cpython-lru_cache-weakref: Python/fileutils.c annotate

annotate Python/fileutils.c @ 81258:5bb289e4fb35 3.3

(Merge 3.2) Issue #16455: On FreeBSD and Solaris, if the locale is C, the ASCII/surrogateescape codec is now used, instead of the locale encoding, to decode the command line arguments. This change fixes inconsistencies with os.fsencode() and os.fsdecode() because these operating systems announces an ASCII locale encoding, whereas the ISO-8859-1 encoding is used in practice.

author	Victor Stinner <victor.stinner@gmail.com>
date	Thu, 03 Jan 2013 01:21:07 +0100
parents	2a67d79490ae c256764e2b3f
children	8bbd215940f1 4bf19545c016

rev	line source
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	1 #include "Python.h"
80406 5db6d9ddf6e8 Issue #15835: Define PATH_MAX on HP-UX. Stefan Krah <skrah@bytereef.org> parents: 75353 diff changeset	2 #include "osdefs.h"
65164 1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	3 #ifdef MS_WINDOWS
1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	4 # include <windows.h>
1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	5 #endif
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	6
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	7 #ifdef HAVE_LANGINFO_H
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	8 #include <locale.h>
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	9 #include <langinfo.h>
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	10 #endif
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	11
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	12 #ifdef __APPLE__
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	13 extern wchar_t* _Py_DecodeUTF8_surrogateescape(const char *s, Py_ssize_t size);
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	14 #endif
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	15
75353 c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	16 PyObject *
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	17 _Py_device_encoding(int fd)
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	18 {
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	19 #if defined(MS_WINDOWS) \|\| defined(MS_WIN64)
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	20 UINT cp;
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	21 #endif
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	22 if (!_PyVerify_fd(fd) \|\| !isatty(fd)) {
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	23 Py_RETURN_NONE;
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	24 }
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	25 #if defined(MS_WINDOWS) \|\| defined(MS_WIN64)
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	26 if (fd == 0)
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	27 cp = GetConsoleCP();
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	28 else if (fd == 1 \|\| fd == 2)
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	29 cp = GetConsoleOutputCP();
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	30 else
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	31 cp = 0;
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	32 /* GetConsoleCP() and GetConsoleOutputCP() return 0 if the application
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	33 has no console */
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	34 if (cp != 0)
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	35 return PyUnicode_FromFormat("cp%u", (unsigned int)cp);
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	36 #elif defined(CODESET)
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	37 {
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	38 char *codeset = nl_langinfo(CODESET);
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	39 if (codeset != NULL && codeset[0] != 0)
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	40 return PyUnicode_FromString(codeset);
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	41 }
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	42 #endif
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	43 Py_RETURN_NONE;
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	44 }
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	45
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	46 #if !defined(__APPLE__) && !defined(MS_WINDOWS)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	47 extern int _Py_normalize_encoding(const char , char , size_t);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	48
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	49 /* Workaround FreeBSD and OpenIndiana locale encoding issue with the C locale.
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	50 On these operating systems, nl_langinfo(CODESET) announces an alias of the
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	51 ASCII encoding, whereas mbstowcs() and wcstombs() functions use the
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	52 ISO-8859-1 encoding. The problem is that os.fsencode() and os.fsdecode() use
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	53 locale.getpreferredencoding() codec. For example, if command line arguments
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	54 are decoded by mbstowcs() and encoded back by os.fsencode(), we get a
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	55 UnicodeEncodeError instead of retrieving the original byte string.
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	56
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	57 The workaround is enabled if setlocale(LC_CTYPE, NULL) returns "C",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	58 nl_langinfo(CODESET) announces "ascii" (or an alias to ASCII), and at least
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	59 one byte in range 0x80-0xff can be decoded from the locale encoding. The
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	60 workaround is also enabled on error, for example if getting the locale
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	61 failed.
