Mercurial > cpython-lru_cache-weakref

annotate Python/fileutils.c @ 65411:df0e348e7176

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
_Py_wrealpath() uses _Py_char2wchar() to decode the result, to support surrogate characters.
author Victor Stinner <victor.stinner@haypocalc.com>
date Sat, 16 Oct 2010 22:55:47 +0000
parents 6dcba93d6262
children 3794c9b50824
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
1 #include "Python.h"
65164
1f3919af8a53 Fix fileutils for Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65163
diff changeset
2 #ifdef MS_WINDOWS
1f3919af8a53 Fix fileutils for Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65163
diff changeset
3 # include <windows.h>
1f3919af8a53 Fix fileutils for Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65163
diff changeset
4 #endif
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
5
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
6 #ifdef HAVE_STAT
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
7
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
8 /* Decode a byte string from the locale encoding with the
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
9 surrogateescape error handler (undecodable bytes are decoded as characters
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
10 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
11 character, escape the bytes using the surrogateescape error handler instead
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
12 of decoding them.
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
13
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
14 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
15
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
16 Return a pointer to a newly allocated (wide) character string (use
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
17 PyMem_Free() to free the memory), or NULL on error (conversion error or
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
18 memory error). */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
19 wchar_t*
65346
cc1d8a2db348 Mark _Py_char2wchar() input argument as constant
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65312
diff changeset
20 _Py_char2wchar(const char* arg)
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
21 {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
22 wchar_t *res;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
23 #ifdef HAVE_BROKEN_MBSTOWCS
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
24 /* Some platforms have a broken implementation of
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
25 * mbstowcs which does not count the characters that
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
26 * would result from conversion. Use an upper bound.
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
27 */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
28 size_t argsize = strlen(arg);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
29 #else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
30 size_t argsize = mbstowcs(NULL, arg, 0);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
31 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
32 size_t count;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
33 unsigned char *in;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
34 wchar_t *out;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
35 #ifdef HAVE_MBRTOWC
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
36 mbstate_t mbs;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
37 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
38 if (argsize != (size_t)-1) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
39 res = (wchar_t *)PyMem_Malloc((argsize+1)*sizeof(wchar_t));
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
40 if (!res)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
41 goto oom;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
42 count = mbstowcs(res, arg, argsize+1);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
43 if (count != (size_t)-1) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
44 wchar_t *tmp;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
45 /* Only use the result if it contains no
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
46 surrogate characters. */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
47 for (tmp = res; *tmp != 0 &&
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
48 (*tmp < 0xd800 || *tmp > 0xdfff); tmp++)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
49 ;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
50 if (*tmp == 0)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
51 return res;
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 PyMem_Free(res);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
54 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
55 /* Conversion failed. Fall back to escaping with surrogateescape. */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
56 #ifdef HAVE_MBRTOWC
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
57 /* Try conversion with mbrtwoc (C99), and escape non-decodable bytes. */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
58
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
59 /* Overallocate; as multi-byte characters are in the argument, the
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
60 actual output could use less memory. */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
61 argsize = strlen(arg) + 1;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
62 res = (wchar_t*)PyMem_Malloc(argsize*sizeof(wchar_t));
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
63 if (!res) goto oom;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
64 in = (unsigned char*)arg;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
65 out = res;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
66 memset(&mbs, 0, sizeof mbs);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
67 while (argsize) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
68 size_t converted = mbrtowc(out, (char*)in, argsize, &mbs);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
69 if (converted == 0)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
70 /* Reached end of string; null char stored. */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
71 break;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
72 if (converted == (size_t)-2) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
73 /* Incomplete character. This should never happen,
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
74 since we provide everything that we have -
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
75 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
76 misunderstood how mbrtowc works. */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
77 fprintf(stderr, "unexpected mbrtowc result -2\n");
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
78 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
79 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
80 if (converted == (size_t)-1) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
81 /* Conversion error. Escape as UTF-8b, and start over
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
82 in the initial shift state. */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
83 *out++ = 0xdc00 + *in++;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
84 argsize--;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
85 memset(&mbs, 0, sizeof mbs);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
86 continue;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
87 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
88 if (*out >= 0xd800 && *out <= 0xdfff) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
89 /* Surrogate character. Escape the original
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
90 byte sequence with surrogateescape. */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
91 argsize -= converted;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
92 while (converted--)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
93 *out++ = 0xdc00 + *in++;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
94 continue;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
95 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
96 /* successfully converted some bytes */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
97 in += converted;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
98 argsize -= converted;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
99 out++;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
100 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
101 #else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
102 /* 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
103 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
104 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
105 res = PyMem_Malloc((strlen(arg)+1)*sizeof(wchar_t));
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
106 if (!res) goto oom;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
107 in = (unsigned char*)arg;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
108 out = res;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
109 while(*in)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
110 if(*in < 128)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
111 *out++ = *in++;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
112 else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
113 *out++ = 0xdc00 + *in++;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
114 *out = 0;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
115 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
116 return res;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
117 oom:
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
118 fprintf(stderr, "out of memory\n");
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
119 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
120 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
121
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
122 /* 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
123 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
124 converted to bytes 0x80..0xFF).
