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.
29 #include <sys/cdefs.h>
30 __FBSDID("$FreeBSD$");
32 #include <sys/types.h>
34 #include <sys/param.h>
35 #include <sys/utsname.h>
51 #define DEFAULT_MANPATH "/usr/share/man"
53 #define TOP_LEVEL_DIR 0 /* signifies a top-level man directory */
54 #define MAN_SECTION_DIR 1 /* signifies a man section directory */
55 #define UNKNOWN 2 /* signifies an unclassifiable directory */
57 #define TEST_EXISTS 0x01
59 #define TEST_FILE 0x04
60 #define TEST_READABLE 0x08
61 #define TEST_WRITABLE 0x10
63 static int verbose
; /* -v flag: be verbose with warnings */
64 static int pretend
; /* -n, -p flags: print out what would be done
65 instead of actually doing it */
66 static int force
; /* -f flag: force overwriting all cat pages */
67 static int rm_junk
; /* -r flag: remove garbage pages */
68 static char *locale
; /* user's locale if -L is used */
69 static char *lang_locale
; /* short form of locale */
70 static const char *machine
, *machine_arch
;
71 static int exit_code
; /* exit code to use when finished */
74 * -T argument for nroff
76 static const char *nroff_device
= "ascii";
79 * Mapping from locale to nroff device
81 static const char *locale_device
[] = {
83 "ISO8859-1", "latin1",
84 "ISO8859-15", "latin1",
88 #define BZ2_CMD "bzip2"
89 #define BZ2_EXT ".bz2"
90 #define BZ2CAT_CMD "bz"
94 enum Ziptype
{NONE
, BZIP
, GZIP
};
97 static int starting_dir
;
98 static char tmp_file
[MAXPATHLEN
];
99 static struct stat test_st
;
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
,
126 struct hash_entry
*new_entry
;
127 struct hash_entry
**chain
;
129 new_entry
= (struct hash_entry
*) malloc(sizeof(struct hash_entry
));
130 if (new_entry
== NULL
)
131 err(1, "can't insert into hashtable");
132 chain
= &table
[inode_number
& HASH_MASK
];
133 new_entry
->inode_number
= inode_number
;
134 new_entry
->device_number
= device_number
;
135 new_entry
->data
= data
;
136 new_entry
->next
= *chain
;
141 * Finds a string in a hashtable keyed by inode & device number.
144 find_hashtable(struct hash_entry
**table
,
148 struct hash_entry
*chain
;
150 chain
= table
[inode_number
& HASH_MASK
];
151 while (chain
!= NULL
) {
152 if (chain
->inode_number
== inode_number
&&
153 chain
->device_number
== device_number
)
161 trap_signal(int sig __unused
)
163 if (tmp_file
[0] != '\0')
169 * Deals with junk files in the man or cat section directories.
172 junk(const char *mandir
, const char *name
, const char *reason
)
175 fprintf(stderr
, "%s/%s: %s\n", mandir
, name
, reason
);
177 fprintf(stderr
, "rm %s/%s\n", mandir
, name
);
178 if (!pretend
&& unlink(name
) < 0)
179 warn("%s/%s", mandir
, name
);
184 * Returns TOP_LEVEL_DIR for .../man, MAN_SECTION_DIR for .../manXXX,
185 * and UNKNOWN for everything else.
188 directory_type(char *dir
)
193 p
= strrchr(dir
, '/');
194 if (p
== NULL
|| p
[1] != '\0')
202 if (strncmp(p
, "man", 3) == 0) {
205 return TOP_LEVEL_DIR
;
206 while (isalnum((unsigned char)*p
) || *p
== '_') {
208 return MAN_SECTION_DIR
;
215 * Tests whether the given file name (without a preceding path)
216 * is a proper man page name (like "mk-amd-map.8.gz").
217 * Only alphanumerics and '_' are allowed after the last '.' and
218 * the last '.' can't be the first or last characters.
221 is_manpage_name(char *name
)
223 char *lastdot
= NULL
;
227 if (!isalnum((unsigned char)*n
)) {
246 return lastdot
> name
&& lastdot
+ 1 < n
;
250 is_bzipped(char *name
)
252 int len
= strlen(name
);
253 return len
>= 5 && strcmp(&name
[len
- 4], BZ2_EXT
) == 0;
257 is_gzipped(char *name
)
259 int len
= strlen(name
);
260 return len
>= 4 && strcmp(&name
[len
- 3], GZ_EXT
) == 0;
264 * Converts manXXX to catXXX.
