1# Copyright (C) 2006, 2008, 2009, 2010 Canonical Ltd
2#
3# This program is free software; you can redistribute it and/or modify
4# it under the terms of the GNU General Public License as published by
5# the Free Software Foundation; either version 2 of the License, or
6# (at your option) any later version.
7#
8# This program is distributed in the hope that it will be useful,
9# but WITHOUT ANY WARRANTY; without even the implied warranty of
10# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
11# GNU General Public License for more details.
12#
13# You should have received a copy of the GNU General Public License
14# along with this program; if not, write to the Free Software
15# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
16#
17# cython: language_level=3
18
19"""Wrapper for readdir which returns files ordered by inode."""
20
21
22import os
23import sys
24
25cdef extern from "python-compat.h":
26    pass
27
28
29cdef extern from 'errno.h':
30    int ENOENT
31    int ENOTDIR
32    int EAGAIN
33    int EINTR
34    char *strerror(int errno)
35    # not necessarily a real variable, but this should be close enough
36    int errno
37
38cdef extern from 'unistd.h':
39    int chdir(char *path)
40    int close(int fd)
41    int fchdir(int fd)
42    char *getcwd(char *, int size)
43
44cdef extern from 'stdlib.h':
45    void *malloc(int)
46    void free(void *)
47
48
49cdef extern from 'sys/types.h':
50    ctypedef long ssize_t
51    ctypedef unsigned long size_t
52    ctypedef long time_t
53    ctypedef unsigned long ino_t
54    ctypedef unsigned long long off_t
55    ctypedef int mode_t
56
57
58cdef extern from 'sys/stat.h':
59    cdef struct stat:
60        int st_mode
61        off_t st_size
62        int st_dev
63        ino_t st_ino
64        int st_mtime
65        int st_ctime
66    int lstat(char *path, stat *buf)
67    int S_ISDIR(int mode)
68    int S_ISCHR(int mode)
69    int S_ISBLK(int mode)
70    int S_ISREG(int mode)
71    int S_ISFIFO(int mode)
72    int S_ISLNK(int mode)
73    int S_ISSOCK(int mode)
74
75
76cdef extern from 'fcntl.h':
77    int O_RDONLY
78    int open(char *pathname, int flags, mode_t mode)
79
80
81cdef extern from 'Python.h':
82    int PyErr_CheckSignals() except -1
83    char * PyBytes_AS_STRING(object)
84    ctypedef struct PyObject:
85        pass
86    Py_ssize_t PyBytes_Size(object s)
87    object PyList_GetItem(object lst, Py_ssize_t index)
88    void *PyList_GetItem_object_void "PyList_GET_ITEM" (object lst, int index)
89    int PyList_Append(object lst, object item) except -1
90    void *PyTuple_GetItem_void_void "PyTuple_GET_ITEM" (void* tpl, int index)
91    int PyTuple_SetItem(void *, Py_ssize_t pos, object item) except -1
92    int PyTuple_SetItem_obj "PyTuple_SetItem" (void *, Py_ssize_t pos, PyObject * item) except -1
93    void Py_INCREF(object o)
94    void Py_DECREF(object o)
95    void PyBytes_Concat(PyObject **string, object newpart)
96
97
98cdef extern from 'dirent.h':
99    ctypedef struct dirent:
100        char d_name[256]
101        ino_t d_ino
102    # the opaque C library DIR type.
103    ctypedef struct DIR
104    # should be DIR *, pyrex barfs.
105    DIR * opendir(char * name)
106    int closedir(DIR * dir)
107    dirent *readdir(DIR *dir)
108
109cdef object _directory
110_directory = 'directory'
111cdef object _chardev
112_chardev = 'chardev'
113cdef object _block
114_block = 'block'
115cdef object _file
116_file = 'file'
117cdef object _fifo
118_fifo = 'fifo'
119cdef object _symlink
120_symlink = 'symlink'
121cdef object _socket
122_socket = 'socket'
123cdef object _unknown
124_unknown = 'unknown'
125
126# add a typedef struct dirent dirent to workaround pyrex
127cdef extern from 'readdir.h':
128    pass
129
130
131cdef class _Stat:
132    """Represent a 'stat' result."""
