p[0] == '.' && is_sep(p[1])) {
-				strcpy(buf, pythonhome);
+				wcscpy(buf, pythonhome);
-				buf = strchr(buf, '\0');
+				buf = wcschr(buf, L'\0');
 				p++;
 				n--;
 			}
-			strncpy(buf, p, n);
+			wcsncpy(buf, p, n);
 			buf += n;
 			if (q == NULL)
 				break;
@@ -628,10 +599,10 @@ calculate_path(void)
 	}
 	if (argv0_path) {
 		*buf++ = DELIM;
-		strcpy(buf, argv0_path);
+		wcscpy(buf, argv0_path);
-		buf = strchr(buf, '\0');
+		buf = wcschr(buf, L'\0');
 	}
-	*buf = '\0';
+	*buf = L'\0';
 	/* Now to pull one last hack/trick.  If sys.prefix is
 	   empty, then try and find it somewhere on the paths
 	   we calculated.  We scan backwards, as our general policy
@@ -640,12 +611,12 @@ calculate_path(void)
 	   on the path, and that our 'prefix' directory is
 	   the parent of that.
 	*/
-	if (*prefix=='\0') {
+	if (*prefix==L'\0') {
-		char lookBuf[MAXPATHLEN+1];
+		wchar_t lookBuf[MAXPATHLEN+1];
-		char *look = buf - 1; /* 'buf' is at the end of the buffer */
+		wchar_t *look = buf - 1; /* 'buf' is at the end of the buffer */
 		while (1) {
 			Py_ssize_t nchars;
-			char *lookEnd = look;
+			wchar_t *lookEnd = look;
 			/* 'look' will end up one character before the
 			   start of the path in question - even if this
 			   is one character before the start of the buffer
@@ -653,8 +624,8 @@ calculate_path(void)
 			while (look >= module_search_path && *look != DELIM)
 				look--;
 			nchars = lookEnd-look;
-			strncpy(lookBuf, look+1, nchars);
+			wcsncpy(lookBuf, look+1, nchars);
-			lookBuf[nchars] = '\0';
+			lookBuf[nchars] = L'\0';
 			/* Up one level to the parent */
 			reduce(lookBuf);
 			if (search_for_prefix(lookBuf, LANDMARK)) {
@@ -671,7 +642,7 @@ calculate_path(void)
 
 /* External interface */
 
-char *
+wchar_t *
 Py_GetPath(void)
 {
 	if (!module_search_path)
@@ -679,7 +650,7 @@ Py_GetPath(void)
 	return module_search_path;
 }
 
-char *
+wchar_t *
 Py_GetPrefix(void)
 {
 	if (!module_search_path)
@@ -687,13 +658,13 @@ Py_GetPrefix(void)
 	return prefix;
 }
 
-char *
+wchar_t *
 Py_GetExecPrefix(void)
 {
 	return Py_GetPrefix();
 }
 
-char *
+wchar_t *
 Py_GetProgramFullPath(void)
 {
 	if (!module_search_path)

PC/pyconfig.h

@@ -80,7 +80,7 @@ WIN32 is still required for the locale module.
 #define MS_WIN32 /* only support win32 and greater. */
 #define MS_WINDOWS
 #ifndef PYTHONPATH
-#	define PYTHONPATH ".\\DLLs;.\\lib;.\\lib\\plat-win;.\\lib\\lib-tk"
+#	define PYTHONPATH L".\\DLLs;.\\lib;.\\lib\\plat-win;.\\lib\\lib-tk"
 #endif
 #define NT_THREADS
 #define WITH_THREAD

Python/frozenmain.c

@@ -2,6 +2,7 @@
 /* Python interpreter main program for frozen scripts */
 
 #include "Python.h"
+#include <locale.h>
 
 #ifdef MS_WINDOWS
 extern void PyWinFreeze_ExeInit(void);
@@ -15,9 +16,13 @@ int
 Py_FrozenMain(int argc, char **argv)
 {
 	char *p;
-	int n, sts;
+	int i, n, sts;
 	int inspect = 0;
 	int unbuffered = 0;
+	char *oldloc;
+	wchar_t **argv_copy = PyMem_Malloc(sizeof(wchar_t*)*argc);
+	/* We need a second copies, as Python might modify the first one. */
+	wchar_t **argv_copy2 = PyMem_Malloc(sizeof(wchar_t*)*argc);
 
 	Py_FrozenFlag = 1; /* Suppress errors from getpath.c */
 
@@ -32,10 +37,33 @@ Py_FrozenMain(int argc, char **argv)
 		setbuf(stderr, (char *)NULL);
 	}
 
