1 /* Read, sort and compare two directories. Used for GNU DIFF.
3 Copyright (C) 1988-1989, 1992-1995, 1998, 2001-2002, 2004, 2006-2007,
4 2009-2013 Free Software Foundation, Inc.
6 This file is part of GNU DIFF.
8 This program is free software: you can redistribute it and/or modify
9 it under the terms of the GNU General Public License as published by
10 the Free Software Foundation, either version 3 of the License, or
11 (at your option) any later version.
13 This program is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with this program. If not, see <http://www.gnu.org/licenses/>. */
24 #include <filenamecat.h>
28 /* Read the directory named by DIR and store into DIRDATA a sorted vector
29 of filenames for its contents. DIR->desc == -1 means this directory is
30 known to be nonexistent, so set DIRDATA to an empty vector.
31 Return -1 (setting errno) if error, 0 otherwise. */
35 size_t nnames
; /* Number of names. */
36 char const **names
; /* Sorted names of files in dir, followed by 0. */
37 char *data
; /* Allocated storage for file names. */
40 /* Whether file names in directories should be compared with
41 locale-specific sorting. */
42 static bool locale_specific_sorting
;
44 /* Where to go if locale-specific sorting fails. */
45 static jmp_buf failed_locale_specific_sorting
;
47 static bool dir_loop (struct comparison
const *, int);
48 static int compare_names_for_qsort (void const *, void const *);
51 /* Read a directory and get its vector of names. */
54 dir_read (struct file_data
const *dir
, struct dirdata
*dirdata
)
56 register struct dirent
*next
;
59 /* Address of block containing the files that are described. */
62 /* Number of files in directory. */
65 /* Allocated and used storage for file name data. */
67 size_t data_alloc
, data_used
;
76 /* Open the directory and check for errors. */
77 register DIR *reading
= opendir (dir
->name
);
81 /* Initialize the table of filenames. */
85 dirdata
->data
= data
= xmalloc (data_alloc
);
87 /* Read the directory entries, and insert the subfiles
88 into the 'data' table. */
90 while ((errno
= 0, (next
= readdir (reading
)) != 0))
92 char *d_name
= next
->d_name
;
93 size_t d_size
= _D_EXACT_NAMLEN (next
) + 1;
95 /* Ignore "." and "..". */
97 && (d_name
[1] == 0 || (d_name
[1] == '.' && d_name
[2] == 0)))
100 if (excluded_file_name (excluded
, d_name
))
103 while (data_alloc
< data_used
+ d_size
)
105 if (PTRDIFF_MAX
/ 2 <= data_alloc
)
107 dirdata
->data
= data
= xrealloc (data
, data_alloc
*= 2);
110 memcpy (data
+ data_used
, d_name
, d_size
);
124 if (closedir (reading
) != 0)
129 /* Create the 'names' table from the 'data' table. */
130 if (PTRDIFF_MAX
/ sizeof *names
- 1 <= nnames
)
132 dirdata
->names
= names
= xmalloc ((nnames
+ 1) * sizeof *names
);
133 dirdata
->nnames
= nnames
;
134 for (i
= 0; i
< nnames
; i
++)
137 data
+= strlen (data
) + 1;
143 /* Compare file names, returning a value compatible with strcmp. */
146 compare_names (char const *name1
, char const *name2
)
148 if (locale_specific_sorting
)
152 if (ignore_file_name_case
)
153 r
= strcasecoll (name1
, name2
);
155 r
= strcoll (name1
, name2
);
158 error (0, errno
, _("cannot compare file names '%s' and '%s'"),
160 longjmp (failed_locale_specific_sorting
, 1);
165 return file_name_cmp (name1
, name2
);
168 /* Compare names FILE1 and FILE2 when sorting a directory.
169 Prefer filtered comparison, breaking ties with file_name_cmp. */
172 compare_names_for_qsort (void const *file1
, void const *file2
)
174 char const *const *f1
= file1
;
175 char const *const *f2
= file2
;
176 int diff
= compare_names (*f1
, *f2
);
177 return diff
? diff
: file_name_cmp (*f1
, *f2
);
180 /* Compare the contents of two directories named in CMP.
181 This is a top-level routine; it does everything necessary for diff
184 CMP->file[0].desc == -1 says directory CMP->file[0] doesn't exist,
185 but pretend it is empty. Likewise for CMP->file[1].
187 HANDLE_FILE is a caller-provided subroutine called to handle each file.
188 It gets three operands: CMP, name of file in dir 0, name of file in dir 1.
189 These names are relative to the original working directory.
191 For a file that appears in only one of the dirs, one of the name-args
192 to HANDLE_FILE is zero.
