2017-01-21 13:01:39 +00:00
|
|
|
/* $Id: mandocdb.c,v 1.237 2017/01/11 17:39:53 schwarze Exp $ */
|
2012-10-18 09:55:16 +00:00
|
|
|
/*
|
2014-01-31 19:59:03 +00:00
|
|
|
* Copyright (c) 2011, 2012 Kristaps Dzonsons <kristaps@bsd.lv>
|
2017-01-21 13:01:39 +00:00
|
|
|
* Copyright (c) 2011-2017 Ingo Schwarze <schwarze@openbsd.org>
|
|
|
|
* Copyright (c) 2016 Ed Maste <emaste@freebsd.org>
|
2012-10-18 09:55:16 +00:00
|
|
|
*
|
|
|
|
* Permission to use, copy, modify, and distribute this software for any
|
|
|
|
* purpose with or without fee is hereby granted, provided that the above
|
|
|
|
* copyright notice and this permission notice appear in all copies.
|
|
|
|
*
|
2016-01-15 23:08:59 +00:00
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
|
2012-10-18 09:55:16 +00:00
|
|
|
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
|
2016-01-15 23:08:59 +00:00
|
|
|
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
|
2012-10-18 09:55:16 +00:00
|
|
|
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
|
|
|
|
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
|
|
|
|
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
|
|
|
|
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
|
|
|
|
*/
|
|
|
|
#include "config.h"
|
|
|
|
|
2014-12-02 07:34:06 +00:00
|
|
|
#include <sys/types.h>
|
2014-11-22 18:08:25 +00:00
|
|
|
#include <sys/stat.h>
|
|
|
|
#include <sys/wait.h>
|
2012-10-18 09:55:16 +00:00
|
|
|
|
|
|
|
#include <assert.h>
|
|
|
|
#include <ctype.h>
|
2016-01-15 23:08:59 +00:00
|
|
|
#if HAVE_ERR
|
|
|
|
#include <err.h>
|
|
|
|
#endif
|
2012-10-18 09:55:16 +00:00
|
|
|
#include <errno.h>
|
|
|
|
#include <fcntl.h>
|
2014-12-02 07:34:06 +00:00
|
|
|
#if HAVE_FTS
|
2014-11-22 18:08:25 +00:00
|
|
|
#include <fts.h>
|
2014-12-02 07:34:06 +00:00
|
|
|
#else
|
|
|
|
#include "compat_fts.h"
|
|
|
|
#endif
|
2014-01-31 19:59:03 +00:00
|
|
|
#include <limits.h>
|
2016-07-23 11:45:42 +00:00
|
|
|
#if HAVE_SANDBOX_INIT
|
|
|
|
#include <sandbox.h>
|
|
|
|
#endif
|
2017-01-21 13:01:39 +00:00
|
|
|
#include <stdarg.h>
|
2014-11-22 18:08:25 +00:00
|
|
|
#include <stddef.h>
|
2012-10-18 09:55:16 +00:00
|
|
|
#include <stdio.h>
|
|
|
|
#include <stdint.h>
|
|
|
|
#include <stdlib.h>
|
|
|
|
#include <string.h>
|
|
|
|
#include <unistd.h>
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
#include "mandoc_aux.h"
|
|
|
|
#include "mandoc_ohash.h"
|
|
|
|
#include "mandoc.h"
|
|
|
|
#include "roff.h"
|
2012-10-18 09:55:16 +00:00
|
|
|
#include "mdoc.h"
|
2014-11-22 18:08:25 +00:00
|
|
|
#include "man.h"
|
2016-01-15 23:08:59 +00:00
|
|
|
#include "manconf.h"
|
2014-11-22 18:08:25 +00:00
|
|
|
#include "mansearch.h"
|
2017-01-21 13:01:39 +00:00
|
|
|
#include "dba_array.h"
|
|
|
|
#include "dba.h"
|
2014-11-22 18:08:25 +00:00
|
|
|
|
|
|
|
extern const char *const mansearch_keynames[];
|
|
|
|
|
|
|
|
enum op {
|
|
|
|
OP_DEFAULT = 0, /* new dbs from dir list or default config */
|
|
|
|
OP_CONFFILE, /* new databases from custom config file */