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
125
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
126 This function is the reverse of _Py_char2wchar().
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
127
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
128 Return a pointer to a newly allocated byte string (use PyMem_Free() to free
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
129 the memory), or NULL on error (conversion error or memory error). */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
130 char*
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
131 _Py_wchar2char(const wchar_t *text)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
132 {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
133 const size_t len = wcslen(text);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
134 char *result = NULL, *bytes = NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
135 size_t i, size, converted;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
136 wchar_t c, buf[2];
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
137
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
138 /* The function works in two steps:
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
139 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
140 2. outputs the bytes */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
141 size = 0;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
142 buf[1] = 0;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
143 while (1) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
144 for (i=0; i < len; i++) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
145 c = text[i];
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
146 if (c >= 0xdc80 && c <= 0xdcff) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
147 /* UTF-8b surrogate */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
148 if (bytes != NULL) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
149 *bytes++ = c - 0xdc00;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
150 size--;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
151 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
152 else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
153 size++;
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 else {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
157 buf[0] = c;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
158 if (bytes != NULL)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
159 converted = wcstombs(bytes, buf, size);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
160 else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
161 converted = wcstombs(NULL, buf, 0);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
162 if (converted == (size_t)-1) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
163 if (result != NULL)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
164 PyMem_Free(result);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
165 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
166 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
167 if (bytes != NULL) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
168 bytes += converted;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
169 size -= converted;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
170 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
171 else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
172 size += converted;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
173 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
174 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
175 if (result != NULL) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
176 *bytes = 0;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
177 break;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
178 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
179
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
180 size += 1; /* nul byte at the end */
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
181 result = PyMem_Malloc(size);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
182 if (result == NULL)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
183 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
184 bytes = result;
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 return result;
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 /* In principle, this should use HAVE__WSTAT, and _wstat
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
190 should be detected by autoconf. However, no current
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
191 POSIX system provides that function, so testing for
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
192 it is pointless.
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
193 Not sure whether the MS_WINDOWS guards are necessary:
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
194 perhaps for cygwin/mingw builds?
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
195 */
65164
1f3919af8a53 Fix fileutils for Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65163
diff changeset
196 #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
197
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
198 /* 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
199
65164
1f3919af8a53 Fix fileutils for Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65163
diff changeset
200 int
1f3919af8a53 Fix fileutils for Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65163
diff changeset
201 _Py_wstat(const wchar_t* path, struct stat *buf)
1f3919af8a53 Fix fileutils for Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65163
diff changeset
202 {
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
203 int err;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
204 char *fname;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
205 fname = _Py_wchar2char(path);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
206 if (fname == NULL) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
207 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
208 return -1;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
209 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
210 err = stat(fname, buf);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
211 PyMem_Free(fname);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
212 return err;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
213 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
214 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
215
65168
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
216 /* 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
217 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
218
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
219 Return 0 on success, -1 on _wstat() / stat() error or (if PyErr_Occurred())
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
220 unicode error. */
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
221
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
222 int
65165
2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65164
diff changeset
223 _Py_stat(PyObject *path, struct stat *statbuf)
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
224 {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
225 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
226 int err;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
227 struct _stat wstatbuf;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
228
65165
2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65164
diff changeset
229 err = _wstat(PyUnicode_AS_UNICODE(path), &wstatbuf);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
230 if (!err)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
231 statbuf->st_mode = wstatbuf.st_mode;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
232 return err;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
233 #else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
234 int ret;
65165
2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65164
diff changeset
235 PyObject *bytes = PyUnicode_EncodeFSDefault(path);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
236 if (bytes == NULL)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
237 return -1;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
238 ret = stat(PyBytes_AS_STRING(bytes), statbuf);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
239 Py_DECREF(bytes);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
240 return ret;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
241 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
242 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
243
65168
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
244 /* 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
245 encoding and use fopen() otherwise. */
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
246
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
247 FILE *
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
248 _Py_wfopen(const wchar_t *path, const wchar_t *mode)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
249 {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
250 #ifndef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
251 FILE *f;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
252 char *cpath;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
253 char cmode[10];
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
254 size_t r;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
255 r = wcstombs(cmode, mode, 10);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
256 if (r == (size_t)-1 || r >= 10) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
257 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
258 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
259 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
260 cpath = _Py_wchar2char(path);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
261 if (cpath == NULL)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
262 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
263 f = fopen(cpath, cmode);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
264 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
265 return f;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
266 #else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
267 return _wfopen(path, mode);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
268 #endif
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
65168
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
271 /* 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
272 call fopen() otherwise.
