src - OpenBSD base system

diff options


context:
space:
mode:

author	Ingo Schwarze <schwarze@cvs.openbsd.org>	2011-12-19 02:26:34 +0000
committer	Ingo Schwarze <schwarze@cvs.openbsd.org>	2011-12-19 02:26:34 +0000
commit	7624e132604813b4df8d86067616c584406c5ce7 (patch)
tree	3b56acba5d4ca75672edea9782ecad021cc3ac84
parent	a1bf4d36dbed03ececc602e56370c7a377c68b6e (diff)

Improvements from kristaps@:

(1) Make the database format simpler and smaller by - storing the file type as a single character and - storing paths relative to the dir containing the database. The latter allows to move trees around. Both together typically save 15-25% of the index size. (2) Make sure apropos(1) "any" really covers all search keys. (3) Make manpath_parseline() static, drop manpath_parseconf() completely.

Diffstat

-rw-r--r--

usr.bin/mandoc/apropos_db.c

-rw-r--r--

usr.bin/mandoc/apropos_db.h

-rw-r--r--

usr.bin/mandoc/mandocdb.8

-rw-r--r--

usr.bin/mandoc/mandocdb.c

-rw-r--r--

usr.bin/mandoc/manpath.c

-rw-r--r--

usr.bin/mandoc/manpath.h

6 files changed, 80 insertions, 72 deletions

diff --git a/usr.bin/mandoc/apropos_db.c b/usr.bin/mandoc/apropos_db.c
index ec1be3164e3..4c39b0fe3f9 100644
--- a/usr.bin/mandoc/apropos_db.c
+++ b/usr.bin/mandoc/apropos_db.c

@@ -1,4 +1,4 @@

-/* $Id: apropos_db.c,v 1.14 2011/12/10 22:18:20 schwarze Exp $ */

+/* $Id: apropos_db.c,v 1.15 2011/12/19 02:26:33 schwarze Exp $ */

@@ -105,7 +105,7 @@ static const struct type types[] = {

{ TYPE_Va, "Va" },

{ TYPE_Va, "Vt" },

{ TYPE_Xr, "Xr" },

- { INT_MAX, "any" },

+ { UINT64_MAX, "any" },

{ 0, NULL }

};

@@ -186,7 +186,7 @@ btree_read(const DBT *k, const DBT *v,

static size_t

norm_utf8(unsigned int cp, char out[7])

{

- size_t rc;

+ int rc;

rc = 0;

@@ -227,7 +227,7 @@ norm_utf8(unsigned int cp, char out[7])

return(0);

out[rc] = '\0';

- return(rc);

+ return((size_t)rc);

}

@@ -354,6 +354,7 @@ index_read(const DBT *key, const DBT *val, int index,

{

size_t left;

char *np, *cp;

+ char type;

#define INDEX_BREAD(_dst) \

do { \

@@ -364,13 +365,23 @@ index_read(const DBT *key, const DBT *val, int index,

cp = np + 1; \

} while (/* CONSTCOND */ 0)

- left = val->size;

- cp = (char *)val->data;

+ if (0 == (left = val->size))

+ return(0);

+ cp = val->data;

rec->res.rec = *(recno_t *)key->data;

rec->res.volume = index;

- INDEX_BREAD(rec->res.type);

+ if ('d' == (type = *cp++))

+ rec->res.type = RESTYPE_MDOC;

+ else if ('a' == type)

+ rec->res.type = RESTYPE_MAN;

+ else if ('c' == type)

+ rec->res.type = RESTYPE_CAT;

+ else

+ return(0);

+ left--;

INDEX_BREAD(rec->res.file);

INDEX_BREAD(rec->res.cat);

INDEX_BREAD(rec->res.title);

@@ -570,7 +581,6 @@ static void

recfree(struct rec *rec)

{

- free(rec->res.type);

free(rec->res.file);

free(rec->res.cat);

free(rec->res.title);

diff --git a/usr.bin/mandoc/apropos_db.h b/usr.bin/mandoc/apropos_db.h
index 302d8435694..82f5ac3d543 100644
--- a/usr.bin/mandoc/apropos_db.h
+++ b/usr.bin/mandoc/apropos_db.h