|
|
|
|
OP_UPDATE, /* delete/add entries in existing database */
|
|
|
|
OP_DELETE, /* delete entries from existing database */
|
|
|
|
OP_TEST /* change no databases, report potential problems */
|
2012-10-18 09:55:16 +00:00
|
|
|
};
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
struct str {
|
|
|
|
const struct mpage *mpage; /* if set, the owning parse */
|
|
|
|
uint64_t mask; /* bitmask in sequence */
|
2014-12-25 21:33:25 +00:00
|
|
|
char key[]; /* rendered text */
|
2012-10-18 09:55:16 +00:00
|
|
|
};
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
struct inodev {
|
|
|
|
ino_t st_ino;
|
|
|
|
dev_t st_dev;
|
|
|
|
};
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
struct mpage {
|
|
|
|
struct inodev inodev; /* used for hashing routine */
|
2017-01-21 13:01:39 +00:00
|
|
|
struct dba_array *dba;
|
2014-11-22 18:08:25 +00:00
|
|
|
char *sec; /* section from file content */
|
|
|
|
char *arch; /* architecture from file content */
|
|
|
|
char *title; /* title from file content */
|
|
|
|
char *desc; /* description from file content */
|
2017-01-21 13:01:39 +00:00
|
|
|
struct mpage *next; /* singly linked list */
|
2014-11-22 18:08:25 +00:00
|
|
|
struct mlink *mlinks; /* singly linked list */
|
2014-12-25 21:33:25 +00:00
|
|
|
int name_head_done;
|
2017-01-21 13:01:39 +00:00
|
|
|
enum form form; /* format from file content */
|
2012-10-18 09:55:16 +00:00
|
|
|
};
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
struct mlink {
|
|
|
|
char file[PATH_MAX]; /* filename rel. to manpath */
|
|
|
|
char *dsec; /* section from directory */
|
|
|
|
char *arch; /* architecture from directory */
|
|
|
|
char *name; /* name from file name (not empty) */
|
|
|
|
char *fsec; /* section from file name suffix */
|
|
|
|
struct mlink *next; /* singly linked list */
|
|
|
|
struct mpage *mpage; /* parent */
|
|
|
|
int gzip; /* filename has a .gz suffix */
|
2017-01-21 13:01:39 +00:00
|
|
|
enum form dform; /* format from directory */
|
|
|
|
enum form fform; /* format from file name suffix */
|
2012-10-18 09:55:16 +00:00
|
|
|
};
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
typedef int (*mdoc_fp)(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
|
|
|
struct mdoc_handler {
|
2014-11-22 18:08:25 +00:00
|
|
|
mdoc_fp fp; /* optional handler */
|
|
|
|
uint64_t mask; /* set unless handler returns 0 */
|
2017-01-21 13:01:39 +00:00
|
|
|
int taboo; /* node flags that must not be set */
|
2012-10-18 09:55:16 +00:00
|
|
|
};
|
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
|
|
|
|
int mandocdb(int, char *[]);
|
|
|
|
|
|
|
|
static void dbadd(struct dba *, struct mpage *);
|
2014-11-22 18:08:25 +00:00
|
|
|
static void dbadd_mlink(const struct mlink *mlink);
|
2017-01-21 13:01:39 +00:00
|
|
|
static void dbprune(struct dba *);
|
|
|
|
static void dbwrite(struct dba *);
|
2014-11-22 18:08:25 +00:00
|
|
|
static void filescan(const char *);
|
2017-01-21 13:01:39 +00:00
|
|
|
#if HAVE_FTS_COMPARE_CONST
