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

annotate Python/fileutils.c @ 80415:f3e512b5ffb3

Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions

author	Victor Stinner <victor.stinner@gmail.com>
date	Mon, 12 Nov 2012 23:32:21 +0100
parents	48fbdaf3a849
children	3b344eac6b0f

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));
80415 f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	70 if (size != NULL) {
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	71 if (wstr != NULL)
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	72 *size = wcslen(wstr);
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	73 else
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	74 *size = (size_t)-1;
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	75 }
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	76 return wstr;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	77 #else
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	78 wchar_t *res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	79 #ifdef HAVE_BROKEN_MBSTOWCS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	80 /* Some platforms have a broken implementation of
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	81 * mbstowcs which does not count the characters that
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	82 * would result from conversion. Use an upper bound.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	83 */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	84 size_t argsize = strlen(arg);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	85 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	86 size_t argsize = mbstowcs(NULL, arg, 0);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	87 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	88 size_t count;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	89 unsigned char *in;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	90 wchar_t *out;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	91 #ifdef HAVE_MBRTOWC
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	92 mbstate_t mbs;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	93 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	94 if (argsize != (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	95 res = (wchar_t )PyMem_Malloc((argsize+1)sizeof(wchar_t));
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	96 if (!res)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	97 goto oom;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	98 count = mbstowcs(res, arg, argsize+1);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	99 if (count != (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	100 wchar_t *tmp;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	101 /* Only use the result if it contains no
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	102 surrogate characters. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	103 for (tmp = res; *tmp != 0 &&
80057 13dd8199c112 Issue #16330: Use surrogate-related macros Victor Stinner <victor.stinner@gmail.com> parents: 75353 diff changeset	104 !Py_UNICODE_IS_SURROGATE(*tmp); tmp++)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	105 ;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	106 if (*tmp == 0) {
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	107 if (size != NULL)
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	108 *size = count;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	109 return res;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	110 }
65162 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 PyMem_Free(res);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	113 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	114 /* Conversion failed. Fall back to escaping with surrogateescape. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	115 #ifdef HAVE_MBRTOWC
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	116 /* Try conversion with mbrtwoc (C99), and escape non-decodable bytes. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	117
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	118 /* Overallocate; as multi-byte characters are in the argument, the
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	119 actual output could use less memory. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	120 argsize = strlen(arg) + 1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	121 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	122 if (!res)
76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	123 goto oom;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	124 in = (unsigned char*)arg;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	125 out = res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	126 memset(&mbs, 0, sizeof mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	127 while (argsize) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	128 size_t converted = mbrtowc(out, (char*)in, argsize, &mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	129 if (converted == 0)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	130 /* Reached end of string; null char stored. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	131 break;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	132 if (converted == (size_t)-2) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	133 /* Incomplete character. This should never happen,
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	134 since we provide everything that we have -
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	135 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	136 misunderstood how mbrtowc works. */
66094 76a1c5e88352 _Py_char2wchar() frees the memory on conversion error Victor Stinner <victor.stinner@haypocalc.com> parents: 66092 diff changeset	137 PyMem_Free(res);
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	138 if (size != NULL)
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	139 *size = (size_t)-2;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	140 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	141 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	142 if (converted == (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	143 /* Conversion error. Escape as UTF-8b, and start over
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	144 in the initial shift state. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	145 out++ = 0xdc00 + in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	146 argsize--;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	147 memset(&mbs, 0, sizeof mbs);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	148 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	149 }
80057 13dd8199c112 Issue #16330: Use surrogate-related macros Victor Stinner <victor.stinner@gmail.com> parents: 75353 diff changeset	150 if (Py_UNICODE_IS_SURROGATE(*out)) {
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	151 /* Surrogate character. Escape the original
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	152 byte sequence with surrogateescape. */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	153 argsize -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	154 while (converted--)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	155 out++ = 0xdc00 + in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	156 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	157 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	158 /* successfully converted some bytes */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	159 in += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	160 argsize -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	161 out++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	162 }
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	163 #else /* HAVE_MBRTOWC */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	164 /* 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	165 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	166 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	167 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	168 if (!res)
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	169 goto oom;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	170 in = (unsigned char*)arg;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	171 out = res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	172 while(*in)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	173 if(*in < 128)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	174 out++ = in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	175 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	176 out++ = 0xdc00 + in++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	177 *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	178 #endif /* HAVE_MBRTOWC */
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	179 if (size != NULL)
3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	180 *size = out - res;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	181 return res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	182 oom:
74002 279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	183 if (size != NULL)
279b0aee0cfb Add PyUnicode_DecodeLocaleAndSize() and PyUnicode_DecodeLocale() Victor Stinner <victor.stinner@haypocalc.com> parents: 73593 diff changeset	184 *size = (size_t)-1;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	185 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	186 #endif /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	187 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	188
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	189 /* 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	190 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	191 converted to bytes 0x80..0xFF).