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	62
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	63 Values of locale_is_ascii:
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	64
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	65 1: the workaround is used: _Py_wchar2char() uses
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	66 encode_ascii_surrogateescape() and _Py_char2wchar() uses
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	67 decode_ascii_surrogateescape()
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	68 0: the workaround is not used: _Py_wchar2char() uses wcstombs() and
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	69 _Py_char2wchar() uses mbstowcs()
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	70 -1: unknown, need to call check_force_ascii() to get the value
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	71 */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	72 static int force_ascii = -1;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	73
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	74 static int
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	75 check_force_ascii(void)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	76 {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	77 char *loc;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	78 #if defined(HAVE_LANGINFO_H) && defined(CODESET)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	79 char codeset, *alias;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	80 char encoding[100];
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	81 int is_ascii;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	82 unsigned int i;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	83 char* ascii_aliases[] = {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	84 "ascii",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	85 "646",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	86 "ansi-x3.4-1968",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	87 "ansi-x3-4-1968",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	88 "ansi-x3.4-1986",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	89 "cp367",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	90 "csascii",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	91 "ibm367",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	92 "iso646-us",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	93 "iso-646.irv-1991",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	94 "iso-ir-6",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	95 "us",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	96 "us-ascii",
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	97 NULL
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	98 };
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	99 #endif
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	100
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	101 loc = setlocale(LC_CTYPE, NULL);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	102 if (loc == NULL)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	103 goto error;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	104 if (strcmp(loc, "C") != 0) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	105 /* the LC_CTYPE locale is different than C */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	106 return 0;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	107 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	108
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	109 #if defined(HAVE_LANGINFO_H) && defined(CODESET)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	110 codeset = nl_langinfo(CODESET);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	111 if (!codeset \|\| codeset[0] == '\0') {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	112 /* CODESET is not set or empty */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	113 goto error;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	114 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	115 if (!_Py_normalize_encoding(codeset, encoding, sizeof(encoding)))
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	116 goto error;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	117
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	118 is_ascii = 0;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	119 for (alias=ascii_aliases; *alias != NULL; alias++) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	120 if (strcmp(encoding, *alias) == 0) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	121 is_ascii = 1;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	122 break;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	123 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	124 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	125 if (!is_ascii) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	126 /* nl_langinfo(CODESET) is not "ascii" or an alias of ASCII */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	127 return 0;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	128 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	129
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	130 for (i=0x80; i<0xff; i++) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	131 unsigned char ch;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	132 wchar_t wch;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	133 size_t res;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	134
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	135 ch = (unsigned char)i;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	136 res = mbstowcs(&wch, (char*)&ch, 1);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	137 if (res != (size_t)-1) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	138 /* decoding a non-ASCII character from the locale encoding succeed:
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	139 the locale encoding is not ASCII, force ASCII */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	140 return 1;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	141 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	142 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	143 /* None of the bytes in the range 0x80-0xff can be decoded from the locale
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	144 encoding: the locale encoding is really ASCII */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	145 return 0;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	146 #else
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	147 /* nl_langinfo(CODESET) is not available: always force ASCII */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	148 return 1;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	149 #endif
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	150
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	151 error:
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	152 /* if an error occured, force the ASCII encoding */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	153 return 1;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	154 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	155
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	156 static char*
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	157 encode_ascii_surrogateescape(const wchar_t text, size_t error_pos)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	158 {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	159 char result = NULL, out;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	160 size_t len, i;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	161 wchar_t ch;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	162
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	163 if (error_pos != NULL)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	164 *error_pos = (size_t)-1;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	165
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	166 len = wcslen(text);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	167
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	168 result = PyMem_Malloc(len + 1); /* +1 for NUL byte */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	169 if (result == NULL)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	170 return NULL;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	171