|
|
|
|
static int fts_compare(const FTSENT *const *, const FTSENT *const *);
|
|
|
|
#else
|
|
|
|
static int fts_compare(const FTSENT **, const FTSENT **);
|
|
|
|
#endif
|
2014-11-22 18:08:25 +00:00
|
|
|
static void mlink_add(struct mlink *, const struct stat *);
|
|
|
|
static void mlink_check(struct mpage *, struct mlink *);
|
|
|
|
static void mlink_free(struct mlink *);
|
|
|
|
static void mlinks_undupe(struct mpage *);
|
|
|
|
static void mpages_free(void);
|
2017-01-21 13:01:39 +00:00
|
|
|
static void mpages_merge(struct dba *, struct mparse *);
|
2014-11-22 18:08:25 +00:00
|
|
|
static void parse_cat(struct mpage *, int);
|
2016-01-15 23:08:59 +00:00
|
|
|
static void parse_man(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static void parse_mdoc(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static int parse_mdoc_head(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static int parse_mdoc_Fd(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static void parse_mdoc_fname(struct mpage *, const struct roff_node *);
|
|
|
|
static int parse_mdoc_Fn(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static int parse_mdoc_Fo(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static int parse_mdoc_Nd(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static int parse_mdoc_Nm(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static int parse_mdoc_Sh(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static int parse_mdoc_Va(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
|
|
|
static int parse_mdoc_Xr(struct mpage *, const struct roff_meta *,
|
|
|
|
const struct roff_node *);
|
2014-11-22 18:08:25 +00:00
|
|
|
static void putkey(const struct mpage *, char *, uint64_t);
|
2014-12-25 21:33:25 +00:00
|
|
|
static void putkeys(const struct mpage *, char *, size_t, uint64_t);
|
2014-11-22 18:08:25 +00:00
|
|
|
static void putmdockey(const struct mpage *,
|
2017-01-21 13:01:39 +00:00
|
|
|
const struct roff_node *, uint64_t, int);
|
2014-12-25 21:33:25 +00:00
|
|
|
static int render_string(char **, size_t *);
|
2017-01-21 13:01:39 +00:00
|
|
|
static void say(const char *, const char *, ...)
|
|
|
|
__attribute__((__format__ (printf, 2, 3)));
|
2014-12-02 07:34:06 +00:00
|
|
|
static int set_basedir(const char *, int);
|
2014-11-22 18:08:25 +00:00
|
|
|
static int treescan(void);
|
|
|
|
static size_t utf8(unsigned int, char [7]);
|
|
|
|
|
|
|
|
static int nodb; /* no database changes */
|
|
|
|
static int mparse_options; /* abort the parse early */
|
|
|
|
static int use_all; /* use all found files */
|
|
|
|
static int debug; /* print what we're doing */
|
|
|
|
static int warnings; /* warn about crap */
|
|
|
|
static int write_utf8; /* write UTF-8 output; else ASCII */
|
|
|
|
static int exitcode; /* to be returned by main */
|
|
|
|
static enum op op; /* operational mode */
|
|
|
|
static char basedir[PATH_MAX]; /* current base directory */
|
2017-01-21 13:01:39 +00:00
|
|
|