@@ -1,4 +1,4 @@

-/* $Id: apropos_db.h,v 1.9 2011/11/28 00:16:38 schwarze Exp $ */

+/* $Id: apropos_db.h,v 1.10 2011/12/19 02:26:33 schwarze Exp $ */

@@ -17,8 +17,14 @@

#ifndef APROPOS_H

#define APROPOS_H

+enum restype {

+ RESTYPE_MAN, /* man(7) file */

+ RESTYPE_MDOC, /* mdoc(7) file */

+ RESTYPE_CAT /* pre-formatted file */

+};

struct res {

- char *type; /* file type: mdoc, man or cat */

+ enum restype type; /* input file type */

char *file; /* file in file-system */

char *cat; /* category (3p, 3, etc.) */

char *title; /* title (FOO, etc.) */

diff --git a/usr.bin/mandoc/mandocdb.8 b/usr.bin/mandoc/mandocdb.8
index c0c72a1224c..51ba0aa6b88 100644
--- a/usr.bin/mandoc/mandocdb.8
+++ b/usr.bin/mandoc/mandocdb.8

@@ -1,4 +1,4 @@

-.\" $Id: mandocdb.8,v 1.12 2011/12/12 01:59:13 schwarze Exp $

+.\" $Id: mandocdb.8,v 1.13 2011/12/19 02:26:33 schwarze Exp $

.\"

@@ -14,7 +14,7 @@

.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF

.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

.\"

-.Dd $Mdocdate: December 12 2011 $

+.Dd $Mdocdate: December 19 2011 $

.Dt MANDOCDB 8

.Os

.Sh NAME

@@ -110,20 +110,19 @@ database with record values consisting of

.Pp

.Bl -enum -compact

.It

-the string

-.Cm mdoc ,

-.Cm man ,

+the character

+.Cm d ,

+.Cm a ,

-.Cm cat

+.Cm c

to indicate the file type

.Po

-file in

.Xr mdoc 7 ,

.Xr man 7 ,

and post-formatted, respectively

.Pc ,

.It

-the filename,

+the filename relative to the databases' path,

.It

the manual section,

.It

@@ -137,9 +136,7 @@ and the description.

.Pp

Each of the above is NUL-terminated.

.Pp

-Both the manual section and description may be zero-length if the record

-is unassigned.

-Entries are sequentially-numbered, but the filenames are unordered.

+If the record value is zero-length, it is unassigned.

.Ss Keyword Database

The keyword database,

.Pa mandoc.db ,

diff --git a/usr.bin/mandoc/mandocdb.c b/usr.bin/mandoc/mandocdb.c
index 3e2152e5c54..7cea6b47866 100644
--- a/usr.bin/mandoc/mandocdb.c
+++ b/usr.bin/mandoc/mandocdb.c

@@ -1,4 +1,4 @@

-/* $Id: mandocdb.c,v 1.25 2011/12/12 01:59:13 schwarze Exp $ */

+/* $Id: mandocdb.c,v 1.26 2011/12/19 02:26:33 schwarze Exp $ */

@@ -17,7 +17,6 @@

#include <sys/param.h>

#include <sys/types.h>

-#include <sys/stat.h>

#include <assert.h>

#include <dirent.h>

@@ -97,7 +96,7 @@ static void index_prune(const struct of *, DB *,

recno_t *, recno_t **, size_t *,

size_t *);

static void ofile_argbuild(int, char *[], struct of **);

-static int ofile_dirbuild(const char *, const char *,

+static void ofile_dirbuild(const char *, const char *,

const char *, int, struct of **);

static void ofile_free(struct of *);

static void pformatted(DB *, struct buf *, struct buf *,

@@ -372,12 +371,8 @@ mandocdb(int argc, char *argv[])

exit((int)MANDOCLEVEL_SYSERR);

}

- if (verb > 2) {

- printf("%s: Opened\n", fbuf);

- printf("%s: Opened\n", ibuf);

- }

ofile_argbuild(argc, argv, &of);

if (NULL == of)

goto out;

@@ -387,13 +382,16 @@ mandocdb(int argc, char *argv[])

&maxrec, &recs, &recsz, &reccur);

- * Go to the root of the respective manual tree

- * such that .so links work. In case of failure,

- * just prod on, even though .so links won't work.