133
134    cdef stat _st
135
136    property st_dev:
137        def __get__(self):
138            return self._st.st_dev
139
140    property st_ino:
141        def __get__(self):
142            return self._st.st_ino
143
144    property st_mode:
145        def __get__(self):
146            return self._st.st_mode
147
148    property st_ctime:
149        def __get__(self):
150            return self._st.st_ctime
151
152    property st_mtime:
153        def __get__(self):
154            return self._st.st_mtime
155
156    property st_size:
157        def __get__(self):
158            return self._st.st_size
159
160    def __repr__(self):
161        """Repr is the same as a Stat object.
162
163        (mode, ino, dev, nlink, uid, gid, size, None(atime), mtime, ctime)
164        """
165        return repr((self.st_mode, 0, 0, 0, 0, 0, self.st_size, None,
166                     self.st_mtime, self.st_ctime))
167
168
169from . import osutils
170
171cdef object _safe_utf8
172_safe_utf8 = osutils.safe_utf8
173
174cdef class UTF8DirReader:
175    """A dir reader for utf8 file systems."""
176
177    def kind_from_mode(self, int mode):
178        """Get the kind of a path from a mode status."""
179        return self._kind_from_mode(mode)
180
181    cdef _kind_from_mode(self, int mode):
182        # Files and directories are the most common - check them first.
183        if S_ISREG(mode):
184            return _file
185        if S_ISDIR(mode):
186            return _directory
187        if S_ISCHR(mode):
188            return _chardev
189        if S_ISBLK(mode):
190            return _block
191        if S_ISLNK(mode):
192            return _symlink
193        if S_ISFIFO(mode):
194            return _fifo
195        if S_ISSOCK(mode):
196            return _socket
197        return _unknown
198
199    def top_prefix_to_starting_dir(self, top, prefix=""):
200        """See DirReader.top_prefix_to_starting_dir."""
201        return (_safe_utf8(prefix), None, None, None, _safe_utf8(top))
202
203    def read_dir(self, prefix, top):
204        """Read a single directory from a utf8 file system.
205
206        All paths in and out are utf8.
207
208        This sub-function is called when we know the filesystem is already in utf8
209        encoding. So we don't need to transcode filenames.
210
211        See DirReader.read_dir for details.
212        """
213        #cdef char *_prefix = prefix
214        #cdef char *_top = top
215        # Use C accelerated directory listing.
216        cdef object newval
217        cdef int index
218        cdef int length
219        cdef void * atuple
220        cdef object name
221        cdef PyObject * new_val_obj
222
223        if PyBytes_Size(prefix):
224            relprefix = prefix + b'/'
225        else:
226            relprefix = b''
227        top_slash = top + b'/'
228
229        # read_dir supplies in should-stat order.
230        # for _, name in sorted(_listdir(top)):
231        result = _read_dir(top)
232        length = len(result)
233        # result.sort()
234        for index from 0 <= index < length:
235            atuple = PyList_GetItem_object_void(result, index)
236            name = <object>PyTuple_GetItem_void_void(atuple, 1)
237            # We have a tuple with (inode, name, None, statvalue, None)
238            # Now edit it:
239            # inode -> path_from_top
240            # direct concat - faster than operator +.
241            new_val_obj = <PyObject *>relprefix
242            Py_INCREF(relprefix)
243            PyBytes_Concat(&new_val_obj, name)
244            if NULL == new_val_obj:
245                # PyBytes_Concat will have setup an exception, but how to get
246                # at it?
247                raise Exception("failed to strcat")
248            PyTuple_SetItem_obj(atuple, 0, new_val_obj)
249            # 1st None -> kind
250            newval = self._kind_from_mode(
251                (<_Stat>PyTuple_GetItem_void_void(atuple, 3)).st_mode)
252            Py_INCREF(newval)
253            PyTuple_SetItem(atuple, 2, newval)
254            # 2nd None -> abspath # for all - the caller may need to stat files
255            # etc.
256            # direct concat - faster than operator +.
257            new_val_obj = <PyObject *>top_slash
258            Py_INCREF(top_slash)
259            PyBytes_Concat(&new_val_obj, name)
260            if NULL == new_val_obj:
261                # PyBytes_Concat will have setup an exception, but how to get
262                # at it?