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	172 out = result;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	173 for (i=0; i<len; i++) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	174 ch = text[i];
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	175
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	176 if (ch <= 0x7f) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	177 /* ASCII character */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	178 *out++ = (char)ch;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	179 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	180 else if (0xdc80 <= ch && ch <= 0xdcff) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	181 /* UTF-8b surrogate */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	182 *out++ = (char)(ch - 0xdc00);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	183 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	184 else {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	185 if (error_pos != NULL)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	186 *error_pos = i;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	187 PyMem_Free(result);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	188 return NULL;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	189 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	190 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	191 *out = '\0';
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	192 return result;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	193 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	194 #endif /* !defined(__APPLE__) && !defined(MS_WINDOWS) */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	195
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	196 #if !defined(__APPLE__) && (!defined(MS_WINDOWS) \|\| !defined(HAVE_MBRTOWC))
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	197 static wchar_t*
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	198 decode_ascii_surrogateescape(const char arg, size_t size)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	199 {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	200 wchar_t *res;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	201 unsigned char *in;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	202 wchar_t *out;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	203
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	204 res = PyMem_Malloc((strlen(arg)+1)*sizeof(wchar_t));
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	205 if (!res)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	206 return NULL;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	207
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	208 in = (unsigned char*)arg;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	209 out = res;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	210 while(*in)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	211 if(*in < 128)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	212 out++ = in++;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	213 else
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	214 out++ = 0xdc00 + in++;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	215 *out = 0;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	216 if (size != NULL)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	217 *size = out - res;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	218 return res;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	219 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	220 #endif
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	221
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	222
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	223 /* Decode a byte string from the locale encoding with the
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	224 surrogateescape error handler (undecodable bytes are decoded as characters
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	225 in range U+DC80..U+DCFF). If a byte sequence can be decoded as a surrogate
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	226 character, escape the bytes using the surrogateescape error handler instead
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	227 of decoding them.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	228
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	229 Use _Py_wchar2char() to encode the character string back to a byte string.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	230
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	231 Return a pointer to a newly allocated wide character string (use
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	232 PyMem_Free() to free the memory) and write the number of written wide
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	233 characters excluding the null character into *size if size is not NULL, or
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	234 NULL on error (decoding or memory allocation error). If size is not NULL,
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	235 *size is set to (size_t)-1 on memory error and (size_t)-2 on decoding
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	236 error.
66094 76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	237
76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	238 Conversion errors should never happen, unless there is a bug in the C
76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	239 library. */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	240 wchar_t*
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	241 _Py_char2wchar(const char* arg, size_t *size)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	242 {
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	243 #ifdef __APPLE__
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	244 wchar_t *wstr;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	245 wstr = _Py_DecodeUTF8_surrogateescape(arg, strlen(arg));
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	246 if (size != NULL) {
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	247 if (wstr != NULL)
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	248 *size = wcslen(wstr);
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	249 else
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	250 *size = (size_t)-1;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	251 }
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	252 return wstr;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	253 #else
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	254 wchar_t *res;
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	255 size_t argsize;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	256 size_t count;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	257 unsigned char *in;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	258 wchar_t *out;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	259 #ifdef HAVE_MBRTOWC
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	260 mbstate_t mbs;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	261 #endif
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	262
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	263 #ifndef MS_WINDOWS
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	264 if (force_ascii == -1)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	265 force_ascii = check_force_ascii();
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	266
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	267 if (force_ascii) {
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	268 /* force ASCII encoding to workaround mbstowcs() issue */
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	269 res = decode_ascii_surrogateescape(arg, size);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	270 if (res == NULL)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	271 goto oom;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	272 return res;
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	273 }
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	274 #endif
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	275
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	276 #ifdef HAVE_BROKEN_MBSTOWCS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	277 /* Some platforms have a broken implementation of
