2 * Copyright (c) 2002 John Rochester
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer,
10 * in this position and unchanged.
11 * 2. Redistributions in binary form must reproduce the above copyright
12 * notice, this list of conditions and the following disclaimer in the
13 * documentation and/or other materials provided with the distribution.
14 * 3. The name of the author may not be used to endorse or promote products
15 * derived from this software without specific prior written permission
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
18 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
19 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
20 * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
21 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
22 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
23 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
24 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
25 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
26 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
28 * $FreeBSD: src/usr.bin/catman/catman.c,v 1.9 2003/06/10 02:18:00 ache Exp $
29 * $DragonFly: src/usr.bin/catman/catman.c,v 1.4 2008/11/10 16:10:34 swildner Exp $
32 #include <sys/types.h>
34 #include <sys/param.h>
48 #define DEFAULT_MANPATH "/usr/share/man"
50 #define TOP_LEVEL_DIR 0 /* signifies a top-level man directory */
51 #define MAN_SECTION_DIR 1 /* signifies a man section directory */
52 #define UNKNOWN 2 /* signifies an unclassifiable directory */
54 #define TEST_EXISTS 0x01
56 #define TEST_FILE 0x04
57 #define TEST_READABLE 0x08
58 #define TEST_WRITABLE 0x10
59 #define TEST_EXECUTABLE 0x20
61 static int verbose
; /* -v flag: be verbose with warnings */
62 static int pretend
; /* -n, -p flags: print out what would be done
63 instead of actually doing it */
64 static int force
; /* -f flag: force overwriting all cat pages */
65 static int rm_junk
; /* -r flag: remove garbage pages */
66 static char *locale
; /* user's locale if -L is used */
67 static char *lang_locale
; /* short form of locale */
68 static const char *machine
;
69 static int exit_code
; /* exit code to use when finished */
72 * -T argument for nroff
74 static const char *nroff_device
= "ascii";
77 * Mapping from locale to nroff device
79 static const char *locale_device
[] = {
81 "ISO8859-1", "latin1",
82 "ISO8859-15", "latin1",
86 #define BZ2_CMD "bzip2"
87 #define BZ2_EXT ".bz2"
88 #define BZ2CAT_CMD "bz"
92 enum Ziptype
{NONE
, BZIP
, GZIP
};
95 static gid_t gids
[NGROUPS_MAX
];
97 static int starting_dir
;
98 static char tmp_file
[MAXPATHLEN
];
102 * A hashtable is an array of chains composed of this entry structure.
108 struct hash_entry
*next
;
111 #define HASHTABLE_ALLOC 16384 /* allocation for hashtable (power of 2) */
112 #define HASH_MASK (HASHTABLE_ALLOC - 1)
114 static struct hash_entry
*visited
[HASHTABLE_ALLOC
];
115 static struct hash_entry
*links
[HASHTABLE_ALLOC
];
118 * Inserts a string into a hashtable keyed by inode & device number.
121 insert_hashtable(struct hash_entry
**table
, ino_t inode_number
,
122 dev_t device_number
, const char *data
)
124 struct hash_entry
*new_entry
;
125 struct hash_entry
**chain
;
127 new_entry
= malloc(sizeof(struct hash_entry
));
128 if (new_entry
== NULL
)
129 err(1, "can't insert into hashtable");
130 chain
= &table
[inode_number
& HASH_MASK
];
131 new_entry
->inode_number
= inode_number
;
132 new_entry
->device_number
= device_number
;
133 new_entry
->data
= data
;
134 new_entry
->next
= *chain
;
139 * Finds a string in a hashtable keyed by inode & device number.
142 find_hashtable(struct hash_entry
**table
, ino_t inode_number
,
145 struct hash_entry
*chain
;
147 chain
= table
[inode_number
& HASH_MASK
];
148 while (chain
!= NULL
) {
149 if (chain
->inode_number
== inode_number
&&
150 chain
->device_number
== device_number
)
158 trap_signal(int sig __unused
)
160 if (tmp_file
[0] != '\0')
166 * Deals with junk files in the man or cat section directories.