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 This function is the reverse of _Py_char2wchar().
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	194
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	195 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	196 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	197
1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	198 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	199 on encoding error, or (size_t)-1 otherwise. */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	200 char*
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	201 _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	202 {
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	203 #ifdef __APPLE__
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	204 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	205 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	206 char *cpath;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	207
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	208 unicode = PyUnicode_FromWideChar(text, wcslen(text));
80415 f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	209 if (unicode == NULL)
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	210 return NULL;
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();
80415 f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	216 if (error_pos != NULL)
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	217 *error_pos = (size_t)-1;
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	218 return NULL;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	219 }
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	220
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	221 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	222 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	223 if (cpath == NULL) {
80415 f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	224 PyErr_Clear();
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	225 Py_DECREF(bytes);
80415 f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	226 if (error_pos != NULL)
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	227 *error_pos = (size_t)-1;
80411 48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	228 return NULL;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	229 }
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	230 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	231 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	232 return cpath;
48fbdaf3a849 Issue #16416: OS data are now always encoded/decoded to/from Victor Stinner <victor.stinner@gmail.com> parents: 80407 diff changeset	233 #else /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	234 const size_t len = wcslen(text);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	235 char result = NULL, bytes = NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	236 size_t i, size, converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	237 wchar_t c, buf[2];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	238
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	239 /* The function works in two steps:
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	240 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	241 2. outputs the bytes */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	242 size = 0;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	243 buf[1] = 0;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	244 while (1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	245 for (i=0; i < len; i++) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	246 c = text[i];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	247 if (c >= 0xdc80 && c <= 0xdcff) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	248 /* UTF-8b surrogate */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	249 if (bytes != NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	250 *bytes++ = c - 0xdc00;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	251 size--;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	252 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	253 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	254 size++;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	255 continue;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	256 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	257 else {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	258 buf[0] = c;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	259 if (bytes != NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	260 converted = wcstombs(bytes, buf, size);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	261 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	262 converted = wcstombs(NULL, buf, 0);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	263 if (converted == (size_t)-1) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	264 if (result != NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	265 PyMem_Free(result);
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	266 if (error_pos != NULL)
1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	267 *error_pos = i;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	268 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	269 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	270 if (bytes != NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	271 bytes += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	272 size -= converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	273 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	274 else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	275 size += converted;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	276 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	277 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	278 if (result != NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	279 *bytes = 0;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	280 break;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	281 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	282
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	283 size += 1; /* nul byte at the end */
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	284 result = PyMem_Malloc(size);
80415 f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	285 if (result == NULL) {
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	286 if (error_pos != NULL)
f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	287 *error_pos = (size_t)-1;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	288 return NULL;
80415 f3e512b5ffb3 Issue #16416: Fix error handling in _Py_wchar2char() _Py_char2wchar() functions Victor Stinner <victor.stinner@gmail.com> parents: 80411 diff changeset	289 }
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	290 bytes = result;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	291 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	292 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	293 #endif /* __APPLE__ */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	294 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	295
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	296 /* In principle, this should use HAVE__WSTAT, and _wstat
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	297 should be detected by autoconf. However, no current
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	298 POSIX system provides that function, so testing for
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	299 it is pointless.