267 get_cat_section(char *section
)
271 cat_section
= strdup(section
);
272 assert(strlen(section
) > 3 && strncmp(section
, "man", 3) == 0);
273 memcpy(cat_section
, "cat", 3);
278 * Tests to see if the given directory has already been visited.
281 already_visited(char *mandir
, char *dir
, int count_visit
)
285 if (stat(dir
, &st
) < 0) {
287 warn("%s/%s", mandir
, dir
);
293 if (find_hashtable(visited
, st
.st_ino
, st
.st_dev
) != NULL
) {
295 warnx("already visited %s/%s", mandir
, dir
);
297 warnx("already visited %s", dir
);
301 insert_hashtable(visited
, st
.st_ino
, st
.st_dev
, "");
306 * Returns a set of TEST_* bits describing a file's type and permissions.
307 * If mod_time isn't NULL, it will contain the file's modification time.
310 test_path(char *name
, time_t *mod_time
)
314 if (stat(name
, &test_st
) < 0)
316 result
= TEST_EXISTS
;
317 if (mod_time
!= NULL
)
318 *mod_time
= test_st
.st_mtime
;
319 if (S_ISDIR(test_st
.st_mode
))
321 else if (S_ISREG(test_st
.st_mode
))
323 if (access(name
, R_OK
))
324 result
|= TEST_READABLE
;
325 if (access(name
, W_OK
))
326 result
|= TEST_WRITABLE
;
331 * Checks whether a file is a symbolic link.
334 is_symlink(char *path
)
338 return lstat(path
, &st
) >= 0 && S_ISLNK(st
.st_mode
);
342 * Tests to see if the given directory can be written to.
345 check_writable(char *mandir
)
347 if (verbose
&& !(test_path(mandir
, NULL
) & TEST_WRITABLE
))
348 fprintf(stderr
, "%s: not writable - will only be able to write to existing cat directories\n", mandir
);
352 * If the directory exists, attempt to make it writable, otherwise
353 * attempt to create it.
356 make_writable_dir(char *mandir
, char *dir
)
360 if ((test
= test_path(dir
, NULL
)) != 0) {
361 if (!(test
& TEST_WRITABLE
) && chmod(dir
, 0755) < 0) {
362 warn("%s/%s: chmod", mandir
, dir
);
367 if (verbose
|| pretend
)
368 fprintf(stderr
, "mkdir %s\n", dir
);
371 if (mkdir(dir
, 0755) < 0) {
372 warn("%s/%s: mkdir", mandir
, dir
);
382 * Processes a single man page source by using nroff to create
383 * the preformatted cat page.
386 process_page(char *mandir
, char *src
, char *cat
, enum Ziptype zipped
)
388 int src_test
, cat_test
;
389 time_t src_mtime
, cat_mtime
;
390 char cmd
[MAXPATHLEN
];
393 const char *link_name
;
395 src_test
= test_path(src
, &src_mtime
);
396 if (!(src_test
& (TEST_FILE
|TEST_READABLE
))) {
397 if (!(src_test
& TEST_DIR
)) {
398 warnx("%s/%s: unreadable", mandir
, src
);
400 if (rm_junk
&& is_symlink(src
))
401 junk(mandir
, src
, "bogus symlink");
405 src_dev
= test_st
.st_dev
;
406 src_ino
= test_st
.st_ino
;
407 cat_test
= test_path(cat
, &cat_mtime
);
408 if (cat_test
& (TEST_FILE
|TEST_READABLE
)) {
409 if (!force
&& cat_mtime
>= src_mtime
) {
411 fprintf(stderr
, "\t%s/%s: up to date\n",
418 * Is the man page a link to one we've already processed?
420 if ((link_name
= find_hashtable(links
, src_ino
, src_dev
)) != NULL
) {
421 if (verbose
|| pretend
) {
422 fprintf(stderr
, "%slink %s -> %s\n",
423 verbose
? "\t" : "", cat
, link_name
);
427 if (link(link_name
, cat
) < 0)
428 warn("%s %s: link", link_name
, cat
);
432 insert_hashtable(links
, src_ino
, src_dev
, strdup(cat
));
433 if (verbose
|| pretend
) {
434 fprintf(stderr
, "%sformat %s -> %s\n",
435 verbose
? "\t" : "", src
, cat
);
439 snprintf(tmp_file
, sizeof tmp_file
, "%s.tmp", cat
);
440 snprintf(cmd
, sizeof cmd
,
441 "%scat %s | tbl | nroff -c -T%s -man | %s > %s.tmp",
442 zipped
== BZIP
? BZ2CAT_CMD
: zipped
== GZIP
? GZCAT_CMD
: "",
444 zipped
== BZIP
? BZ2_CMD
: zipped
== GZIP
? GZ_CMD
: "cat",
446 if (system(cmd
) != 0)
447 err(1, "formatting pipeline");
448 if (rename(tmp_file
, cat
) < 0)
454 * Scan the man section directory for pages and process each one,
455 * then check for junk in the corresponding cat section.
