mingw: prefer unix sep if MSYSTEM environment variable

Co-authored-by: Алексей <alexey.pawlow@gmail.com>
Co-authored-by: Christoph Reiter <reiter.christoph@gmail.com>
Co-authored-by: cat <cat@wolfgirl.org>
Co-authored-by: Naveen M K <naveen521kk@gmail.com>

Include/pylifecycle.h

@@ -21,6 +21,12 @@ PyAPI_FUNC(int) Py_IsInitialized(void);
 PyAPI_FUNC(PyThreadState *) Py_NewInterpreter(void);
 PyAPI_FUNC(void) Py_EndInterpreter(PyThreadState *);
 
+PyAPI_FUNC(wchar_t) Py_GetSepW(const wchar_t *);
+PyAPI_FUNC(char) Py_GetSepA(const char *);
+
+PyAPI_FUNC(void) Py_NormalizeSepsW(wchar_t *);
+PyAPI_FUNC(void) Py_NormalizeSepsA(char *);
+
 
 /* Py_PyAtExit is for the atexit module, Py_AtExit is for low-level
  * exit functions.

Lib/ntpath.py

@@ -11,9 +11,7 @@
 curdir = '.'
 pardir = '..'
 extsep = '.'
-sep = '\\'
 pathsep = ';'
-altsep = '/'
 defpath = '.;C:\\bin'
 devnull = 'nul'
 
@@ -23,6 +21,14 @@
 import genericpath
 from genericpath import *
 
+if sys.platform == "win32" and "MSYSTEM" in os.environ:
+    sep = '/'
+    altsep = '\\'
+else:
+    sep = '\\'
+    altsep = '/'
+bsep = str.encode(sep)
+baltsep = str.encode(altsep)
 
 __all__ = ["normcase","isabs","join","splitdrive","split","splitext",
            "basename","dirname","commonprefix","getsize","getmtime",
@@ -34,9 +40,33 @@
 
 def _get_bothseps(path):
     if isinstance(path, bytes):
-        return b'\\/'
+        return bsep+baltsep
+    else:
+        return sep+altsep
+
+def _get_sep(path):
+    if isinstance(path, bytes):
+        return bsep
     else:
-        return '\\/'
+        return sep
+
+def _get_altsep(path):
+    if isinstance(path, bytes):
+        return baltsep
+    else:
+        return altsep
+
+def _get_colon(path):
+    if isinstance(path, bytes):
+        return b':'
+    else:
+        return ':'
+
+def _get_unc_prefix(path):
+    if isinstance(path, bytes):
+        return b'\\\\?\\UNC\\'
+    else:
+        return '\\\\?\\UNC\\'
 
 # Normalize the case of a pathname and map slashes to backslashes.
 # Other normalizations (such as optimizing '../' away) are not done
@@ -58,14 +88,14 @@ def normcase(s):
             return s
         if isinstance(s, bytes):
             encoding = sys.getfilesystemencoding()
-            s = s.decode(encoding, 'surrogateescape').replace('/', '\\')
+            s = s.decode(encoding, 'surrogateescape').replace(altsep, sep)
             s = _LCMapStringEx(_LOCALE_NAME_INVARIANT,
                                _LCMAP_LOWERCASE, s)
             return s.encode(encoding, 'surrogateescape')
         else:
             return _LCMapStringEx(_LOCALE_NAME_INVARIANT,
                                   _LCMAP_LOWERCASE,
-                                  s.replace('/', '\\'))
+                                  s.replace(altsep, sep))
 except ImportError:
     def normcase(s):
         """Normalize case of pathname.
@@ -74,8 +104,8 @@ def normcase(s):
         """
         s = os.fspath(s)
         if isinstance(s, bytes):
-            return os.fsencode(os.fsdecode(s).replace('/', '\\').lower())
-        return s.replace('/', '\\').lower()
+            return os.fsencode(os.fsdecode(s).replace(altsep, sep).lower())
+        return s.replace(altsep, sep).lower()
 