169 junk(const char *mandir
, const char *name
, const char *reason
)
172 fprintf(stderr
, "%s/%s: %s\n", mandir
, name
, reason
);
174 fprintf(stderr
, "rm %s/%s\n", mandir
, name
);
175 if (!pretend
&& unlink(name
) < 0)
176 warn("%s/%s", mandir
, name
);
181 * Returns TOP_LEVEL_DIR for .../man, MAN_SECTION_DIR for .../manXXX,
182 * and UNKNOWN for everything else.
185 directory_type(char *dir
)
190 p
= strrchr(dir
, '/');
191 if (p
== NULL
|| p
[1] != '\0')
199 if (strncmp(p
, "man", 3) == 0) {
202 return TOP_LEVEL_DIR
;
203 while (isalnum((unsigned char)*p
) || *p
== '_') {
205 return(MAN_SECTION_DIR
);
212 * Tests whether the given file name (without a preceding path)
213 * is a proper man page name (like "mk-amd-map.8.gz").
214 * Only alphanumerics and '_' are allowed after the last '.' and
215 * the last '.' can't be the first or last characters.
218 is_manpage_name(char *name
)
220 char *lastdot
= NULL
;
223 for (n
= name
; *n
!= '\0'; n
++) {
242 return(lastdot
> name
&& lastdot
+ 1 < n
);
246 is_bzipped(char *name
)
248 int len
= strlen(name
);
249 return(len
>= 5 && strcmp(&name
[len
- 4], BZ2_EXT
) == 0);
253 is_gzipped(char *name
)
255 int len
= strlen(name
);
256 return(len
>= 4 && strcmp(&name
[len
- 3], GZ_EXT
) == 0);
260 * Converts manXXX to catXXX.
263 get_cat_section(char *section
)
267 cat_section
= strdup(section
);
268 strncpy(cat_section
, "cat", 3);
273 * Tests to see if the given directory has already been visited.
276 already_visited(char *mandir
, char *dir
, int count_visit
)
280 if (stat(dir
, &st
) < 0) {
282 warn("%s/%s", mandir
, dir
);
288 if (find_hashtable(visited
, st
.st_ino
, st
.st_dev
) != NULL
) {
290 warnx("already visited %s/%s", mandir
, dir
);
292 warnx("already visited %s", dir
);
296 insert_hashtable(visited
, st
.st_ino
, st
.st_dev
, "");
301 * Returns a set of TEST_* bits describing a file's type and permissions.
302 * If mod_time isn't NULL, it will contain the file's modification time.
305 test_path(char *name
, time_t *mod_time
)
309 if (stat(name
, &test_st
) < 0)
311 result
= TEST_EXISTS
;
312 if (mod_time
!= NULL
)
313 *mod_time
= test_st
.st_mtime
;
314 if (S_ISDIR(test_st
.st_mode
))
316 else if (S_ISREG(test_st
.st_mode
))
318 if (test_st
.st_uid
== uid
) {
319 test_st
.st_mode
>>= 6;
322 for (i
= 0; i
< ngids
; i
++) {
323 if (test_st
.st_gid
== gids
[i
]) {
324 test_st
.st_mode
>>= 3;
329 if (test_st
.st_mode
& S_IROTH
)
330 result
|= TEST_READABLE
;
331 if (test_st
.st_mode
& S_IWOTH
)
332 result
|= TEST_WRITABLE
;
333 if (test_st
.st_mode
& S_IXOTH
)
334 result
|= TEST_EXECUTABLE
;
339 * Checks whether a file is a symbolic link.
342 is_symlink(char *path
)
346 return(lstat(path
, &st
) >= 0 && S_ISLNK(st
.st_mode
));
350 * Tests to see if the given directory can be written to.
