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

annotate Python/fileutils.c @ 80411:48fbdaf3a849

Issue #16416: OS data are now always encoded/decoded to/from UTF-8/surrogateescape, instead of the locale encoding (which may be ASCII if no locale environment variable is set), to avoid inconsistencies with os.fsencode() and os.fsdecode() functions which are already using UTF-8/surrogateescape.

author	Victor Stinner <victor.stinner@gmail.com>
date	Mon, 12 Nov 2012 23:04:02 +0100
parents	3857a63cf840
children	f3e512b5ffb3

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
75353 c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	7 #ifdef HAVE_LANGINFO_H
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	8 #include <langinfo.h>
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	9 #endif
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	10
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	11 #ifdef __APPLE__
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	12 extern wchar_t* _Py_DecodeUTF8_surrogateescape(const char *s, Py_ssize_t size);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	13 #endif
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	14
75353 c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	15 PyObject *
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	16 _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	17 {
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	18 #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	19 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	20 #endif
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	21 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	22 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	23 }
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	24 #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	25 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	26 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	27 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	28 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	29 else
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	30 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	31 /* 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	32 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	33 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	34 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	35 #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	36 {
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	37 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	38 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	39 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	40 }
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	41 #endif
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	42 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	43 }
c80d9a0bd5a0 Issue #14153 Create _Py_device_encoding() to prevent _io from having to import Brett Cannon <brett@python.org> parents: 74129 diff changeset	44
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	45 #ifdef HAVE_STAT
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	46
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	47 /* Decode a byte string from the locale encoding with the
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	48 surrogateescape error handler (undecodable bytes are decoded as characters
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	49 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	50 character, escape the bytes using the surrogateescape error handler instead
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	51 of decoding them.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	52
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	53 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	54
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	55 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	56 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	57 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	58 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	59 *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	60 error.
66094 76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	61
76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	62 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	63 library. */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	64 wchar_t*
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	65 _Py_char2wchar(const char* arg, size_t *size)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	66 {
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	67 #ifdef __APPLE__
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	68 wchar_t *wstr;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	69 wstr = _Py_DecodeUTF8_surrogateescape(arg, strlen(arg));
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	70 if (wstr == NULL)
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	71 return NULL;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	72 if (size != NULL)
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	73 *size = wcslen(wstr);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	74 return wstr;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	75 #else
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	76 wchar_t *res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	77 #ifdef HAVE_BROKEN_MBSTOWCS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	78 /* Some platforms have a broken implementation of
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	79 * mbstowcs which does not count the characters that
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	80 * would result from conversion. Use an upper bound.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	81 */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	82 size_t argsize = strlen(arg);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	83 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	84 size_t argsize = mbstowcs(NULL, arg, 0);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	85 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	86 size_t count;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	87 unsigned char *in;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	88 wchar_t *out;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	89 #ifdef HAVE_MBRTOWC
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	90 mbstate_t mbs;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	91 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	92 if (argsize != (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	93 res = (wchar_t )PyMem_Malloc((argsize+1)sizeof(wchar_t));
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	94 if (!res)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	95 goto oom;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	96 count = mbstowcs(res, arg, argsize+1);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	97 if (count != (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	98 wchar_t *tmp;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	99 /* Only use the result if it contains no
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	100 surrogate characters. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	101 for (tmp = res; *tmp != 0 &&
80057 13dd8199c112 Issue #16330: Use surrogate-related macros Victor Stinner <victor.stinner@gmail.com> parents: 75353 diff changeset	102 !Py_UNICODE_IS_SURROGATE(*tmp); tmp++)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	103 ;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	104 if (*tmp == 0) {
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	105 if (size != NULL)
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	106 *size = count;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	107 return res;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	108 }
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	109 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	110 PyMem_Free(res);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	111 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	112 /* Conversion failed. Fall back to escaping with surrogateescape. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	113 #ifdef HAVE_MBRTOWC
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	114 /* Try conversion with mbrtwoc (C99), and escape non-decodable bytes. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	115
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	116 /* Overallocate; as multi-byte characters are in the argument, the
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	117 actual output could use less memory. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	118 argsize = strlen(arg) + 1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	119 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	120 if (!res)