 
 # Return whether a path is absolute.
@@ -87,14 +117,9 @@ def normcase(s):
 def isabs(s):
     """Test whether a path is absolute"""
     s = os.fspath(s)
-    if isinstance(s, bytes):
-        sep = b'\\'
-        altsep = b'/'
-        colon_sep = b':\\'
-    else:
-        sep = '\\'
-        altsep = '/'
-        colon_sep = ':\\'
+    sep = _get_sep(s)
+    altsep = _get_altsep(s)
+    colon_sep = _get_colon(s) + sep
     s = s[:3].replace(altsep, sep)
     # Absolute: UNC, device, and paths with a drive and root.
     # LEGACY BUG: isabs("/x") should be false since the path has no drive.
@@ -106,14 +131,9 @@ def isabs(s):
 # Join two (or more) paths.
 def join(path, *paths):
     path = os.fspath(path)
-    if isinstance(path, bytes):
-        sep = b'\\'
-        seps = b'\\/'
-        colon = b':'
-    else:
-        sep = '\\'
-        seps = '\\/'
-        colon = ':'
+    sep = _get_sep(path)
+    seps = _get_bothseps(path)
+    colon = _get_colon(path)
     try:
         if not paths:
             path[:0] + sep  #23780: Ensure compatible data type even if p is null.
@@ -172,16 +192,10 @@ def splitdrive(p):
     """
     p = os.fspath(p)
     if len(p) >= 2:
-        if isinstance(p, bytes):
-            sep = b'\\'
-            altsep = b'/'
-            colon = b':'
-            unc_prefix = b'\\\\?\\UNC\\'
-        else:
-            sep = '\\'
-            altsep = '/'
-            colon = ':'
-            unc_prefix = '\\\\?\\UNC\\'
+        sep = _get_sep(p)
+        altsep = _get_altsep(p)
+        colon = _get_colon(p)
+        unc_prefix = _get_unc_prefix(p)
         normp = p.replace(altsep, sep)
         if normp[0:2] == sep * 2:
             # UNC drives, e.g. \\server\share or \\?\UNC\server\share
@@ -231,9 +245,9 @@ def split(p):
 def splitext(p):
     p = os.fspath(p)
     if isinstance(p, bytes):
-        return genericpath._splitext(p, b'\\', b'/', b'.')
+        return genericpath._splitext(p, bsep, baltsep, b'.')
     else:
-        return genericpath._splitext(p, '\\', '/', '.')
+        return genericpath._splitext(p, sep, altsep, '.')
 splitext.__doc__ = genericpath._splitext.__doc__
 
 
@@ -496,14 +510,12 @@ def expandvars(path):
     def normpath(path):
         """Normalize path, eliminating double slashes, etc."""
         path = os.fspath(path)
+        sep = _get_sep(path)
+        altsep = _get_altsep(path)
         if isinstance(path, bytes):
-            sep = b'\\'
-            altsep = b'/'
             curdir = b'.'
             pardir = b'..'
         else:
-            sep = '\\'
-            altsep = '/'
             curdir = '.'
             pardir = '..'
         path = path.replace(altsep, sep)
@@ -719,6 +731,7 @@ def realpath(path, *, strict=False):
                 # strip the prefix anyway.
                 if ex.winerror == initial_winerror:
                     path = spath
+        path = normpath(path)
         return path
 
 
@@ -729,12 +742,11 @@ def realpath(path, *, strict=False):
 def relpath(path, start=None):
     """Return a relative version of a path"""
     path = os.fspath(path)
+    sep = _get_sep(path)
     if isinstance(path, bytes):
-        sep = b'\\'
         curdir = b'.'
         pardir = b'..'
     else:
-        sep = '\\'
         curdir = '.'
         pardir = '..'
 
@@ -789,13 +801,11 @@ def commonpath(paths):
         raise ValueError('commonpath() arg is an empty sequence')
 
     paths = tuple(map(os.fspath, paths))
+    sep = _get_sep(paths[0])
+    altsep = _get_altsep(paths[0])
     if isinstance(paths[0], bytes):
-        sep = b'\\'
-        altsep = b'/'
         curdir = b'.'
     else:
-        sep = '\\'
-        altsep = '/'
         curdir = '.'
 
     try:

Modules/posixmodule.c

@@ -3835,6 +3835,7 @@ posix_getcwd(int use_bytes)
         return PyErr_SetFromWindowsErr(0);
     }
 
+    Py_NormalizeSepsW(wbuf2);
     PyObject *resobj = PyUnicode_FromWideChar(wbuf2, len);
     if (wbuf2 != wbuf) {
         PyMem_RawFree(wbuf2);
@@ -4439,6 +4440,7 @@ os__getfinalpathname_impl(PyObject *module, path_t *path)
         target_path = tmp;
     }
 
+    Py_NormalizeSepsW(target_path);
     result = PyUnicode_FromWideChar(target_path, result_length);
     if (result && path->narrow) {
         Py_SETREF(result, PyUnicode_EncodeFSDefault(result));

Python/initconfig.c

@@ -176,7 +176,7 @@ static const char usage_envvars[] =
 "PYTHONVERBOSE           : trace import statements (-v)\n"
 "PYTHONWARNINGS=arg      : warning control (-W arg)\n";
 