+	if (!argv_copy) {
+		fprintf(stderr, "out of memory");
+		return 1;
+	}
+
+	oldloc = setlocale(LC_ALL, NULL);
+	setlocale(LC_ALL, "");
+	for (i = 0; i < argc; i++) {
+		size_t argsize = mbstowcs(NULL, argv[i], 0);
+		if (argsize == (size_t)-1) {
+			fprintf(stderr, "Could not convert argument %d to string", i);
+			return 1;
+		}
+		argv_copy[i] = PyMem_Malloc((argsize+1)*sizeof(wchar_t));
+		argv_copy2[i] = argv_copy[i];
+		if (!argv_copy[i]) {
+			fprintf(stderr, "out of memory");
+			return 1;
+		}
+		mbstowcs(argv_copy[i], argv[i], argsize+1);
+	}
+	setlocale(LC_ALL, oldloc);
+
 #ifdef MS_WINDOWS
 	PyInitFrozenExtensions();
 #endif /* MS_WINDOWS */
-	Py_SetProgramName(argv[0]);
+	Py_SetProgramName(argv_copy[0]);
 	Py_Initialize();
 #ifdef MS_WINDOWS
 	PyWinFreeze_ExeInit();
@@ -45,7 +73,7 @@ Py_FrozenMain(int argc, char **argv)
 		fprintf(stderr, "Python %s\n%s\n",
 			Py_GetVersion(), Py_GetCopyright());
 
-	PySys_SetArgv(argc, argv);
+	PySys_SetArgv(argc, argv_copy);
 
 	n = PyImport_ImportFrozenModule("__main__");
 	if (n == 0)
@@ -64,5 +92,10 @@ Py_FrozenMain(int argc, char **argv)
 	PyWinFreeze_ExeTerm();
 #endif
 	Py_Finalize();
+	for (i = 0; i < argc; i++) {
+		PyMem_Free(argv_copy2[i]);
+	}
+	PyMem_Free(argv_copy);
+	PyMem_Free(argv_copy2);
 	return sts;
 }

Python/getopt.c

@@ -27,8 +27,11 @@
 /* Modified to support --help and --version, as well as /? on Windows
  * by Georg Brandl. */
 
+#include <Python.h>
 #include <stdio.h>
 #include <string.h>
+#include <wchar.h>
+#include <pygetopt.h>
 
 #ifdef __cplusplus
 extern "C" {
@@ -36,40 +39,40 @@ extern "C" {
 
 int _PyOS_opterr = 1;          /* generate error messages */
 int _PyOS_optind = 1;          /* index into argv array   */
-char *_PyOS_optarg = NULL;     /* optional argument       */
+wchar_t *_PyOS_optarg = NULL;     /* optional argument       */
 
-int _PyOS_GetOpt(int argc, char **argv, char *optstring)
+int _PyOS_GetOpt(int argc, wchar_t **argv, wchar_t *optstring)
 {
-	static char *opt_ptr = "";
+	static wchar_t *opt_ptr = L"";
-	char *ptr;
+	wchar_t *ptr;
-	int option;
+	wchar_t option;
 
 	if (*opt_ptr == '\0') {
 
 		if (_PyOS_optind >= argc)
 			return -1;
 #ifdef MS_WINDOWS
-		else if (strcmp(argv[_PyOS_optind], "/?") == 0) {
+		else if (wcscmp(argv[_PyOS_optind], L"/?") == 0) {
 			++_PyOS_optind;
 			return 'h';
 		}
 #endif
 
-		else if (argv[_PyOS_optind][0] != '-' ||
+		else if (argv[_PyOS_optind][0] != L'-' ||
-		         argv[_PyOS_optind][1] == '\0' /* lone dash */ )
+		         argv[_PyOS_optind][1] == L'\0' /* lone dash */ )
 			return -1;
 
-		else if (strcmp(argv[_PyOS_optind], "--") == 0) {
+		else if (wcscmp(argv[_PyOS_optind], L"--") == 0) {
 			++_PyOS_optind;
 			return -1;
 		}
 
-		else if (strcmp(argv[_PyOS_optind], "--help") == 0) {
+		else if (wcscmp(argv[_PyOS_optind], L"--help") == 0) {
 			++_PyOS_optind;
 			return 'h';
 		}
 