353 check_writable(char *mandir
)
355 if (verbose
&& !(test_path(mandir
, NULL
) & TEST_WRITABLE
))
356 fprintf(stderr
, "%s: not writable - will only be able to write "
357 "to existing cat directories\n", mandir
);
361 * If the directory exists, attempt to make it writable, otherwise
362 * attempt to create it.
365 make_writable_dir(char *mandir
, char *dir
)
369 if ((test
= test_path(dir
, NULL
)) != 0) {
370 if (!(test
& TEST_WRITABLE
) && chmod(dir
, 0755) < 0) {
371 warn("%s/%s: chmod", mandir
, dir
);
376 if (verbose
|| pretend
)
377 fprintf(stderr
, "mkdir %s\n", dir
);
380 if (mkdir(dir
, 0755) < 0) {
381 warn("%s/%s: mkdir", mandir
, dir
);
391 * Processes a single man page source by using nroff to create
392 * the preformatted cat page.
395 process_page(char *mandir
, char *src
, char *cat
, enum Ziptype zipped
)
397 int src_test
, cat_test
;
398 time_t src_mtime
, cat_mtime
;
399 char cmd
[MAXPATHLEN
];
402 const char *link_name
;
404 src_test
= test_path(src
, &src_mtime
);
405 if (!(src_test
& (TEST_FILE
|TEST_READABLE
))) {
406 if (!(src_test
& TEST_DIR
)) {
407 warnx("%s/%s: unreadable", mandir
, src
);
409 if (rm_junk
&& is_symlink(src
))
410 junk(mandir
, src
, "bogus symlink");
414 src_dev
= test_st
.st_dev
;
415 src_ino
= test_st
.st_ino
;
416 cat_test
= test_path(cat
, &cat_mtime
);
417 if (cat_test
& (TEST_FILE
|TEST_READABLE
)) {
418 if (!force
&& cat_mtime
>= src_mtime
) {
420 fprintf(stderr
, "\t%s/%s: up to date\n",
426 * Is the man page a link to one we've already processed?
428 if ((link_name
= find_hashtable(links
, src_ino
, src_dev
)) != NULL
) {
429 if (verbose
|| pretend
)
430 fprintf(stderr
, "%slink %s -> %s\n",
431 verbose
? "\t" : "", cat
, link_name
);
433 link(link_name
, cat
);
436 insert_hashtable(links
, src_ino
, src_dev
, strdup(cat
));
437 if (verbose
|| pretend
) {
438 fprintf(stderr
, "%sformat %s -> %s\n",
439 verbose
? "\t" : "", src
, cat
);
443 snprintf(tmp_file
, sizeof tmp_file
, "%s.tmp", cat
);
444 snprintf(cmd
, sizeof cmd
,
445 "%scat %s | tbl | nroff -T%s -man | col | %s > %s.tmp",
446 zipped
== BZIP
? BZ2CAT_CMD
: zipped
== GZIP
? GZCAT_CMD
: "",
448 zipped
== BZIP
? BZ2_CMD
: zipped
== GZIP
? GZ_CMD
: "cat",
450 if (system(cmd
) != 0)
451 err(1, "formatting pipeline");
452 if (rename(tmp_file
, cat
) < 0)
458 * Scan the man section directory for pages and process each one,
459 * then check for junk in the corresponding cat section.
462 scan_section(char *mandir
, char *section
, char *cat_section
)
464 struct dirent
**entries
;
465 char **expected
= NULL
;
471 char page_path
[MAXPATHLEN
];
472 char cat_path
[MAXPATHLEN
];
473 char zip_path
[MAXPATHLEN
];
476 * scan the man section directory for pages
478 npages
= scandir(section
, &entries
, NULL
, alphasort
);
480 warn("%s/%s", mandir
, section
);
484 if (verbose
|| rm_junk
) {
486 * Maintain a list of all cat pages that should exist,
487 * corresponding to existing man pages.