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	278 * mbstowcs which does not count the characters that
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	279 * would result from conversion. Use an upper bound.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	280 */
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	281 argsize = strlen(arg);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	282 #else
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	283 argsize = mbstowcs(NULL, arg, 0);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	284 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	285 if (argsize != (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	286 res = (wchar_t )PyMem_Malloc((argsize+1)sizeof(wchar_t));
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	287 if (!res)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	288 goto oom;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	289 count = mbstowcs(res, arg, argsize+1);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	290 if (count != (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	291 wchar_t *tmp;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	292 /* Only use the result if it contains no
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	293 surrogate characters. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	294 for (tmp = res; *tmp != 0 &&
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	295 (tmp < 0xd800 \|\| tmp > 0xdfff); tmp++)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	296 ;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	297 if (*tmp == 0) {
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	298 if (size != NULL)
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	299 *size = count;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	300 return res;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	301 }
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	302 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	303 PyMem_Free(res);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	304 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	305 /* Conversion failed. Fall back to escaping with surrogateescape. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	306 #ifdef HAVE_MBRTOWC
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	307 /* Try conversion with mbrtwoc (C99), and escape non-decodable bytes. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	308
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	309 /* Overallocate; as multi-byte characters are in the argument, the
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	310 actual output could use less memory. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	311 argsize = strlen(arg) + 1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	312 res = (wchar_t)PyMem_Malloc(argsizesizeof(wchar_t));
66094 76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	313 if (!res)
76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	314 goto oom;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	315 in = (unsigned char*)arg;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	316 out = res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	317 memset(&mbs, 0, sizeof mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	318 while (argsize) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	319 size_t converted = mbrtowc(out, (char*)in, argsize, &mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	320 if (converted == 0)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	321 /* Reached end of string; null char stored. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	322 break;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	323 if (converted == (size_t)-2) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	324 /* Incomplete character. This should never happen,
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	325 since we provide everything that we have -
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	326 unless there is a bug in the C library, or I
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	327 misunderstood how mbrtowc works. */
66094 76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	328 PyMem_Free(res);
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	329 if (size != NULL)
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	330 *size = (size_t)-2;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	331 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	332 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	333 if (converted == (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	334 /* Conversion error. Escape as UTF-8b, and start over
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	335 in the initial shift state. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	336 out++ = 0xdc00 + in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	337 argsize--;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	338 memset(&mbs, 0, sizeof mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	339 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	340 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	341 if (out >= 0xd800 && out <= 0xdfff) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	342 /* Surrogate character. Escape the original
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	343 byte sequence with surrogateescape. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	344 argsize -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	345 while (converted--)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	346 out++ = 0xdc00 + in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	347 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	348 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	349 /* successfully converted some bytes */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	350 in += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	351 argsize -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	352 out++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	353 }
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	354 if (size != NULL)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	355 *size = out - res;
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	356 #else /* HAVE_MBRTOWC */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	357 /* Cannot use C locale for escaping; manually escape as if charset
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	358 is ASCII (i.e. escape all bytes > 128. This will still roundtrip
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	359 correctly in the locale's charset, which must be an ASCII superset. */
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	360 res = decode_ascii_surrogateescape(arg, size);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	361 if (res == NULL)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	362 goto oom;
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	363 #endif /* HAVE_MBRTOWC */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	364 return res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	365 oom:
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	366 if (size != NULL)
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	367 *size = (size_t)-1;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	368 return NULL;
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	369 #endif /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	370 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	371
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	372 /* Encode a (wide) character string to the locale encoding with the
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	373 surrogateescape error handler (characters in range U+DC80..U+DCFF are
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	374 converted to bytes 0x80..0xFF).
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	375
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	376 This function is the reverse of _Py_char2wchar().
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	377
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	378 Return a pointer to a newly allocated byte string (use PyMem_Free() to free
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	379 the memory), or NULL on encoding or memory allocation error.