-		else if (strcmp(argv[_PyOS_optind], "--version") == 0) {
+		else if (wcscmp(argv[_PyOS_optind], L"--version") == 0) {
 			++_PyOS_optind;
 			return 'V';
 		}
@@ -78,27 +81,27 @@ int _PyOS_GetOpt(int argc, char **argv, char *optstring)
 		opt_ptr = &argv[_PyOS_optind++][1]; 
 	}
 
-	if ( (option = *opt_ptr++) == '\0')
+	if ( (option = *opt_ptr++) == L'\0')
 		return -1;
 	
-	if ((ptr = strchr(optstring, option)) == NULL) {
+	if ((ptr = wcschr(optstring, option)) == NULL) {
 		if (_PyOS_opterr)
-			fprintf(stderr, "Unknown option: -%c\n", option);
+		  fprintf(stderr, "Unknown option: -%c\n", (char)option);
 
 		return '_';
 	}
 
-	if (*(ptr + 1) == ':') {
+	if (*(ptr + 1) == L':') {
-		if (*opt_ptr != '\0') {
+		if (*opt_ptr != L'\0') {
 			_PyOS_optarg  = opt_ptr;
-			opt_ptr = "";
+			opt_ptr = L"";
 		}
 
 		else {
 			if (_PyOS_optind >= argc) {
 				if (_PyOS_opterr)
 					fprintf(stderr,
-			    "Argument expected for the -%c option\n", option);
+						"Argument expected for the -%c option\n", (char)option);
 				return '_';
 			}
 

Python/pythonrun.c

@@ -17,6 +17,7 @@
 #include "ast.h"
 #include "eval.h"
 #include "marshal.h"
+#include "osdefs.h"
 
 #ifdef HAVE_SIGNAL_H
 #include <signal.h>
@@ -30,6 +31,7 @@
 #ifdef MS_WINDOWS
 #undef BYTE
 #include "windows.h"
+#define PATH_MAX MAXPATHLEN
 #endif
 
 #ifndef Py_REF_DEBUG
@@ -44,7 +46,7 @@
 extern "C" {
 #endif
 
-extern char *Py_GetPath(void);
+extern wchar_t *Py_GetPath(void);
 
 extern grammar _PyParser_Grammar; /* From graminit.c */
 
@@ -646,35 +648,43 @@ Py_EndInterpreter(PyThreadState *tstate)
 	PyInterpreterState_Delete(interp);
 }
 
-static char *progname = "python";
+static wchar_t *progname = L"python";
 
 void
-Py_SetProgramName(char *pn)
+Py_SetProgramName(wchar_t *pn)
 {
 	if (pn && *pn)
 		progname = pn;
 }
 
-char *
+wchar_t *
 Py_GetProgramName(void)
 {
 	return progname;
 }
 
-static char *default_home = NULL;
+static wchar_t *default_home = NULL;
+static wchar_t env_home[PATH_MAX+1];
 
 void
-Py_SetPythonHome(char *home)
+Py_SetPythonHome(wchar_t *home)
 {
 	default_home = home;
 }
 
-char *
+wchar_t *
 Py_GetPythonHome(void)
 {
-	char *home = default_home;
+	wchar_t *home = default_home;
-	if (home == NULL && !Py_IgnoreEnvironmentFlag)
+	if (home == NULL && !Py_IgnoreEnvironmentFlag) {
-		home = Py_GETENV("PYTHONHOME");
+		char* chome = Py_GETENV("PYTHONHOME");
+		if (chome) {
+			size_t r = mbstowcs(env_home, chome, PATH_MAX+1);
+			if (r != (size_t)-1 && r <= PATH_MAX)
+				home = env_home;
+		}
+
+	}
 	return home;
 }
 

Python/sysmodule.c

@@ -882,7 +882,7 @@ PySys_ResetWarnOptions(void)
 }
 
 void
-PySys_AddWarnOption(const char *s)
+PySys_AddWarnOption(const wchar_t *s)
 {
 	PyObject *str;
 
