summaryrefslogtreecommitdiffstats
path: root/Python/fileutils.c
blob: c563eaa5fbb7438991509e4d78fd28b7e6486f85 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
#include "Python.h"
#ifdef MS_WINDOWS
#  include <windows.h>
#endif

#ifdef HAVE_STAT

/* Decode a byte string from the locale encoding with the
   surrogateescape error handler (undecodable bytes are decoded as characters
   in range U+DC80..U+DCFF). If a byte sequence can be decoded as a surrogate
   character, escape the bytes using the surrogateescape error handler instead
   of decoding them.

   Use _Py_wchar2char() to encode the character string back to a byte string.

   Return a pointer to a newly allocated wide character string (use
   PyMem_Free() to free the memory) and write the number of written wide
   characters excluding the null character into *size if size is not NULL, or
   NULL on error (conversion or memory allocation error).

   Conversion errors should never happen, unless there is a bug in the C
   library. */
wchar_t*
_Py_char2wchar(const char* arg, size_t *size)
{
    wchar_t *res;
#ifdef HAVE_BROKEN_MBSTOWCS
    /* Some platforms have a broken implementation of
     * mbstowcs which does not count the characters that
     * would result from conversion.  Use an upper bound.
     */
    size_t argsize = strlen(arg);
#else
    size_t argsize = mbstowcs(NULL, arg, 0);
#endif
    size_t count;
    unsigned char *in;
    wchar_t *out;
#ifdef HAVE_MBRTOWC
    mbstate_t mbs;
#endif
    if (argsize != (size_t)-1) {
        res = (wchar_t *)PyMem_Malloc((argsize+1)*sizeof(wchar_t));
        if (!res)
            goto oom;
        count = mbstowcs(res, arg, argsize+1);
        if (count != (size_t)-1) {
            wchar_t *tmp;
            /* Only use the result if it contains no
               surrogate characters. */
            for (tmp = res; *tmp != 0 &&
                         (*tmp < 0xd800 || *tmp > 0xdfff); tmp++)
                ;
            if (*tmp == 0) {
                if (size != NULL)
                    *size = count;
                return res;
            }
        }
        PyMem_Free(res);
    }
    /* Conversion failed. Fall back to escaping with surrogateescape. */
#ifdef HAVE_MBRTOWC
    /* Try conversion with mbrtwoc (C99), and escape non-decodable bytes. */

    /* Overallocate; as multi-byte characters are in the argument, the
       actual output could use less memory. */
    argsize = strlen(arg) + 1;
    res = (wchar_t*)PyMem_Malloc(argsize*sizeof(wchar_t));
    if (!res)
        goto oom;
    in = (unsigned char*)arg;
    out = res;
    memset(&mbs, 0, sizeof mbs);
    while (argsize) {
        size_t converted = mbrtowc(out, (char*)in, argsize, &mbs);
        if (converted == 0)
            /* Reached end of string; null char stored. */
            break;
        if (converted == (size_t)-2) {
            /* Incomplete character. This should never happen,
               since we provide everything that we have -
               unless there is a bug in the C library, or I
               misunderstood how mbrtowc works. */
            fprintf(stderr, "unexpected mbrtowc result -2\n");
            PyMem_Free(res);
            return NULL;
        }
        if (converted == (size_t)-1) {
            /* Conversion error. Escape as UTF-8b, and start over
               in the initial shift state. */
            *out++ = 0xdc00 + *in++;
            argsize--;
            memset(&mbs, 0, sizeof mbs);
            continue;
        }
        if (*out >= 0xd800 && *out <= 0xdfff) {
            /* Surrogate character.  Escape the original
               byte sequence with surrogateescape. */
            argsize -= converted;
            while (converted--)
                *out++ = 0xdc00 + *in++;
            continue;
        }
        /* successfully converted some bytes */
        in += converted;
        argsize -= converted;
        out++;
    }
#else
    /* Cannot use C locale for escaping; manually escape as if charset
       is ASCII (i.e. escape all bytes > 128. This will still roundtrip
       correctly in the locale's charset, which must be an ASCII superset. */
    res = PyMem_Malloc((strlen(arg)+1)*sizeof(wchar_t));
    if (!res) goto oom;
    in = (unsigned char*)arg;
    out = res;
    while(*in)
        if(*in < 128)
            *out++ = *in++;
        else
            *out++ = 0xdc00 + *in++;
    *out = 0;
#endif
    if (size != NULL)
        *size = out - res;
    return res;
oom:
    fprintf(stderr, "out of memory\n");
    return NULL;
}

/* Encode a (wide) character string to the locale encoding with the
   surrogateescape error handler (characters in range U+DC80..U+DCFF are
   converted to bytes 0x80..0xFF).