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
273
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
274 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
275 (if PyErr_Occurred()) on unicode error */
65162
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 FILE*
65165
2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65164
diff changeset
278 _Py_fopen(PyObject *path, const char *mode)
65162
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 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
281 wchar_t wmode[10];
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
282 int usize;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
283
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
284 usize = MultiByteToWideChar(CP_ACP, 0, mode, -1, wmode, sizeof(wmode));
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
285 if (usize == 0)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
286 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
287
65165
2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65164
diff changeset
288 return _wfopen(PyUnicode_AS_UNICODE(path), wmode);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
289 #else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
290 FILE *f;
65165
2147cc152227 _Py_stat() and _Py_fopen(): avoid PyUnicode_AsWideCharString() on Windows
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65164
diff changeset
291 PyObject *bytes = PyUnicode_EncodeFSDefault(path);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
292 if (bytes == NULL)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
293 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
294 f = fopen(PyBytes_AS_STRING(bytes), mode);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
295 Py_DECREF(bytes);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
296 return f;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
297 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
298 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
299
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
300 #ifdef HAVE_READLINK
65168
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
301
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
302 /* Read value of symbolic link. Encode the path to the locale encoding, decode
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
303 the result from the locale encoding. */
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
304
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
305 int
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
306 _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
307 {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
308 char *cpath;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
309 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
310 wchar_t *wbuf;
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
311 int res;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
312 size_t r1;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
313
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
314 cpath = _Py_wchar2char(path);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
315 if (cpath == NULL) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
316 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
317 return -1;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
318 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
319 res = (int)readlink(cpath, cbuf, PATH_MAX);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
320 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
321 if (res == -1)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
322 return -1;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
323 if (res == PATH_MAX) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
324 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
325 return -1;
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 cbuf[res] = '\0'; /* buf will be null terminated */
65409
e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65346
diff changeset
328 wbuf = _Py_char2wchar(cbuf);
65410
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65409
diff changeset
329 if (wbuf == NULL) {
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65409
diff changeset
330 errno = EINVAL;
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65409
diff changeset
331 return -1;
6dcba93d6262 _Py_wreadlink(): catch _Py_char2wchar() failure
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65409
diff changeset
332 }
65409
e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65346
diff changeset
333 r1 = wcslen(wbuf);
e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65346
diff changeset
334 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
335 PyMem_Free(wbuf);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
336 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
337 return -1;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
338 }
65409
e1dc9861b43f _Py_wreadlink() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65346
diff changeset
339 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
340 PyMem_Free(wbuf);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
341 return (int)r1;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
342 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
343 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
344
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
345 #ifdef HAVE_REALPATH
65168
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
346
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
347 /* Return the canonicalized absolute pathname. Encode path to the locale
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
348 encoding, decode the result from the locale encoding. */
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
349
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
350 wchar_t*
65166
8fd5d06462e6 _Py_wrealpath() requires the size of the output buffer
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65165
diff changeset
351 _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
352 wchar_t *resolved_path, size_t resolved_path_size)
65162
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 char *cpath;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
355 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
356 wchar_t *wresolved_path;
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
357 char *res;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
358 size_t r;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
359 cpath = _Py_wchar2char(path);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
360 if (cpath == NULL) {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
361 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
362 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
363 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
364 res = realpath(cpath, cresolved_path);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
365 PyMem_Free(cpath);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
366 if (res == NULL)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
367 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
368
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65410
diff changeset
369 wresolved_path = _Py_char2wchar(cresolved_path);
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65410
diff changeset
370 if (wresolved_path == NULL) {
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
371 errno = EINVAL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
372 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
373 }
65411
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65410
diff changeset
374 r = wcslen(wresolved_path);
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65410
diff changeset
375 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
376 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
377 errno = EINVAL;
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65410
diff changeset
378 return NULL;
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65410
diff changeset
379 }
df0e348e7176 _Py_wrealpath() uses _Py_char2wchar() to decode the result, to support
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65410
diff changeset
380 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
381 PyMem_Free(wresolved_path);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
382 return resolved_path;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
383 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
384 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
385
65312
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65168
diff changeset
386 /* Get the current directory. size is the buffer size in wide characters
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65168
diff changeset
387 including the null character. Decode the path from the locale encoding. */
65168
1c4774b4e98a fileutils.c: document which encodings are used
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65166
diff changeset
388
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
389 wchar_t*
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
390 _Py_wgetcwd(wchar_t *buf, size_t size)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
391 {
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
392 #ifdef MS_WINDOWS
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
393 return _wgetcwd(buf, size);
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
394 #else
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
395 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
396 wchar_t *wname;
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65168
diff changeset
397
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
398 if (getcwd(fname, PATH_MAX) == NULL)
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
399 return NULL;
65312
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65168
diff changeset
400 wname = _Py_char2wchar(fname);
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65168
diff changeset
401 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
402 return NULL;
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65168
diff changeset
403 if (size <= wcslen(wname)) {
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65168
diff changeset
404 PyMem_Free(wname);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
405 return NULL;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
406 }
65312
a27fc0ff3678 _Py_wgetcwd() decodes the path using _Py_char2wchar() to support surrogates
Victor Stinner <victor.stinner@haypocalc.com>
parents: 65168
diff changeset
407 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
408 PyMem_Free(wname);
65162
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
409 return buf;
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
410 #endif
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
411 }
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
412
39014f0abc32 Create fileutils.c/.h
Victor Stinner <victor.stinner@haypocalc.com>
parents:
diff changeset
413 #endif