@@ -892,7 +892,7 @@ PySys_AddWarnOption(const char *s)
 		if (warnoptions == NULL)
 			return;
 	}
-	str = PyUnicode_FromString(s);
+	str = PyUnicode_FromWideChar(s, -1);
 	if (str != NULL) {
 		PyList_Append(warnoptions, str);
 		Py_DECREF(str);
@@ -1222,12 +1222,12 @@ _PySys_Init(void)
 	SET_SYS_FROM_STRING("platform",
 			    PyUnicode_FromString(Py_GetPlatform()));
 	SET_SYS_FROM_STRING("executable",
-			    PyUnicode_DecodeFSDefault(
+			    PyUnicode_FromWideChar(
-				Py_GetProgramFullPath()));
+				   Py_GetProgramFullPath(), -1));
 	SET_SYS_FROM_STRING("prefix",
-			    PyUnicode_DecodeFSDefault(Py_GetPrefix()));
+			    PyUnicode_FromWideChar(Py_GetPrefix(), -1));
 	SET_SYS_FROM_STRING("exec_prefix",
-		   	    PyUnicode_DecodeFSDefault(Py_GetExecPrefix()));
+		   	    PyUnicode_FromWideChar(Py_GetExecPrefix(), -1));
 	SET_SYS_FROM_STRING("maxsize",
 			    PyLong_FromSsize_t(PY_SSIZE_T_MAX));
 	SET_SYS_FROM_STRING("float_info",
@@ -1280,15 +1280,15 @@ _PySys_Init(void)
 }
 
 static PyObject *
-makepathobject(const char *path, int delim)
+makepathobject(const wchar_t *path, wchar_t delim)
 {
 	int i, n;
-	const char *p;
+	const wchar_t *p;
 	PyObject *v, *w;
 
 	n = 1;
 	p = path;
-	while ((p = strchr(p, delim)) != NULL) {
+	while ((p = wcschr(p, delim)) != NULL) {
 		n++;
 		p++;
 	}
@@ -1296,10 +1296,10 @@ makepathobject(const char *path, int delim)
 	if (v == NULL)
 		return NULL;
 	for (i = 0; ; i++) {
-		p = strchr(path, delim);
+		p = wcschr(path, delim);
 		if (p == NULL)
-			p = strchr(path, '\0'); /* End of string */
+			p = wcschr(path, L'\0'); /* End of string */
-		w = PyUnicode_DecodeFSDefaultAndSize(path, (Py_ssize_t) (p - path));
+		w = PyUnicode_FromWideChar(path, (Py_ssize_t)(p - path));
 		if (w == NULL) {
 			Py_DECREF(v);
 			return NULL;
@@ -1313,7 +1313,7 @@ makepathobject(const char *path, int delim)
 }
 
 void
-PySys_SetPath(const char *path)
+PySys_SetPath(const wchar_t *path)
 {
 	PyObject *v;
 	if ((v = makepathobject(path, DELIM)) == NULL)
@@ -1324,12 +1324,12 @@ PySys_SetPath(const char *path)
 }
 
 static PyObject *
-makeargvobject(int argc, char **argv)
+makeargvobject(int argc, wchar_t **argv)
 {
 	PyObject *av;
 	if (argc <= 0 || argv == NULL) {
 		/* Ensure at least one (empty) argument is seen */
-		static char *empty_argv[1] = {""};
+		static wchar_t *empty_argv[1] = {L""};
 		argv = empty_argv;
 		argc = 1;
 	}
@@ -1351,7 +1351,7 @@ makeargvobject(int argc, char **argv)
 			} else
 				v = PyUnicode_FromString(argv[i]);
 #else
-			PyObject *v = PyUnicode_FromString(argv[i]);
+			PyObject *v = PyUnicode_FromWideChar(argv[i], -1);
 #endif
 			if (v == NULL) {
 				Py_DECREF(av);
@@ -1364,13 +1364,38 @@ makeargvobject(int argc, char **argv)
 	return av;
 }
 