458 scan_section(char *mandir
, char *section
, char *cat_section
)
460 struct dirent
**entries
;
461 char **expected
= NULL
;
467 char page_path
[MAXPATHLEN
];
468 char cat_path
[MAXPATHLEN
];
469 char zip_path
[MAXPATHLEN
];
472 * scan the man section directory for pages
474 npages
= scandir(section
, &entries
, NULL
, alphasort
);
476 warn("%s/%s", mandir
, section
);
480 if (verbose
|| rm_junk
) {
482 * Maintain a list of all cat pages that should exist,
483 * corresponding to existing man pages.
485 expected
= (char **) calloc(npages
, sizeof(char *));
487 for (i
= 0; i
< npages
; free(entries
[i
++])) {
488 page_name
= entries
[i
]->d_name
;
489 snprintf(page_path
, sizeof page_path
, "%s/%s", section
,
491 if (!is_manpage_name(page_name
)) {
492 if (!(test_path(page_path
, NULL
) & TEST_DIR
)) {
493 junk(mandir
, page_path
,
494 "invalid man page name");
498 zipped
= is_bzipped(page_name
) ? BZIP
:
499 is_gzipped(page_name
) ? GZIP
: NONE
;
500 if (zipped
!= NONE
) {
501 snprintf(cat_path
, sizeof cat_path
, "%s/%s",
502 cat_section
, page_name
);
503 if (expected
!= NULL
)
504 expected
[nexpected
++] = strdup(page_name
);
505 process_page(mandir
, page_path
, cat_path
, zipped
);
508 * We've got an uncompressed man page,
509 * check to see if there's a (preferred)
512 snprintf(zip_path
, sizeof zip_path
, "%s%s",
514 if (test_path(zip_path
, NULL
) != 0) {
515 junk(mandir
, page_path
,
516 "man page unused due to existing " GZ_EXT
);
520 "warning, %s is uncompressed\n",
523 snprintf(cat_path
, sizeof cat_path
, "%s/%s",
524 cat_section
, page_name
);
525 if (expected
!= NULL
) {
526 asprintf(&expected
[nexpected
++],
529 process_page(mandir
, page_path
, cat_path
, NONE
);
534 if (expected
== NULL
)
537 * scan cat sections for junk
539 npages
= scandir(cat_section
, &entries
, NULL
, alphasort
);
541 for (i
= 0; i
< npages
; free(entries
[i
++])) {
542 const char *junk_reason
;
545 page_name
= entries
[i
]->d_name
;
546 if (strcmp(page_name
, ".") == 0 || strcmp(page_name
, "..") == 0)
549 * Keep the index into the expected cat page list
550 * ahead of the name we've found.
552 while (e
< nexpected
&&
553 (cmp
= strcmp(page_name
, expected
[e
])) > 0)
557 /* we have an unexpected page */
558 snprintf(cat_path
, sizeof cat_path
, "%s/%s", cat_section
,
560 if (!is_manpage_name(page_name
)) {
561 if (test_path(cat_path
, NULL
) & TEST_DIR
)
563 junk_reason
= "invalid cat page name";
564 } else if (!is_gzipped(page_name
) && e
+ 1 < nexpected
&&
565 strncmp(page_name
, expected
[e
+ 1], strlen(page_name
)) == 0 &&
566 strlen(expected
[e
+ 1]) == strlen(page_name
) + 3) {
567 junk_reason
= "cat page unused due to existing " GZ_EXT
;
569 junk_reason
= "cat page without man page";
570 junk(mandir
, cat_path
, junk_reason
);
573 while (e
< nexpected
)
580 * Processes a single man section.
583 process_section(char *mandir
, char *section
)
587 if (already_visited(mandir
, section
, 1))
590 fprintf(stderr
, " section %s\n", section
);
591 cat_section
= get_cat_section(section
);
592 if (make_writable_dir(mandir
, cat_section
))
593 scan_section(mandir
, section
, cat_section
);
598 select_sections(const struct dirent
*entry
)
603 name
= strdup(entry
->d_name
);
604 ret
= directory_type(name
) == MAN_SECTION_DIR
;
610 * Processes a single top-level man directory. If section isn't NULL,
611 * it will only process that section sub-directory, otherwise it will
612 * process all of them.