76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	121 goto oom;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	122 in = (unsigned char*)arg;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	123 out = res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	124 memset(&mbs, 0, sizeof mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	125 while (argsize) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	126 size_t converted = mbrtowc(out, (char*)in, argsize, &mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	127 if (converted == 0)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	128 /* Reached end of string; null char stored. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	129 break;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	130 if (converted == (size_t)-2) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	131 /* Incomplete character. This should never happen,
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	132 since we provide everything that we have -
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	133 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	134 misunderstood how mbrtowc works. */
66094 76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	135 PyMem_Free(res);
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	136 if (size != NULL)
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	137 *size = (size_t)-2;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	138 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	139 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	140 if (converted == (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	141 /* Conversion error. Escape as UTF-8b, and start over
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	142 in the initial shift state. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	143 out++ = 0xdc00 + in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	144 argsize--;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	145 memset(&mbs, 0, sizeof mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	146 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	147 }
80057 13dd8199c112 Issue #16330: Use surrogate-related macros Victor Stinner <victor.stinner@gmail.com> parents: 75353 diff changeset	148 if (Py_UNICODE_IS_SURROGATE(*out)) {
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	149 /* Surrogate character. Escape the original
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	150 byte sequence with surrogateescape. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	151 argsize -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	152 while (converted--)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	153 out++ = 0xdc00 + in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	154 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	155 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	156 /* successfully converted some bytes */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	157 in += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	158 argsize -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	159 out++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	160 }
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	161 #else /* HAVE_MBRTOWC */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	162 /* 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	163 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	164 correctly in the locale's charset, which must be an ASCII superset. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	165 res = PyMem_Malloc((strlen(arg)+1)*sizeof(wchar_t));
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	166 if (!res)
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	167 goto oom;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	168 in = (unsigned char*)arg;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	169 out = res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	170 while(*in)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	171 if(*in < 128)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	172 out++ = in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	173 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	174 out++ = 0xdc00 + in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	175 *out = 0;
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	176 #endif /* HAVE_MBRTOWC */
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	177 if (size != NULL)
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	178 *size = out - res;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	179 return res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	180 oom:
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	181 if (size != NULL)
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	182 *size = (size_t)-1;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	183 return NULL;
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	184 #endif /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	185 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	186
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	187 /* 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	188 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	189 converted to bytes 0x80..0xFF).
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	190
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	191 This function is the reverse of _Py_char2wchar().
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	192
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	193 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	194 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	195
1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	196 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	197 on encoding error, or (size_t)-1 otherwise. */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	198 char*
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	199 _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	200 {
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	201 #ifdef __APPLE__
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	202 Py_ssize_t len;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	203 PyObject unicode, bytes = NULL;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	204 char *cpath;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	205
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	206 unicode = PyUnicode_FromWideChar(text, wcslen(text));
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	207 if (unicode == NULL) {
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	208 Py_DECREF(unicode);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	209 return NULL;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	210 }
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	211
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	212 bytes = _PyUnicode_AsUTF8String(unicode, "surrogateescape");
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	213 Py_DECREF(unicode);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	214 if (bytes == NULL) {
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	215 PyErr_Clear();
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	216 return NULL;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	217 }
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	218
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	219 len = PyBytes_GET_SIZE(bytes);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	220 cpath = PyMem_Malloc(len+1);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	221 if (cpath == NULL) {
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	222 Py_DECREF(bytes);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	223 return NULL;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	224 }
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	225 memcpy(cpath, PyBytes_AsString(bytes), len + 1);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	226 Py_DECREF(bytes);
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	227 return cpath;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	228 #else /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	229 const size_t len = wcslen(text);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	230 char result = NULL, bytes = NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	231 size_t i, size, converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	232 wchar_t c, buf[2];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	233
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	234 if (error_pos != NULL)
1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	235 *error_pos = (size_t)-1;
1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	236
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	237 /* The function works in two steps:
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	238 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	239 2. outputs the bytes */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	240 size = 0;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	241 buf[1] = 0;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	242 while (1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	243 for (i=0; i < len; i++) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	244 c = text[i];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	245 if (c >= 0xdc80 && c <= 0xdcff) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	246 /* UTF-8b surrogate */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	247 if (bytes != NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	248 *bytes++ = c - 0xdc00;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	249 size--;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	250 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	251 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	252 size++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	253 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	254 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	255 else {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	256 buf[0] = c;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	257 if (bytes != NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	258 converted = wcstombs(bytes, buf, size);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	259 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	260 converted = wcstombs(NULL, buf, 0);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	261 if (converted == (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	262 if (result != NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	263 PyMem_Free(result);
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	264 if (error_pos != NULL)
1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	265 *error_pos = i;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	266 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	267 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	268 if (bytes != NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	269 bytes += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	270 size -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	271 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	272 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	273 size += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	274 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	275 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	276 if (result != NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	277 *bytes = 0;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	278 break;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	279 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	280
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	281 size += 1; /* nul byte at the end */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	282 result = PyMem_Malloc(size);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	283 if (result == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	284 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	285 bytes = result;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	286 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	287 return result;
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	288 #endif /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	289 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	290
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	291 /* In principle, this should use HAVE__WSTAT, and _wstat
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	292 should be detected by autoconf. However, no current
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	293 POSIX system provides that function, so testing for
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	294 it is pointless.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	295 Not sure whether the MS_WINDOWS guards are necessary:
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	296 perhaps for cygwin/mingw builds?
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	297 */
65164 1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	298 #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	299
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	300 /* 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	301
65164 1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	302 int
1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	303 _Py_wstat(const wchar_t* path, struct stat *buf)
1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	304 {
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	305 int err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	306 char *fname;
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	307 fname = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	308 if (fname == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	309 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	310 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	311 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	312 err = stat(fname, buf);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	313 PyMem_Free(fname);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	314 return err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	315 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	316 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	317
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	318 /* 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	319 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	320
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	321 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	322 raised. */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	323
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	324 int
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	325 _Py_stat(PyObject path, struct stat statbuf)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	326 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	327 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	328 int err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	329 struct _stat wstatbuf;
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	330 wchar_t *wpath;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	331
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	332 wpath = PyUnicode_AsUnicode(path);
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	333 if (wpath == NULL)
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	334 return -2;
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	335 err = _wstat(wpath, &wstatbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	336 if (!err)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	337 statbuf->st_mode = wstatbuf.st_mode;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	338 return err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	339 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	340 int ret;
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	341 PyObject *bytes = PyUnicode_EncodeFSDefault(path);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	342 if (bytes == NULL)
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	343 return -2;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	344 ret = stat(PyBytes_AS_STRING(bytes), statbuf);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	345 Py_DECREF(bytes);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	346 return ret;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	347 #endif
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
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	350 /* 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	351 encoding and use fopen() otherwise. */
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	352
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	353 FILE *
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	354 _Py_wfopen(const wchar_t path, const wchar_t mode)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	355 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	356 #ifndef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	357 FILE *f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	358 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	359 char cmode[10];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	360 size_t r;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	361 r = wcstombs(cmode, mode, 10);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	362 if (r == (size_t)-1 \|\| r >= 10) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	363 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	364 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	365 }
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	366 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	367 if (cpath == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	368 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	369 f = fopen(cpath, cmode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	370 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	371 return f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	372 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	373 return _wfopen(path, mode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	374 #endif
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
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	377 /* 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	378 call fopen() otherwise.