194 Returns the maximum of all the values returned by HANDLE_FILE,
195 or EXIT_TROUBLE if trouble is encountered in opening files. */
198 diff_dirs (struct comparison
const *cmp
,
199 int (*handle_file
) (struct comparison
const *,
200 char const *, char const *))
202 struct dirdata dirdata
[2];
203 int volatile val
= EXIT_SUCCESS
;
206 if ((cmp
->file
[0].desc
== -1 || dir_loop (cmp
, 0))
207 && (cmp
->file
[1].desc
== -1 || dir_loop (cmp
, 1)))
209 error (0, 0, _("%s: recursive directory loop"),
210 cmp
->file
[cmp
->file
[0].desc
== -1].name
);
214 /* Get contents of both dirs. */
215 for (i
= 0; i
< 2; i
++)
216 if (! dir_read (&cmp
->file
[i
], &dirdata
[i
]))
218 perror_with_name (cmp
->file
[i
].name
);
222 if (val
== EXIT_SUCCESS
)
224 char const **volatile names
[2];
225 names
[0] = dirdata
[0].names
;
226 names
[1] = dirdata
[1].names
;
228 /* Use locale-specific sorting if possible, else native byte order. */
229 locale_specific_sorting
= true;
230 if (setjmp (failed_locale_specific_sorting
))
231 locale_specific_sorting
= false;
233 /* Sort the directories. */
234 for (i
= 0; i
< 2; i
++)
235 qsort (names
[i
], dirdata
[i
].nnames
, sizeof *dirdata
[i
].names
,
236 compare_names_for_qsort
);
238 /* If '-S name' was given, and this is the topmost level of comparison,
239 ignore all file names less than the specified starting name. */
241 if (starting_file
&& ! cmp
->parent
)
243 while (*names
[0] && compare_names (*names
[0], starting_file
) < 0)
245 while (*names
[1] && compare_names (*names
[1], starting_file
) < 0)
249 /* Loop while files remain in one or both dirs. */
250 while (*names
[0] || *names
[1])
252 /* Compare next name in dir 0 with next name in dir 1.
254 pretend the "next name" in that dir is very large. */
255 int nameorder
= (!*names
[0] ? 1 : !*names
[1] ? -1
256 : compare_names (*names
[0], *names
[1]));
258 /* Prefer a file_name_cmp match if available. This algorithm is
259 O(N**2), where N is the number of names in a directory
260 that compare_names says are all equal, but in practice N
261 is so small it's not worth tuning. */
264 int raw_order
= file_name_cmp (*names
[0], *names
[1]);
267 int greater_side
= raw_order
< 0;
268 int lesser_side
= 1 - greater_side
;
269 char const **lesser
= names
[lesser_side
];
270 char const *greater_name
= *names
[greater_side
];
274 *p
&& compare_names (*p
, greater_name
) == 0;
277 int c
= file_name_cmp (*p
, greater_name
);
282 memmove (lesser
+ 1, lesser
,
283 (char *) p
- (char *) lesser
);
284 *lesser
= greater_name
;
292 int v1
= (*handle_file
) (cmp
,
293 0 < nameorder
? 0 : *names
[0]++,
294 nameorder
< 0 ? 0 : *names
[1]++);
300 for (i
= 0; i
< 2; i
++)
302 free (dirdata
[i
].names
);
303 free (dirdata
[i
].data
);
309 /* Return nonzero if CMP is looping recursively in argument I. */
311 static bool _GL_ATTRIBUTE_PURE
312 dir_loop (struct comparison
const *cmp
, int i
)
314 struct comparison
const *p
= cmp
;
315 while ((p
= p
->parent
))
316 if (0 < same_file (&p
->file
[i
].stat
, &cmp
->file
[i
].stat
))
321 /* Find a matching filename in a directory. */
324 find_dir_file_pathname (char const *dir
, char const *file
)
326 /* The 'IF_LINT (volatile)' works around what appears to be a bug in
327 gcc 4.8.0 20120825; see
328 <http://lists.gnu.org/archive/html/bug-diffutils/2012-08/msg00007.html>.
330 char const * IF_LINT (volatile) match
= file
;
333 struct dirdata dirdata
;
334 dirdata
.names
= NULL
;
337 if (ignore_file_name_case
)
339 struct file_data filedata
;
343 if (dir_read (&filedata
, &dirdata
))
345 locale_specific_sorting
= true;
346 if (setjmp (failed_locale_specific_sorting
))
347 match
= file
; /* longjmp may mess up MATCH. */
350 for (char const **p
= dirdata
.names
; *p
; p
++)
351 if (compare_names (*p
, file
) == 0)
353 if (file_name_cmp (*p
, file
) == 0)
365 val
= file_name_concat (dir
, match
, NULL
);
366 free (dirdata
.names
);