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	300 Not sure whether the MS_WINDOWS guards are necessary:
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	301 perhaps for cygwin/mingw builds?
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	302 */
65164 1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	303 #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	304
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	305 /* 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	306
65164 1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	307 int
1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	308 _Py_wstat(const wchar_t* path, struct stat *buf)
1f3919af8a53 Fix fileutils for Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65163 diff changeset	309 {
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	310 int err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	311 char *fname;
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	312 fname = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	313 if (fname == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	314 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	315 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	316 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	317 err = stat(fname, buf);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	318 PyMem_Free(fname);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	319 return err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	320 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	321 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	322
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	323 /* 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	324 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	325
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	326 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	327 raised. */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	328
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	329 int
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	330 _Py_stat(PyObject path, struct stat statbuf)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	331 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	332 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	333 int err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	334 struct _stat wstatbuf;
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	335 wchar_t *wpath;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	336
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	337 wpath = PyUnicode_AsUnicode(path);
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	338 if (wpath == NULL)
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	339 return -2;
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	340 err = _wstat(wpath, &wstatbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	341 if (!err)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	342 statbuf->st_mode = wstatbuf.st_mode;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	343 return err;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	344 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	345 int ret;
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	346 PyObject *bytes = PyUnicode_EncodeFSDefault(path);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	347 if (bytes == NULL)
74058 2a140b01d343 import.c now catchs _Py_stat() exceptions Victor Stinner <victor.stinner@haypocalc.com> parents: 74002 diff changeset	348 return -2;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	349 ret = stat(PyBytes_AS_STRING(bytes), statbuf);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	350 Py_DECREF(bytes);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	351 return ret;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	352 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	353 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	354
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	355 /* 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	356 encoding and use fopen() otherwise. */
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	357
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	358 FILE *
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	359 _Py_wfopen(const wchar_t path, const wchar_t mode)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	360 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	361 #ifndef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	362 FILE *f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	363 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	364 char cmode[10];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	365 size_t r;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	366 r = wcstombs(cmode, mode, 10);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	367 if (r == (size_t)-1 \|\| r >= 10) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	368 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	369 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	370 }
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	371 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	372 if (cpath == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	373 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	374 f = fopen(cpath, cmode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	375 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	376 return f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	377 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	378 return _wfopen(path, mode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	379 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	380 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	381
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	382 /* 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	383 call fopen() otherwise.
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	384
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	385 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	386 (if PyErr_Occurred()) on unicode error */
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	387
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	388 FILE*
65165 2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows Victor Stinner <victor.stinner@haypocalc.com> parents: 65164 diff changeset	389 _Py_fopen(PyObject path, const char mode)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	390 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	391 #ifdef MS_WINDOWS
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	392 wchar_t *wpath;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	393 wchar_t wmode[10];
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	394 int usize;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	395
74129 33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	396 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	397 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	398 "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	399 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	400 return NULL;
33dea851f918 Issue #13626: Add support for SSL Diffie-Hellman key exchange, through the Antoine Pitrou <solipsis@pitrou.net> parents: 74085 diff changeset	401 }
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	402 wpath = PyUnicode_AsUnicode(path);
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	403 if (wpath == NULL)
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	404 return NULL;
94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	405
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	406 usize = MultiByteToWideChar(CP_ACP, 0, mode, -1, wmode, sizeof(wmode));
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	407 if (usize == 0)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	408 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	409
73593 94a38ad4aebf Catch PyUnicode_AS_UNICODE() errors in fileutils.c Victor Stinner <victor.stinner@haypocalc.com> parents: 66094 diff changeset	410 return _wfopen(wpath, wmode);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	411 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	412 FILE *f;
74085 a882aa30a713 _Py_fopen now allows bytes filenames under non-Windows platforms. Antoine Pitrou <solipsis@pitrou.net> parents: 74058 diff changeset	413 PyObject *bytes;
a882aa30a713 _Py_fopen now allows bytes filenames under non-Windows platforms. Antoine Pitrou <solipsis@pitrou.net> parents: 74058 diff changeset	414 if (!PyUnicode_FSConverter(path, &bytes))