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	379
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	380 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	381 (if PyErr_Occurred()) on unicode error */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	382
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	383 FILE*
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	384 _Py_fopen(PyObject path, const char mode)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	385 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	386 #ifdef MS_WINDOWS
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	387 wchar_t *wpath;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	388 wchar_t wmode[10];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	389 int usize;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	390
74129 33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	391 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	392 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	393 "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	394 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	395 return NULL;
33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	396 }
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	397 wpath = PyUnicode_AsUnicode(path);
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	398 if (wpath == NULL)
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	399 return NULL;
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	400
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	401 usize = MultiByteToWideChar(CP_ACP, 0, mode, -1, wmode, sizeof(wmode));
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	402 if (usize == 0)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	403 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	404
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	405 return _wfopen(wpath, wmode);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	406 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	407 FILE *f;
74085 a882aa30a713 _Py_fopen now allows bytes filenames under non-Windows platforms. Antoine Pitrou <solipsis@pitrou.net> parents: 74058 diff changeset	408 PyObject *bytes;
a882aa30a713 _Py_fopen now allows bytes filenames under non-Windows platforms. Antoine Pitrou <solipsis@pitrou.net> parents: 74058 diff changeset	409 if (!PyUnicode_FSConverter(path, &bytes))
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	410 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	411 f = fopen(PyBytes_AS_STRING(bytes), mode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	412 Py_DECREF(bytes);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	413 return f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	414 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	415 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	416
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	417 #ifdef HAVE_READLINK
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	418
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	419 /* 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	420 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	421
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	422 int
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	423 _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	424 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	425 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	426 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	427 wchar_t *wbuf;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	428 int res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	429 size_t r1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	430
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	431 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	432 if (cpath == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	433 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	434 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	435 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	436 res = (int)readlink(cpath, cbuf, PATH_MAX);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	437 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	438 if (res == -1)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	439 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	440 if (res == PATH_MAX) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	441 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	442 return -1;
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 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	445 wbuf = _Py_char2wchar(cbuf, &r1);
65410 6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	446 if (wbuf == NULL) {
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	447 errno = EINVAL;
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	448 return -1;
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	449 }
65409 e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65346 diff changeset	450 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	451 PyMem_Free(wbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	452 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	453 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	454 }
65409 e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65346 diff changeset	455 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	456 PyMem_Free(wbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	457 return (int)r1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	458 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	459 #endif
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 #ifdef HAVE_REALPATH
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	462
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	463 /* 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	464 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	465 Return NULL on error. */
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	466
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	467 wchar_t*
65166 8fd5d06462e6 _Py_wrealpath() requires the size of the output buffer Victor Stinner <victor.stinner@haypocalc.com> parents: 65165 diff changeset	468 _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	469 wchar_t *resolved_path, size_t resolved_path_size)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	470 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	471 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	472 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	473 wchar_t *wresolved_path;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	474 char *res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	475 size_t r;
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	476 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	477 if (cpath == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	478 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	479 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	480 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	481 res = realpath(cpath, cresolved_path);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	482 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	483 if (res == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	484 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	485
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	486 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	487 if (wresolved_path == NULL) {
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	488 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	489 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	490 }
65411 df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	491 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	492 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	493 errno = EINVAL;
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	494 return NULL;
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	495 }
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	496 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	497 PyMem_Free(wresolved_path);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	498 return resolved_path;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	499 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	500 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	501
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	502 /* 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	503 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	504 Return NULL on error. */
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	505
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	506 wchar_t*
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	507 _Py_wgetcwd(wchar_t *buf, size_t size)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	508 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	509 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	510 return _wgetcwd(buf, size);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	511 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	512 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	513 wchar_t *wname;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	514 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	515
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	516 if (getcwd(fname, PATH_MAX) == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	517 return NULL;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	518 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	519 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	520 return NULL;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	521 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	522 PyMem_Free(wname);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	523 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	524 }
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	525 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	526 PyMem_Free(wname);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	527 return buf;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	528 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	529 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	530
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	531 #endif

Mercurial > cpython-lru_cache-weakref

annotate Python/fileutils.c @ 80411:48fbdaf3a849