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	380
1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	381 If error_pos is not NULL: *error_pos is the index of the invalid character
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	382 on encoding error, or (size_t)-1 otherwise. */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	383 char*
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	384 _Py_wchar2char(const wchar_t text, size_t error_pos)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	385 {
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	386 #ifdef __APPLE__
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	387 Py_ssize_t len;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	388 PyObject unicode, bytes = NULL;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	389 char *cpath;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	390
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	391 unicode = PyUnicode_FromWideChar(text, wcslen(text));
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	392 if (unicode == NULL)
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	393 return NULL;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	394
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	395 bytes = _PyUnicode_AsUTF8String(unicode, "surrogateescape");
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	396 Py_DECREF(unicode);
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	397 if (bytes == NULL) {
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	398 PyErr_Clear();
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	399 if (error_pos != NULL)
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	400 *error_pos = (size_t)-1;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	401 return NULL;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	402 }
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	403
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	404 len = PyBytes_GET_SIZE(bytes);
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	405 cpath = PyMem_Malloc(len+1);
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	406 if (cpath == NULL) {
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	407 PyErr_Clear();
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	408 Py_DECREF(bytes);
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	409 if (error_pos != NULL)
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	410 *error_pos = (size_t)-1;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	411 return NULL;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	412 }
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	413 memcpy(cpath, PyBytes_AsString(bytes), len + 1);
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	414 Py_DECREF(bytes);
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	415 return cpath;
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	416 #else /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	417 const size_t len = wcslen(text);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	418 char result = NULL, bytes = NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	419 size_t i, size, converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	420 wchar_t c, buf[2];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	421
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	422 #ifndef MS_WINDOWS
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	423 if (force_ascii == -1)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	424 force_ascii = check_force_ascii();
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	425
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	426 if (force_ascii)
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	427 return encode_ascii_surrogateescape(text, error_pos);
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	428 #endif
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	429
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	430 /* The function works in two steps:
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	431 1. compute the length of the output buffer in bytes (size)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	432 2. outputs the bytes */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	433 size = 0;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	434 buf[1] = 0;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	435 while (1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	436 for (i=0; i < len; i++) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	437 c = text[i];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	438 if (c >= 0xdc80 && c <= 0xdcff) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	439 /* UTF-8b surrogate */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	440 if (bytes != NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	441 *bytes++ = c - 0xdc00;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	442 size--;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	443 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	444 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	445 size++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	446 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	447 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	448 else {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	449 buf[0] = c;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	450 if (bytes != NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	451 converted = wcstombs(bytes, buf, size);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	452 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	453 converted = wcstombs(NULL, buf, 0);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	454 if (converted == (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	455 if (result != NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	456 PyMem_Free(result);
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	457 if (error_pos != NULL)
1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	458 *error_pos = i;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	459 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	460 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	461 if (bytes != NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	462 bytes += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	463 size -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	464 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	465 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	466 size += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	467 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	468 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	469 if (result != NULL) {
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	470 *bytes = '\0';
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	471 break;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	472 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	473
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	474 size += 1; /* nul byte at the end */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	475 result = PyMem_Malloc(size);
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	476 if (result == NULL) {
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	477 if (error_pos != NULL)
c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	478 *error_pos = (size_t)-1;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	479 return NULL;
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	480 }
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	481 bytes = result;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	482 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	483 return result;
80702 c838c9b117f1 Issue #16416: On Mac OS X, operating system data are now always Victor Stinner <victor.stinner@gmail.com> parents: 66094 diff changeset	484 #endif /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	485 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	486
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	487 /* In principle, this should use HAVE__WSTAT, and _wstat
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	488 should be detected by autoconf. However, no current
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	489 POSIX system provides that function, so testing for
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	490 it is pointless.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	491 Not sure whether the MS_WINDOWS guards are necessary:
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	492 perhaps for cygwin/mingw builds?
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	493 */
65164 1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	494 #if defined(HAVE_STAT) && !defined(MS_WINDOWS)
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	495
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	496 /* Get file status. Encode the path to the locale encoding. */
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	497
65164 1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	498 int
1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	499 _Py_wstat(const wchar_t* path, struct stat *buf)
1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	500 {
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	501 int err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	502 char *fname;
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	503 fname = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	504 if (fname == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	505 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	506 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	507 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	508 err = stat(fname, buf);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	509 PyMem_Free(fname);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	510 return err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	511 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	512 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	513
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	514 #ifdef HAVE_STAT
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	515
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	516 /* Call _wstat() on Windows, or encode the path to the filesystem encoding and
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	517 call stat() otherwise. Only fill st_mode attribute on Windows.