615 process_mandir(char *dir_name
, char *section
)
617 if (fchdir(starting_dir
) < 0)
619 if (already_visited(NULL
, dir_name
, section
== NULL
))
621 check_writable(dir_name
);
623 fprintf(stderr
, "man directory %s\n", dir_name
);
625 fprintf(stderr
, "cd %s\n", dir_name
);
626 if (chdir(dir_name
) < 0) {
627 warn("%s: chdir", dir_name
);
631 if (section
!= NULL
) {
632 process_section(dir_name
, section
);
634 struct dirent
**entries
;
635 char *machine_dir
, *arch_dir
;
639 nsections
= scandir(".", &entries
, select_sections
, alphasort
);
641 warn("%s", dir_name
);
645 for (i
= 0; i
< nsections
; i
++) {
646 process_section(dir_name
, entries
[i
]->d_name
);
647 asprintf(&machine_dir
, "%s/%s", entries
[i
]->d_name
,
649 if (test_path(machine_dir
, NULL
) & TEST_DIR
)
650 process_section(dir_name
, machine_dir
);
652 if (strcmp(machine_arch
, machine
) != 0) {
653 asprintf(&arch_dir
, "%s/%s", entries
[i
]->d_name
,
655 if (test_path(arch_dir
, NULL
) & TEST_DIR
)
656 process_section(dir_name
, arch_dir
);
666 * Processes one argument, which may be a colon-separated list of
670 process_argument(const char *arg
)
679 err(1, "out of memory");
680 while ((dir
= strsep(&parg
, ":")) != NULL
) {
681 switch (directory_type(dir
)) {
683 if (locale
!= NULL
) {
684 asprintf(&mandir
, "%s/%s", dir
, locale
);
685 process_mandir(mandir
, NULL
);
687 if (lang_locale
!= NULL
) {
688 asprintf(&mandir
, "%s/%s", dir
,
690 process_mandir(mandir
, NULL
);
694 process_mandir(dir
, NULL
);
697 case MAN_SECTION_DIR
: {
698 mandir
= strdup(dirname(dir
));
699 section
= strdup(basename(dir
));
700 process_mandir(mandir
, section
);
706 warnx("%s: directory name not in proper man form", dir
);
714 determine_locale(void)
718 if ((locale
= setlocale(LC_CTYPE
, "")) == NULL
) {
719 warnx("-L option used, but no locale found\n");
722 sep
= strchr(locale
, '_');
723 if (sep
!= NULL
&& isupper((unsigned char)sep
[1])
724 && isupper((unsigned char)sep
[2])) {
725 asprintf(&lang_locale
, "%.*s%s", (int)(sep
- locale
),
728 sep
= nl_langinfo(CODESET
);
729 if (sep
!= NULL
&& *sep
!= '\0' && strcmp(sep
, "US-ASCII") != 0) {
732 for (i
= 0; locale_device
[i
] != NULL
; i
+= 2) {
733 if (strcmp(sep
, locale_device
[i
]) == 0) {
734 nroff_device
= locale_device
[i
+ 1];
740 if (lang_locale
!= NULL
)
741 fprintf(stderr
, "short locale is %s\n", lang_locale
);
742 fprintf(stderr
, "nroff device is %s\n", nroff_device
);
749 fprintf(stderr
, "usage: %s [-fLnrv] [directories ...]\n",
755 main(int argc
, char **argv
)
759 if ((uid
= getuid()) == 0) {
760 fprintf(stderr
, "don't run %s as root, use:\n echo", argv
[0]);
761 for (optind
= 0; optind
< argc
; optind
++) {
762 fprintf(stderr
, " %s", argv
[optind
]);
764 fprintf(stderr
, " | nice -5 su -m man\n");
767 while ((opt
= getopt(argc
, argv
, "vnfLrh")) != -1) {
789 if ((starting_dir
= open(".", 0)) < 0) {
793 signal(SIGINT
, trap_signal
);
794 signal(SIGHUP
, trap_signal
);
795 signal(SIGQUIT
, trap_signal
);
796 signal(SIGTERM
, trap_signal
);
798 if ((machine
= getenv("MACHINE")) == NULL
) {
799 static struct utsname utsname
;
801 if (uname(&utsname
) == -1)
803 machine
= utsname
.machine
;
806 if ((machine_arch
= getenv("MACHINE_ARCH")) == NULL
)
807 machine_arch
= MACHINE_ARCH
;
809 if (optind
== argc
) {
810 const char *manpath
= getenv("MANPATH");
812 manpath
= DEFAULT_MANPATH
;
813 process_argument(manpath
);
815 while (optind
< argc
)
816 process_argument(argv
[optind
++]);