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	415 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	416 f = fopen(PyBytes_AS_STRING(bytes), mode);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	417 Py_DECREF(bytes);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	418 return f;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	419 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	420 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	421
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	422 #ifdef HAVE_READLINK
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	423
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	424 /* 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	425 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	426
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	427 int
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	428 _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	429 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	430 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	431 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	432 wchar_t *wbuf;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	433 int res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	434 size_t r1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	435
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	436 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	437 if (cpath == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	438 errno = EINVAL;
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 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	441 res = (int)readlink(cpath, cbuf, PATH_MAX);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	442 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	443 if (res == -1)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	444 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	445 if (res == PATH_MAX) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	446 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	447 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	448 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	449 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	450 wbuf = _Py_char2wchar(cbuf, &r1);
65410 6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	451 if (wbuf == NULL) {
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	452 errno = EINVAL;
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 diff changeset	453 return -1;
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure Victor Stinner <victor.stinner@haypocalc.com> parents: 65409 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 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	456 PyMem_Free(wbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	457 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	458 return -1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	459 }
65409 e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65346 diff changeset	460 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	461 PyMem_Free(wbuf);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	462 return (int)r1;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	463 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	464 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	465
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	466 #ifdef HAVE_REALPATH
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	467
1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	468 /* 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	469 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	470 Return NULL on error. */
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	471
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	472 wchar_t*
65166 8fd5d06462e6 _Py_wrealpath() requires the size of the output buffer Victor Stinner <victor.stinner@haypocalc.com> parents: 65165 diff changeset	473 _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	474 wchar_t *resolved_path, size_t resolved_path_size)
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	475 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	476 char *cpath;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	477 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	478 wchar_t *wresolved_path;
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	479 char *res;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	480 size_t r;
66092 1b5f47e2e0f0 PyUnicode_EncodeFS() raises an exception if _Py_wchar2char() fails Victor Stinner <victor.stinner@haypocalc.com> parents: 65412 diff changeset	481 cpath = _Py_wchar2char(path, NULL);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	482 if (cpath == NULL) {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	483 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	484 return NULL;
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 res = realpath(cpath, cresolved_path);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	487 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	488 if (res == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	489 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	490
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	491 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	492 if (wresolved_path == NULL) {
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	493 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	494 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	495 }
65411 df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	496 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	497 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	498 errno = EINVAL;
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	499 return NULL;
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	500 }
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support Victor Stinner <victor.stinner@haypocalc.com> parents: 65410 diff changeset	501 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	502 PyMem_Free(wresolved_path);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	503 return resolved_path;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	504 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	505 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	506
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	507 /* 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	508 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	509 Return NULL on error. */
65168 1c4774b4e98a fileutils.c: document which encodings are used Victor Stinner <victor.stinner@haypocalc.com> parents: 65166 diff changeset	510
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	511 wchar_t*
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	512 _Py_wgetcwd(wchar_t *buf, size_t size)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	513 {
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	514 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	515 return _wgetcwd(buf, size);
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	516 #else
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	517 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	518 wchar_t *wname;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	519 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	520
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	521 if (getcwd(fname, PATH_MAX) == NULL)
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	522 return NULL;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	523 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	524 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	525 return NULL;
65412 3794c9b50824 Add an optional size argument to _Py_char2wchar() Victor Stinner <victor.stinner@haypocalc.com> parents: 65411 diff changeset	526 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	527 PyMem_Free(wname);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	528 return NULL;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	529 }
65312 a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates Victor Stinner <victor.stinner@haypocalc.com> parents: 65168 diff changeset	530 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	531 PyMem_Free(wname);
65162 39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	532 return buf;
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	533 #endif
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	534 }
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	535
39014f0abc32 Create fileutils.c/.h Victor Stinner <victor.stinner@haypocalc.com> parents: diff changeset	536 #endif

Mercurial > cpython-lru_cache-weakref

annotate Python/fileutils.c @ 80415:f3e512b5ffb3