-#if defined(MS_WINDOWS)
+#if defined(_MSC_VER)
 #  define PYTHONHOMEHELP "<prefix>\\python{major}{minor}"
 #else
 #  define PYTHONHOMEHELP "<prefix>/lib/pythonX.X"

Python/pathconfig.c

@@ -18,6 +18,140 @@
 extern "C" {
 #endif
 
+#ifdef __MINGW32__
+#define wcstok  wcstok_s
+#include <windows.h>
+#endif
+
+static int
+Py_StartsWithA(const char * str, const char * prefix)
+{
+    while(*prefix)
+    {
+        if(*prefix++ != *str++)
+            return 0;
+    }
+
+    return 1;
+}
+
+static int
+Py_StartsWithW(const wchar_t * str, const wchar_t * prefix)
+{
+    while(*prefix)
+    {
+        if(*prefix++ != *str++)
+            return 0;
+    }
+
+    return 1;
+}
+
+char
+Py_GetSepA(const char *name)
+{
+    char* msystem = (char*)2; /* So that non Windows use / as sep */
+    static char sep = '\0';
+#ifdef _WIN32
+    /* https://msdn.microsoft.com/en-gb/library/windows/desktop/aa365247%28v=vs.85%29.aspx
+     * The "\\?\" prefix .. indicate that the path should be passed to the system with minimal
+     * modification, which means that you cannot use forward slashes to represent path separators
+     */
+    if (name != NULL && Py_StartsWithA(name, "\\\\?\\") != 0)
+    {
+        return '\\';
+    }
+#endif
+    if (sep != '\0')
+        return sep;
+#if defined(__MINGW32__)
+    msystem = Py_GETENV("MSYSTEM");
+#endif
+    if (msystem != NULL)
+        sep = '/';
+    else
+        sep = '\\';
+    return sep;
+}
+
+static char
+Py_GetAltSepA(const char *name)
+{
+    char sep = Py_GetSepA(name);
+    if (sep == '/')
+        return '\\';
+    return '/';
+}
+
+void
+Py_NormalizeSepsA(char *name)
+{
+    assert(name != NULL);
+    char sep = Py_GetSepA(name);
+    char altsep = Py_GetAltSepA(name);
+    char* seps;
+    if (name[0] != '\0' && name[1] == ':') {
+        name[0] = toupper(name[0]);
+    }
+    seps = strchr(name, altsep);
+    while(seps) {
+        *seps = sep;
+        seps = strchr(seps, altsep);
+    }
+}
+
+wchar_t
+Py_GetSepW(const wchar_t *name)
+{
+    char* msystem = (char*)2; /* So that non Windows use / as sep */
+    static wchar_t sep = L'\0';
+#ifdef _WIN32
+    /* https://msdn.microsoft.com/en-gb/library/windows/desktop/aa365247%28v=vs.85%29.aspx
+     * The "\\?\" prefix .. indicate that the path should be passed to the system with minimal
+     * modification, which means that you cannot use forward slashes to represent path separators
+     */
+    if (name != NULL && Py_StartsWithW(name, L"\\\\?\\") != 0)
+    {
+        return L'\\';
+    }
+#endif
+    if (sep != L'\0')
+        return sep;
+#if defined(__MINGW32__)
+    msystem = Py_GETENV("MSYSTEM");
+#endif
+    if (msystem != NULL)
+        sep = L'/';
+    else
+        sep = L'\\';
+    return sep;
+}
+
+static wchar_t
+Py_GetAltSepW(const wchar_t *name)
+{
+    char sep = Py_GetSepW(name);
+    if (sep == L'/')
+        return L'\\';
+    return L'/';
+}
+
+void
+Py_NormalizeSepsW(wchar_t *name)
+{
+    assert(name != NULL);
+    wchar_t sep = Py_GetSepW(name);
+    wchar_t altsep = Py_GetAltSepW(name);
+    wchar_t* seps;
+    if (name[0] != L'\0' && name[1] == L':') {
+        name[0] = towupper(name[0]);
+    }
+    seps = wcschr(name, altsep);
+    while(seps) {
+        *seps = sep;
+        seps = wcschr(seps, altsep);
+    }
+}
 
 /* External interface */
 
@@ -317,6 +451,7 @@ _Py_SetProgramFullPath(const wchar_t *program_full_path)
     if (has_value && _Py_path_config.program_full_path == NULL) {
         path_out_of_memory(__func__);
     }
+    Py_NormalizeSepsW(_Py_path_config.program_name);
 }