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	518
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	519 Return 0 on success, -1 on _wstat() / stat() error, -2 if an exception was
2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	520 raised. */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	521
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	522 int
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	523 _Py_stat(PyObject path, struct stat statbuf)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	524 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	525 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	526 int err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	527 struct _stat wstatbuf;
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	528 wchar_t *wpath;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	529
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	530 wpath = PyUnicode_AsUnicode(path);
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	531 if (wpath == NULL)
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	532 return -2;
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	533 err = _wstat(wpath, &wstatbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	534 if (!err)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	535 statbuf->st_mode = wstatbuf.st_mode;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	536 return err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	537 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	538 int ret;
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	539 PyObject *bytes = PyUnicode_EncodeFSDefault(path);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	540 if (bytes == NULL)
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	541 return -2;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	542 ret = stat(PyBytes_AS_STRING(bytes), statbuf);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	543 Py_DECREF(bytes);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	544 return ret;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	545 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	546 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	547
81257 c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	548 #endif
c256764e2b3f Issue #16455: On FreeBSD and Solaris, if the locale is C, the Victor Stinner <victor.stinner@gmail.com> parents: 80704 diff changeset	549
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	550 /* Open a file. Use _wfopen() on Windows, encode the path to the locale
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	551 encoding and use fopen() otherwise. */
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	552
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	553 FILE *
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	554 _Py_wfopen(const wchar_t path, const wchar_t mode)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	555 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	556 #ifndef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	557 FILE *f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	558 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	559 char cmode[10];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	560 size_t r;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	561 r = wcstombs(cmode, mode, 10);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	562 if (r == (size_t)-1 \|\| r >= 10) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	563 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	564 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	565 }
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	566 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	567 if (cpath == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	568 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	569 f = fopen(cpath, cmode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	570 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	571 return f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	572 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	573 return _wfopen(path, mode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	574 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	575 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	576
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	577 /* Call _wfopen() on Windows, or encode the path to the filesystem encoding and
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	578 call fopen() otherwise.
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	579
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	580 Return the new file object on success, or NULL if the file cannot be open or
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	581 (if PyErr_Occurred()) on unicode error */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	582
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	583 FILE*
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	584 _Py_fopen(PyObject path, const char mode)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	585 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	586 #ifdef MS_WINDOWS
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	587 wchar_t *wpath;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	588 wchar_t wmode[10];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	589 int usize;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	590
74129 33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	591 if (!PyUnicode_Check(path)) {
33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	592 PyErr_Format(PyExc_TypeError,
33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	593 "str file path expected under Windows, got %R",
33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	594 Py_TYPE(path));
33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	595 return NULL;
33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	596 }
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	597 wpath = PyUnicode_AsUnicode(path);
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	598 if (wpath == NULL)
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	599 return NULL;
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	600
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	601 usize = MultiByteToWideChar(CP_ACP, 0, mode, -1, wmode, sizeof(wmode));
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	602 if (usize == 0)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	603 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	604
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	605 return _wfopen(wpath, wmode);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	606 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	607 FILE *f;
74085 a882aa30a713 _Py_fopen now allows bytes filenames under non-Windows platforms. Antoine Pitrou <solipsis@pitrou.net> parents: 74058 diff changeset	608 PyObject *bytes;
a882aa30a713 _Py_fopen now allows bytes filenames under non-Windows platforms. Antoine Pitrou <solipsis@pitrou.net> parents: 74058 diff changeset	609 if (!PyUnicode_FSConverter(path, &bytes))
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	610 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	611 f = fopen(PyBytes_AS_STRING(bytes), mode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	612 Py_DECREF(bytes);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	613 return f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	614 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	615 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	616
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	617 #ifdef HAVE_READLINK
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	618
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	619 /* Read value of symbolic link. Encode the path to the locale encoding, decode
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	620 the result from the locale encoding. Return -1 on error. */
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	621
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	622 int