489 expected
= (char **) calloc(npages
, sizeof(char *));
491 for (i
= 0; i
< npages
; free(entries
[i
++])) {
492 page_name
= entries
[i
]->d_name
;
493 snprintf(page_path
, sizeof page_path
, "%s/%s", section
,
495 if (!is_manpage_name(page_name
)) {
496 if (!(test_path(page_path
, NULL
) & TEST_DIR
)) {
497 junk(mandir
, page_path
,
498 "invalid man page name");
502 zipped
= is_bzipped(page_name
) ? BZIP
:
503 is_gzipped(page_name
) ? GZIP
: NONE
;
504 if (zipped
!= NONE
) {
505 snprintf(cat_path
, sizeof cat_path
, "%s/%s",
506 cat_section
, page_name
);
507 if (expected
!= NULL
)
508 expected
[nexpected
++] = strdup(page_name
);
509 process_page(mandir
, page_path
, cat_path
, zipped
);
512 * We've got an uncompressed man page,
513 * check to see if there's a (preferred)
516 snprintf(zip_path
, sizeof zip_path
, "%s%s",
518 if (test_path(zip_path
, NULL
) != 0) {
519 junk(mandir
, page_path
,
520 "man page unused due to existing " GZ_EXT
);
524 "warning, %s is uncompressed\n",
527 snprintf(cat_path
, sizeof cat_path
, "%s/%s",
528 cat_section
, page_name
);
529 if (expected
!= NULL
) {
530 asprintf(&expected
[nexpected
++],
533 process_page(mandir
, page_path
, cat_path
, NONE
);
538 if (expected
== NULL
)
541 * scan cat sections for junk
543 npages
= scandir(cat_section
, &entries
, NULL
, alphasort
);
545 for (i
= 0; i
< npages
; free(entries
[i
++])) {
546 const char *junk_reason
;
549 page_name
= entries
[i
]->d_name
;
550 if (strcmp(page_name
, ".") == 0 || strcmp(page_name
, "..") == 0)
553 * Keep the index into the expected cat page list
554 * ahead of the name we've found.
556 while (e
< nexpected
&&
557 (cmp
= strcmp(page_name
, expected
[e
])) > 0)
561 /* we have an unexpected page */
562 snprintf(cat_path
, sizeof cat_path
, "%s/%s", cat_section
,
564 if (!is_manpage_name(page_name
)) {
565 if (test_path(cat_path
, NULL
) & TEST_DIR
)
567 junk_reason
= "invalid cat page name";
568 } else if (!is_gzipped(page_name
) && e
+ 1 < nexpected
&&
569 strncmp(page_name
, expected
[e
+ 1], strlen(page_name
)) == 0 &&
570 strlen(expected
[e
+ 1]) == strlen(page_name
) + 3) {
571 junk_reason
= "cat page unused due to existing " GZ_EXT
;
573 junk_reason
= "cat page without man page";
574 junk(mandir
, cat_path
, junk_reason
);
577 while (e
< nexpected
)
584 * Processes a single man section.
587 process_section(char *mandir
, char *section
)
591 if (already_visited(mandir
, section
, 1))
594 fprintf(stderr
, " section %s\n", section
);
595 cat_section
= get_cat_section(section
);
596 if (make_writable_dir(mandir
, cat_section
))
597 scan_section(mandir
, section
, cat_section
);
602 select_sections(struct dirent
*entry
)
604 return(directory_type(entry
->d_name
) == MAN_SECTION_DIR
);
608 * Processes a single top-level man directory. If section isn't NULL,
609 * it will only process that section sub-directory, otherwise it will
610 * process all of them.