   This function is the reverse of _Py_char2wchar().

   Return a pointer to a newly allocated byte string (use PyMem_Free() to free
   the memory), or NULL on conversion or memory allocation error.

   If error_pos is not NULL: *error_pos is the index of the invalid character
   on conversion error, or (size_t)-1 otherwise. */
char*
_Py_wchar2char(const wchar_t *text, size_t *error_pos)
{
    const size_t len = wcslen(text);
    char *result = NULL, *bytes = NULL;
    size_t i, size, converted;
    wchar_t c, buf[2];

    if (error_pos != NULL)
        *error_pos = (size_t)-1;

    /* The function works in two steps:
       1. compute the length of the output buffer in bytes (size)
       2. outputs the bytes */
    size = 0;
    buf[1] = 0;
    while (1) {
        for (i=0; i < len; i++) {
            c = text[i];
            if (c >= 0xdc80 && c <= 0xdcff) {
                /* UTF-8b surrogate */
                if (bytes != NULL) {
                    *bytes++ = c - 0xdc00;
                    size--;
                }
                else
                    size++;
                continue;
            }
            else {
                buf[0] = c;
                if (bytes != NULL)
                    converted = wcstombs(bytes, buf, size);
                else
                    converted = wcstombs(NULL, buf, 0);
                if (converted == (size_t)-1) {
                    if (result != NULL)
                        PyMem_Free(result);
                    if (error_pos != NULL)
                        *error_pos = i;
                    return NULL;
                }
                if (bytes != NULL) {
                    bytes += converted;
                    size -= converted;
                }
                else
                    size += converted;
            }
        }
        if (result != NULL) {
            *bytes = 0;
            break;
        }

        size += 1; /* nul byte at the end */
        result = PyMem_Malloc(size);
        if (result == NULL)
            return NULL;
        bytes = result;
    }
    return result;
}

/* In principle, this should use HAVE__WSTAT, and _wstat
   should be detected by autoconf. However, no current
   POSIX system provides that function, so testing for
   it is pointless.
   Not sure whether the MS_WINDOWS guards are necessary:
   perhaps for cygwin/mingw builds?
*/
#if defined(HAVE_STAT) && !defined(MS_WINDOWS)

/* Get file status. Encode the path to the locale encoding. */

int
_Py_wstat(const wchar_t* path, struct stat *buf)
{
    int err;
    char *fname;
    fname = _Py_wchar2char(path, NULL);
    if (fname == NULL) {
        errno = EINVAL;
        return -1;
    }
    err = stat(fname, buf);
    PyMem_Free(fname);
    return err;
}
#endif

/* Call _wstat() on Windows, or encode the path to the filesystem encoding and
   call stat() otherwise. Only fill st_mode attribute on Windows.

   Return 0 on success, -1 on _wstat() / stat() error or (if PyErr_Occurred())
   unicode error. */

int
_Py_stat(PyObject *path, struct stat *statbuf)
{
#ifdef MS_WINDOWS
    int err;
    struct _stat wstatbuf;

    err = _wstat(PyUnicode_AS_UNICODE(path), &wstatbuf);
    if (!err)
        statbuf->st_mode = wstatbuf.st_mode;
    return err;
#else
    int ret;
    PyObject *bytes = PyUnicode_EncodeFSDefault(path);
    if (bytes == NULL)
        return -1;
    ret = stat(PyBytes_AS_STRING(bytes), statbuf);
    Py_DECREF(bytes);
    return ret;
#endif
}

/* Open a file. Use _wfopen() on Windows, encode the path to the locale
   encoding and use fopen() otherwise. */

FILE *
_Py_wfopen(const wchar_t *path, const wchar_t *mode)
{
#ifndef MS_WINDOWS
    FILE *f;
    char *cpath;
    char cmode[10];
    size_t r;
    r = wcstombs(cmode, mode, 10);
    if (r == (size_t)-1 || r >= 10) {
        errno = EINVAL;
        return NULL;
    }
    cpath = _Py_wchar2char(path, NULL);
    if (cpath == NULL)
        return NULL;
    f = fopen(cpath, cmode);
    PyMem_Free(cpath);
    return f;
#else
    return _wfopen(path, mode);
#endif
}

/* Call _wfopen() on Windows, or encode the path to the filesystem encoding and
   call fopen() otherwise.