+#ifdef HAVE_REALPATH
+static wchar_t*
+_wrealpath(const wchar_t *path, wchar_t *resolved_path)
+{
+	char cpath[PATH_MAX];
+	char cresolved_path[PATH_MAX];
+	char *res;
+	size_t r;
+	r = wcstombs(cpath, path, PATH_MAX);
+	if (r == (size_t)-1 || r >= PATH_MAX) {
+		errno = EINVAL;
+		return NULL;
+	}
+	res = realpath(cpath, cresolved_path);
+	if (res == NULL)
+		return NULL;
+	r = mbstowcs(resolved_path, cresolved_path, PATH_MAX);
+	if (r == (size_t)-1 || r >= PATH_MAX) {
+		errno = EINVAL;
+		return NULL;
+	}
+	return resolved_path;
+}
+#endif
+
 void
-PySys_SetArgv(int argc, char **argv)
+PySys_SetArgv(int argc, wchar_t **argv)
 {
 #if defined(HAVE_REALPATH)
-	char fullpath[MAXPATHLEN];
+	wchar_t fullpath[MAXPATHLEN];
 #elif defined(MS_WINDOWS)
-	char fullpath[MAX_PATH];
+	wchar_t fullpath[MAX_PATH];
 #endif
 	PyObject *av = makeargvobject(argc, argv);
 	PyObject *path = PySys_GetObject("path");
@@ -1379,53 +1404,54 @@ PySys_SetArgv(int argc, char **argv)
 	if (PySys_SetObject("argv", av) != 0)
 		Py_FatalError("can't assign sys.argv");
 	if (path != NULL) {
-		char *argv0 = argv[0];
+		wchar_t *argv0 = argv[0];
-		char *p = NULL;
+		wchar_t *p = NULL;
 		Py_ssize_t n = 0;
 		PyObject *a;
+		extern int _Py_wreadlink(const wchar_t *, wchar_t *, size_t);
 #ifdef HAVE_READLINK
-		char link[MAXPATHLEN+1];
+		wchar_t link[MAXPATHLEN+1];
-		char argv0copy[2*MAXPATHLEN+1];
+		wchar_t argv0copy[2*MAXPATHLEN+1];
 		int nr = 0;
-		if (argc > 0 && argv0 != NULL && strcmp(argv0, "-c") != 0)
+		if (argc > 0 && argv0 != NULL && wcscmp(argv0, L"-c") != 0)
-			nr = readlink(argv0, link, MAXPATHLEN);
+			nr = _Py_wreadlink(argv0, link, MAXPATHLEN);
 		if (nr > 0) {
 			/* It's a symlink */
 			link[nr] = '\0';
 			if (link[0] == SEP)
 				argv0 = link; /* Link to absolute path */
-			else if (strchr(link, SEP) == NULL)
+			else if (wcschr(link, SEP) == NULL)
 				; /* Link without path */
 			else {
 				/* Must join(dirname(argv0), link) */
-				char *q = strrchr(argv0, SEP);
+				wchar_t *q = wcsrchr(argv0, SEP);
 				if (q == NULL)
 					argv0 = link; /* argv0 without path */
 				else {
 					/* Must make a copy */
-					strcpy(argv0copy, argv0);
+					wcscpy(argv0copy, argv0);
-					q = strrchr(argv0copy, SEP);
+					q = wcsrchr(argv0copy, SEP);
-					strcpy(q+1, link);
+					wcscpy(q+1, link);
 					argv0 = argv0copy;
 				}
 			}
 		}
 #endif /* HAVE_READLINK */
 #if SEP == '\\' /* Special case for MS filename syntax */
-		if (argc > 0 && argv0 != NULL && strcmp(argv0, "-c") != 0) {
+		if (argc > 0 && argv0 != NULL && wcscmp(argv0, L"-c") != 0) {
-			char *q;
+			wchar_t *q;
 #ifdef MS_WINDOWS
-			char *ptemp;
+			wchar_t *ptemp;
-			if (GetFullPathName(argv0,
+			if (GetFullPathNameW(argv0,
-					   sizeof(fullpath),
+					   sizeof(fullpath)/sizeof(fullpath[0]),
 					   fullpath,
 					   &ptemp)) {
 				argv0 = fullpath;
 			}
 #endif
-			p = strrchr(argv0, SEP);
+			p = wcsrchr(argv0, SEP);
 			/* Test for alternate separator */
-			q = strrchr(p ? p : argv0, '/');
+			q = wcsrchr(p ? p : argv0, '/');
 			if (q != NULL)
 				p = q;
 			if (p != NULL) {
@@ -1435,13 +1461,13 @@ PySys_SetArgv(int argc, char **argv)
 			}
 		}
 #else /* All other filename syntaxes */
-		if (argc > 0 && argv0 != NULL && strcmp(argv0, "-c") != 0) {
+		if (argc > 0 && argv0 != NULL && wcscmp(argv0, L"-c") != 0) {
 #if defined(HAVE_REALPATH)
-			if (realpath(argv0, fullpath)) {
+			if (_wrealpath(argv0, fullpath)) {
 				argv0 = fullpath;
 			}
 #endif
-			p = strrchr(argv0, SEP);
+			p = wcsrchr(argv0, SEP);
 		}
 		if (p != NULL) {
 			n = p + 1 - argv0;
@@ -1451,7 +1477,7 @@ PySys_SetArgv(int argc, char **argv)
 #endif /* Unix */
 		}
 #endif /* All others */
-		a = PyUnicode_FromStringAndSize(argv0, n);
+		a = PyUnicode_FromWideChar(argv0, n);
 		if (a == NULL)
 			Py_FatalError("no mem for sys.path insertion");
 		if (PyList_Insert(path, 0, a) < 0)