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	623 _Py_wreadlink(const wchar_t path, wchar_t buf, size_t bufsiz)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	624 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	625 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	626 char cbuf[PATH_MAX];
65409 e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65346 diff changeset	627 wchar_t *wbuf;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	628 int res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	629 size_t r1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	630
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	631 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	632 if (cpath == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	633 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	634 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	635 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	636 res = (int)readlink(cpath, cbuf, PATH_MAX);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	637 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	638 if (res == -1)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	639 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	640 if (res == PATH_MAX) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	641 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	642 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	643 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	644 cbuf[res] = '\0'; /* buf will be null terminated */
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	645 wbuf = _Py_char2wchar(cbuf, &r1);
65410 6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	646 if (wbuf == NULL) {
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	647 errno = EINVAL;
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	648 return -1;
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	649 }
65409 e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65346 diff changeset	650 if (bufsiz <= r1) {
e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65346 diff changeset	651 PyMem_Free(wbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	652 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	653 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	654 }
65409 e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65346 diff changeset	655 wcsncpy(buf, wbuf, bufsiz);
e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65346 diff changeset	656 PyMem_Free(wbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	657 return (int)r1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	658 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	659 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	660
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	661 #ifdef HAVE_REALPATH
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	662
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	663 /* Return the canonicalized absolute pathname. Encode path to the locale
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	664 encoding, decode the result from the locale encoding.
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	665 Return NULL on error. */
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	666
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	667 wchar_t*
65166 8fd5d06462e6 _Py_wrealpath() requires the size of the output buffer Victor Stinner <victor.stinner@haypocalc.com> parents: 65165 diff changeset	668 _Py_wrealpath(const wchar_t *path,
8fd5d06462e6 _Py_wrealpath() requires the size of the output buffer Victor Stinner <victor.stinner@haypocalc.com> parents: 65165 diff changeset	669 wchar_t *resolved_path, size_t resolved_path_size)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	670 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	671 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	672 char cresolved_path[PATH_MAX];
65411 df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	673 wchar_t *wresolved_path;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	674 char *res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	675 size_t r;
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	676 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	677 if (cpath == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	678 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	679 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	680 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	681 res = realpath(cpath, cresolved_path);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	682 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	683 if (res == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	684 return NULL;
65411 df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	685
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	686 wresolved_path = _Py_char2wchar(cresolved_path, &r);
65411 df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	687 if (wresolved_path == NULL) {
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	688 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	689 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	690 }
65411 df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	691 if (resolved_path_size <= r) {
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	692 PyMem_Free(wresolved_path);
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	693 errno = EINVAL;
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	694 return NULL;
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	695 }
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	696 wcsncpy(resolved_path, wresolved_path, resolved_path_size);
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	697 PyMem_Free(wresolved_path);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	698 return resolved_path;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	699 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	700 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	701
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	702 /* Get the current directory. size is the buffer size in wide characters
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	703 including the null character. Decode the path from the locale encoding.
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	704 Return NULL on error. */
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	705
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	706 wchar_t*
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	707 _Py_wgetcwd(wchar_t *buf, size_t size)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	708 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	709 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	710 return _wgetcwd(buf, size);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	711 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	712 char fname[PATH_MAX];
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	713 wchar_t *wname;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	714 size_t len;
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	715
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	716 if (getcwd(fname, PATH_MAX) == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	717 return NULL;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	718 wname = _Py_char2wchar(fname, &len);
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	719 if (wname == NULL)
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	720 return NULL;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	721 if (size <= len) {
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	722 PyMem_Free(wname);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	723 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	724 }
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	725 wcsncpy(buf, wname, size);
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	726 PyMem_Free(wname);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	727 return buf;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	728 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	729 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	730

Mercurial > cpython-lru_cache-weakref

annotate Python/fileutils.c @ 81258:5bb289e4fb35 3.3