   Return the new file object on success, or NULL if the file cannot be open or
   (if PyErr_Occurred()) on unicode error */

FILE*
_Py_fopen(PyObject *path, const char *mode)
{
#ifdef MS_WINDOWS
    wchar_t wmode[10];
    int usize;

    usize = MultiByteToWideChar(CP_ACP, 0, mode, -1, wmode, sizeof(wmode));
    if (usize == 0)
        return NULL;

    return _wfopen(PyUnicode_AS_UNICODE(path), wmode);
#else
    FILE *f;
    PyObject *bytes = PyUnicode_EncodeFSDefault(path);
    if (bytes == NULL)
        return NULL;
    f = fopen(PyBytes_AS_STRING(bytes), mode);
    Py_DECREF(bytes);
    return f;
#endif
}

#ifdef HAVE_READLINK

/* Read value of symbolic link. Encode the path to the locale encoding, decode
   the result from the locale encoding. */

int
_Py_wreadlink(const wchar_t *path, wchar_t *buf, size_t bufsiz)
{
    char *cpath;
    char cbuf[PATH_MAX];
    wchar_t *wbuf;
    int res;
    size_t r1;

    cpath = _Py_wchar2char(path, NULL);
    if (cpath == NULL) {
        errno = EINVAL;
        return -1;
    }
    res = (int)readlink(cpath, cbuf, PATH_MAX);
    PyMem_Free(cpath);
    if (res == -1)
        return -1;
    if (res == PATH_MAX) {
        errno = EINVAL;
        return -1;
    }
    cbuf[res] = '\0'; /* buf will be null terminated */
    wbuf = _Py_char2wchar(cbuf, &r1);
    if (wbuf == NULL) {
        errno = EINVAL;
        return -1;
    }
    if (bufsiz <= r1) {
        PyMem_Free(wbuf);
        errno = EINVAL;
        return -1;
    }
    wcsncpy(buf, wbuf, bufsiz);
    PyMem_Free(wbuf);
    return (int)r1;
}
#endif

#ifdef HAVE_REALPATH

/* Return the canonicalized absolute pathname. Encode path to the locale
   encoding, decode the result from the locale encoding. */

wchar_t*
_Py_wrealpath(const wchar_t *path,
              wchar_t *resolved_path, size_t resolved_path_size)
{
    char *cpath;
    char cresolved_path[PATH_MAX];
    wchar_t *wresolved_path;
    char *res;
    size_t r;
    cpath = _Py_wchar2char(path, NULL);
    if (cpath == NULL) {
        errno = EINVAL;
        return NULL;
    }
    res = realpath(cpath, cresolved_path);
    PyMem_Free(cpath);
    if (res == NULL)
        return NULL;

    wresolved_path = _Py_char2wchar(cresolved_path, &r);
    if (wresolved_path == NULL) {
        errno = EINVAL;
        return NULL;
    }
    if (resolved_path_size <= r) {
        PyMem_Free(wresolved_path);
        errno = EINVAL;
        return NULL;
    }
    wcsncpy(resolved_path, wresolved_path, resolved_path_size);
    PyMem_Free(wresolved_path);
    return resolved_path;
}
#endif

/* Get the current directory. size is the buffer size in wide characters
   including the null character. Decode the path from the locale encoding. */

wchar_t*
_Py_wgetcwd(wchar_t *buf, size_t size)
{
#ifdef MS_WINDOWS
    return _wgetcwd(buf, size);
#else
    char fname[PATH_MAX];
    wchar_t *wname;
    size_t len;

    if (getcwd(fname, PATH_MAX) == NULL)
        return NULL;
    wname = _Py_char2wchar(fname, &len);
    if (wname == NULL)
        return NULL;
    if (size <= len) {
        PyMem_Free(wname);
        return NULL;
    }
    wcsncpy(buf, wname, size);
    PyMem_Free(wname);
    return buf;
#endif
}

#endif