263                raise Exception("failed to strcat")
264            PyTuple_SetItem_obj(atuple, 4, new_val_obj)
265        return result
266
267
268cdef raise_os_error(int errnum, char *msg_prefix, path):
269    if errnum == EINTR:
270        PyErr_CheckSignals()
271    raise OSError(errnum, msg_prefix + strerror(errnum), path)
272
273
274cdef _read_dir(path):
275    """Like os.listdir, this reads the contents of a directory.
276
277    :param path: the directory to list.
278    :return: a list of single-owner (the list) tuples ready for editing into
279        the result tuples walkdirs needs to yield. They contain (inode, name,
280        None, statvalue, None).
281    """
282    cdef DIR *the_dir
283    # currently this needs a fixup - the C code says 'dirent' but should say
284    # 'struct dirent'
285    cdef dirent * entry
286    cdef dirent sentinel
287    cdef char *name
288    cdef int stat_result
289    cdef _Stat statvalue
290    global errno
291    cdef int orig_dir_fd
292
293    # Avoid chdir('') because it causes problems on Sun OS, and avoid this if
294    # staying in .
295    if path != b"" and path != b'.':
296        # we change into the requested directory before reading, and back at the
297        # end, because that turns out to make the stat calls measurably faster than
298        # passing full paths every time.
299        orig_dir_fd = open(".", O_RDONLY, 0)
300        if orig_dir_fd == -1:
301            raise_os_error(errno, "open: ", ".")
302        if -1 == chdir(path):
303            # Ignore the return value, because we are already raising an
304            # exception
305            close(orig_dir_fd)
306            raise_os_error(errno, "chdir: ", path)
307    else:
308        orig_dir_fd = -1
309
310    try:
311        the_dir = opendir(b".")
312        if NULL == the_dir:
313            raise_os_error(errno, "opendir: ", path)
314        try:
315            result = []
316            entry = &sentinel
317            while entry != NULL:
318                # Unlike most libc functions, readdir needs errno set to 0
319                # beforehand so that eof can be distinguished from errors.  See
320                # <https://bugs.launchpad.net/bzr/+bug/279381>
321                while True:
322                    errno = 0
323                    entry = readdir(the_dir)
324                    if entry == NULL and (errno == EAGAIN or errno == EINTR):
325                        if errno == EINTR:
326                            PyErr_CheckSignals()
327                        # try again
328                        continue
329                    else:
330                        break
331                if entry == NULL:
332                    if errno == ENOTDIR or errno == 0:
333                        # We see ENOTDIR at the end of a normal directory.
334                        # As ENOTDIR for read_dir(file) is triggered on opendir,
335                        # we consider ENOTDIR to be 'no error'.
336                        continue
337                    else:
338                        raise_os_error(errno, "readdir: ", path)
339                name = entry.d_name
340                if not (name[0] == c"." and (
341                    (name[1] == 0) or
342                    (name[1] == c"." and name[2] == 0))
343                    ):
344                    statvalue = _Stat()
345                    stat_result = lstat(entry.d_name, &statvalue._st)
346                    if stat_result != 0:
347                        if errno != ENOENT:
348                            raise_os_error(errno, "lstat: ",
349                                path + b"/" + entry.d_name)
350                        else:
351                            # the file seems to have disappeared after being
352                            # seen by readdir - perhaps a transient temporary
353                            # file.  there's no point returning it.
354                            continue
355                    # We append a 5-tuple that can be modified in-place by the C
356                    # api:
357                    # inode to sort on (to replace with top_path)
358                    # name (to keep)
359                    # kind (None, to set)
360                    # statvalue (to keep)
361                    # abspath (None, to set)
362                    PyList_Append(result, (entry.d_ino, entry.d_name, None,
363                        statvalue, None))
364        finally:
365            if -1 == closedir(the_dir):
366                raise_os_error(errno, "closedir: ", path)
367    finally:
368        if -1 != orig_dir_fd:
369            failed = False
370            if -1 == fchdir(orig_dir_fd):
371                # try to close the original directory anyhow
372                failed = True
373            if -1 == close(orig_dir_fd) or failed:
374                raise_os_error(errno, "return to orig_dir: ", "")
375
376    return result
377
378
379# vim: tw=79 ai expandtab sw=4 sts=4
380