613 process_mandir(char *dir_name
, char *section
)
615 fchdir(starting_dir
);
616 if (already_visited(NULL
, dir_name
, section
== NULL
))
618 check_writable(dir_name
);
620 fprintf(stderr
, "man directory %s\n", dir_name
);
622 fprintf(stderr
, "cd %s\n", dir_name
);
623 if (chdir(dir_name
) < 0) {
624 warn("%s: chdir", dir_name
);
628 if (section
!= NULL
) {
629 process_section(dir_name
, section
);
631 struct dirent
**entries
;
636 nsections
= scandir(".", &entries
, select_sections
, alphasort
);
638 warn("%s", dir_name
);
642 for (i
= 0; i
< nsections
; i
++) {
643 process_section(dir_name
, entries
[i
]->d_name
);
644 asprintf(&machine_dir
, "%s/%s", entries
[i
]->d_name
,
646 if (test_path(machine_dir
, NULL
) & TEST_DIR
)
647 process_section(dir_name
, machine_dir
);
656 * Processes one argument, which may be a colon-separated list of
660 process_argument(const char *arg
)
669 err(1, "out of memory");
670 while ((dir
= strsep(&parg
, ":")) != NULL
) {
671 switch (directory_type(dir
)) {
673 if (locale
!= NULL
) {
674 asprintf(&mandir
, "%s/%s", dir
, locale
);
675 process_mandir(mandir
, NULL
);
677 if (lang_locale
!= NULL
) {
678 asprintf(&mandir
, "%s/%s", dir
,
680 process_mandir(mandir
, NULL
);
684 process_mandir(dir
, NULL
);
687 case MAN_SECTION_DIR
: {
688 mandir
= strdup(dirname(dir
));
689 section
= strdup(basename(dir
));
690 process_mandir(mandir
, section
);
696 warnx("%s: directory name not in proper man form", dir
);
704 determine_locale(void)
708 if ((locale
= setlocale(LC_CTYPE
, "")) == NULL
) {
709 warnx("-L option used, but no locale found\n");
712 sep
= strchr(locale
, '_');
713 if (sep
!= NULL
&& isupper(sep
[1]) && isupper(sep
[2]))
714 asprintf(&lang_locale
, "%.*s%s", (int)(sep
- locale
), locale
,
716 sep
= nl_langinfo(CODESET
);
717 if (sep
!= NULL
&& *sep
!= '\0' && strcmp(sep
, "US-ASCII") != 0) {
720 for (i
= 0; locale_device
[i
] != NULL
; i
+= 2) {
721 if (strcmp(sep
, locale_device
[i
]) == 0) {
722 nroff_device
= locale_device
[i
+ 1];
728 if (lang_locale
!= NULL
)
729 fprintf(stderr
, "short locale is %s\n", lang_locale
);
730 fprintf(stderr
, "nroff device is %s\n", nroff_device
);
737 fprintf(stderr
, "usage: %s [-fLnrv] [directories ...]\n",
743 main(int argc
, char **argv
)
747 if ((uid
= getuid()) == 0) {
748 fprintf(stderr
, "don't run %s as root, use:\n echo", argv
[0]);
749 for (optind
= 0; optind
< argc
; optind
++)
750 fprintf(stderr
, " %s", argv
[optind
]);
751 fprintf(stderr
, " | nice -5 su -m man\n");
754 while ((opt
= getopt(argc
, argv
, "vnfLrh")) != -1) {
776 ngids
= getgroups(NGROUPS_MAX
, gids
);
777 if ((starting_dir
= open(".", 0)) < 0)
780 signal(SIGINT
, trap_signal
);
781 signal(SIGHUP
, trap_signal
);
782 signal(SIGQUIT
, trap_signal
);
783 signal(SIGTERM
, trap_signal
);
785 if ((machine
= getenv("MACHINE")) == NULL
)
788 if (optind
== argc
) {
789 const char *manpath
= getenv("MANPATH");
791 manpath
= DEFAULT_MANPATH
;
792 process_argument(manpath
);
794 while (optind
< argc
)
795 process_argument(argv
[optind
++]);