+ * Go to the root of the respective manual tree.

+ * This must work or no manuals may be found (they're

+ * indexed relative to the root).

if (OP_UPDATE == op) {

- chdir(dir);

+ if (-1 == chdir(dir)) {

+ perror(dir);

+ exit((int)MANDOCLEVEL_SYSERR);

+ }

index_merge(of, mp, &dbuf, &buf, hash,

db, fbuf, idx, ibuf,

maxrec, recs, reccur);

@@ -462,9 +460,12 @@ mandocdb(int argc, char *argv[])

ofile_free(of);

of = NULL;

- if ( ! ofile_dirbuild(dirs.paths[i], NULL, NULL,

- 0, &of))

+ if (-1 == chdir(dirs.paths[i])) {

+ perror(dirs.paths[i]);

exit((int)MANDOCLEVEL_SYSERR);

+ }

+ ofile_dirbuild(".", NULL, NULL, 0, &of);

if (NULL == of)

continue;

@@ -472,12 +473,16 @@ mandocdb(int argc, char *argv[])

of = of->first;

- * Go to the root of the respective manual tree

- * such that .so links work. In case of failure,

- * just prod on, even though .so links won't work.

+ * Go to the root of the respective manual tree.

+ * This must work or no manuals may be found (they're

+ * indexed relative to the root).

- chdir(dirs.paths[i]);

+ if (-1 == chdir(dirs.paths[i])) {

+ perror(dirs.paths[i]);

+ exit((int)MANDOCLEVEL_SYSERR);

+ }

index_merge(of, mp, &dbuf, &buf, hash, db, fbuf,

idx, ibuf, maxrec, recs, reccur);

}

@@ -516,6 +521,7 @@ index_merge(const struct of *of, struct mparse *mp,

size_t sv;

unsigned seq;

struct db_val vbuf;

+ char type;

for (rec = 0; of; of = of->next) {

fn = of->fname;

@@ -596,7 +602,8 @@ index_merge(const struct of *of, struct mparse *mp,

dbuf->len = 0;

- buf_append(dbuf, mdoc ? "mdoc" : (man ? "man" : "cat"));

+ type = mdoc ? 'd' : (man ? 'a' : 'c');

+ buf_appendb(dbuf, &type, 1);

buf_appendb(dbuf, fn, strlen(fn) + 1);

buf_appendb(dbuf, msec, strlen(msec) + 1);

buf_appendb(dbuf, mtitle, strlen(mtitle) + 1);

@@ -684,7 +691,7 @@ index_prune(const struct of *ofile, DB *db, const char *dbf,

recno_t **recs, size_t *recsz, size_t *reccur)

{

const struct of *of;

- const char *fn, *cp;

+ const char *fn;

struct db_val *vbuf;

unsigned seq, sseq;

DBT key, val;

@@ -695,7 +702,6 @@ index_prune(const struct of *ofile, DB *db, const char *dbf,

while (0 == (ch = (*idx->seq)(idx, &key, &val, seq))) {

seq = R_NEXT;

*maxrec = *(recno_t *)key.data;

- cp = val.data;

/* Deleted records are zero-sized. Skip them. */

@@ -709,11 +715,8 @@ index_prune(const struct of *ofile, DB *db, const char *dbf,

* Failing any of these, we go into our error handler.

- if (NULL == (fn = memchr(cp, '\0', val.size)))

- break;

- if (++fn - cp >= (int)val.size)

- break;

- if (NULL == memchr(fn, '\0', val.size - (fn - cp)))

+ fn = (char *)val.data + 1;

+ if (NULL == memchr(fn, '\0', val.size - 1))

break;

@@ -1443,12 +1446,11 @@ ofile_argbuild(int argc, char *argv[], struct of **of)

* everything else is a manual.

* Pass in a pointer to a NULL structure for the first invocation.

-static int

+static void

ofile_dirbuild(const char *dir, const char* psec, const char *parch,

int p_src_form, struct of **of)

{

char buf[MAXPATHLEN];

- struct stat sb;

size_t sz;

DIR *d;

const char *fn, *sec, *arch;

@@ -1510,14 +1512,13 @@ ofile_dirbuild(const char *dir, const char* psec, const char *parch,

if (verb > 2)

printf("%s: Scanning\n", buf);

- if ( ! ofile_dirbuild(buf, sec, arch,

- src_form, of))

- return(0);

+ ofile_dirbuild(buf, sec, arch, src_form, of);

}

if (DT_REG != dp->d_type ||

- (NULL == psec && !use_all) ||

- !strcmp(MANDOC_DB, fn) ||

- !strcmp(MANDOC_IDX, fn))

+ (NULL == psec && !use_all) ||

+ ! strcmp(MANDOC_DB, fn) ||

+ ! strcmp(MANDOC_IDX, fn))

continue;

@@ -1581,13 +1582,15 @@ ofile_dirbuild(const char *dir, const char* psec, const char *parch,

"%s: Path too long\n", buf);

continue;

}

- if (0 == stat(buf, &sb))

+ if (0 == access(buf, R_OK))

continue;

}

+ assert('.' == dir[0]);

+ assert('/' == dir[1]);

buf[0] = '\0';

- strlcat(buf, dir, MAXPATHLEN);

+ strlcat(buf, dir + 2, MAXPATHLEN);

strlcat(buf, "/", MAXPATHLEN);

sz = strlcat(buf, fn, MAXPATHLEN);

if (sz >= MAXPATHLEN) {

@@ -1629,7 +1632,6 @@ ofile_dirbuild(const char *dir, const char* psec, const char *parch,

}

closedir(d);

- return(1);

}

static void

diff --git a/usr.bin/mandoc/manpath.c b/usr.bin/mandoc/manpath.c
index 86084a7aca5..ed58c403ccf 100644
--- a/usr.bin/mandoc/manpath.c
+++ b/usr.bin/mandoc/manpath.c

@@ -1,4 +1,4 @@

-/* $Id: manpath.c,v 1.2 2011/12/12 01:59:13 schwarze Exp $ */

+/* $Id: manpath.c,v 1.3 2011/12/19 02:26:33 schwarze Exp $ */

@@ -16,7 +16,8 @@

* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

-#include <sys/types.h>

+#include <sys/param.h>

#include <assert.h>

#include <ctype.h>

#include <limits.h>

@@ -31,6 +32,7 @@

#define MAN_CONF_KEY "_whatdb"

static void manpath_add(struct manpaths *, const char *);

+static void manpath_parseline(struct manpaths *, char *);

void

manpath_parse(struct manpaths *dirs, const char *file,

@@ -43,7 +45,7 @@ manpath_parse(struct manpaths *dirs, const char *file,

defp = getenv("MANPATH");

if (NULL == defp)

- manpath_parseconf(dirs, file);

+ manpath_manconf(dirs, file ? file : MAN_CONF_FILE);

else

manpath_parseline(dirs, defp);

}

@@ -51,7 +53,7 @@ manpath_parse(struct manpaths *dirs, const char *file,

* Parse a FULL pathname from a colon-separated list of arrays.

-void

+static void

manpath_parseline(struct manpaths *dirs, char *path)

{

char *dir;

@@ -89,13 +91,6 @@ manpath_add(struct manpaths *dirs, const char *dir)

}

void

-manpath_parseconf(struct manpaths *dirs, const char *file)

- manpath_manconf(dirs, file ? file : MAN_CONF_FILE);

-void

manpath_free(struct manpaths *p)

{

int i;

diff --git a/usr.bin/mandoc/manpath.h b/usr.bin/mandoc/manpath.h
index 632834d6fc4..4b2233d17a2 100644
--- a/usr.bin/mandoc/manpath.h
+++ b/usr.bin/mandoc/manpath.h

@@ -1,4 +1,4 @@

-/* $Id: manpath.h,v 1.2 2011/12/12 01:59:13 schwarze Exp $ */

+/* $Id: manpath.h,v 1.3 2011/12/19 02:26:33 schwarze Exp $ */

@@ -31,8 +31,6 @@ __BEGIN_DECLS

void manpath_manconf(struct manpaths *, const char *);

void manpath_parse(struct manpaths *, const char *, char *, char *);

-void manpath_parseconf(struct manpaths *, const char *);

-void manpath_parseline(struct manpaths *, char *);

void manpath_free(struct manpaths *);

__END_DECLS