static struct mpage *mpage_head; /* list of distinct manual pages */
|
2014-11-22 18:08:25 +00:00
|
|
|
static struct ohash mpages; /* table of distinct manual pages */
|
|
|
|
static struct ohash mlinks; /* table of directory entries */
|
|
|
|
static struct ohash names; /* table of all names */
|
|
|
|
static struct ohash strings; /* table of all strings */
|
|
|
|
static uint64_t name_mask;
|
|
|
|
|
2012-10-18 09:55:16 +00:00
|
|
|
static const struct mdoc_handler mdocs[MDOC_MAX] = {
|
2017-01-21 13:01:39 +00:00
|
|
|
{ NULL, 0, 0 }, /* Ap */
|
|
|
|
{ NULL, 0, NODE_NOPRT }, /* Dd */
|
|
|
|
{ NULL, 0, NODE_NOPRT }, /* Dt */
|
|
|
|
{ NULL, 0, NODE_NOPRT }, /* Os */
|
|
|
|
{ parse_mdoc_Sh, TYPE_Sh, 0 }, /* Sh */
|
|
|
|
{ parse_mdoc_head, TYPE_Ss, 0 }, /* Ss */
|
|
|
|
{ NULL, 0, 0 }, /* Pp */
|
|
|
|
{ NULL, 0, 0 }, /* D1 */
|
|
|
|
{ NULL, 0, 0 }, /* Dl */
|
|
|
|
{ NULL, 0, 0 }, /* Bd */
|
|
|
|
{ NULL, 0, 0 }, /* Ed */
|
|
|
|
{ NULL, 0, 0 }, /* Bl */
|
|
|
|
{ NULL, 0, 0 }, /* El */
|
|
|
|
{ NULL, 0, 0 }, /* It */
|
|
|
|
{ NULL, 0, 0 }, /* Ad */
|
|
|
|
{ NULL, TYPE_An, 0 }, /* An */
|
|
|
|
{ NULL, TYPE_Ar, 0 }, /* Ar */
|
|
|
|
{ NULL, TYPE_Cd, 0 }, /* Cd */
|
|
|
|
{ NULL, TYPE_Cm, 0 }, /* Cm */
|
|
|
|
{ NULL, TYPE_Dv, 0 }, /* Dv */
|
|
|
|
{ NULL, TYPE_Er, 0 }, /* Er */
|
|
|
|
{ NULL, TYPE_Ev, 0 }, /* Ev */
|
|
|
|
{ NULL, 0, 0 }, /* Ex */
|
|
|
|
{ NULL, TYPE_Fa, 0 }, /* Fa */
|
|
|
|
{ parse_mdoc_Fd, 0, 0 }, /* Fd */
|
|
|
|
{ NULL, TYPE_Fl, 0 }, /* Fl */
|
|
|
|
{ parse_mdoc_Fn, 0, 0 }, /* Fn */
|
|
|
|
{ NULL, TYPE_Ft, 0 }, /* Ft */
|
|
|
|
{ NULL, TYPE_Ic, 0 }, /* Ic */
|
|
|
|
{ NULL, TYPE_In, 0 }, /* In */
|
|
|
|
{ NULL, TYPE_Li, 0 }, /* Li */
|
|
|
|
{ parse_mdoc_Nd, 0, 0 }, /* Nd */
|
|
|
|
{ parse_mdoc_Nm, 0, 0 }, /* Nm */
|
|
|
|
{ NULL, 0, 0 }, /* Op */
|
|
|
|
{ NULL, 0, 0 }, /* Ot */
|
|
|
|
{ NULL, TYPE_Pa, NODE_NOSRC }, /* Pa */
|
|
|
|
{ NULL, 0, 0 }, /* Rv */
|
|
|
|
{ NULL, TYPE_St, 0 }, /* St */
|
|
|
|
{ parse_mdoc_Va, TYPE_Va, 0 }, /* Va */
|
|
|
|
{ parse_mdoc_Va, TYPE_Vt, 0 }, /* Vt */
|
|
|
|
{ parse_mdoc_Xr, 0, 0 }, /* Xr */
|
|
|
|
{ NULL, 0, 0 }, /* %A */
|
|
|
|
{ NULL, 0, 0 }, /* %B */
|
|
|
|
{ NULL, 0, 0 }, /* %D */
|
|
|
|
{ NULL, 0, 0 }, /* %I */
|
|
|
|
{ NULL, 0, 0 }, /* %J */
|
|
|
|
{ NULL, 0, 0 }, /* %N */
|
|
|
|
{ NULL, 0, 0 }, /* %O */
|
|
|
|
{ NULL, 0, 0 }, /* %P */
|
|
|
|
{ NULL, 0, 0 }, /* %R */
|
|
|
|
{ NULL, 0, 0 }, /* %T */
|
|
|
|
{ NULL, 0, 0 }, /* %V */
|
|
|
|
{ NULL, 0, 0 }, /* Ac */
|
|
|
|
{ NULL, 0, 0 }, /* Ao */
|
|
|
|
{ NULL, 0, 0 }, /* Aq */
|
|
|
|
{ NULL, TYPE_At, 0 }, /* At */
|
|
|
|
{ NULL, 0, 0 }, /* Bc */
|
|
|
|
{ NULL, 0, 0 }, /* Bf */
|
|
|
|
{ NULL, 0, 0 }, /* Bo */
|
|
|
|
{ NULL, 0, 0 }, /* Bq */
|
|
|
|
{ NULL, TYPE_Bsx, NODE_NOSRC }, /* Bsx */
|
|
|
|
{ NULL, TYPE_Bx, NODE_NOSRC }, /* Bx */
|
|
|
|
{ NULL, 0, 0 }, /* Db */
|
|
|
|
{ NULL, 0, 0 }, /* Dc */
|
|
|
|
{ NULL, 0, 0 }, /* Do */
|
|
|
|
{ NULL, 0, 0 }, /* Dq */
|
|
|
|
{ NULL, 0, 0 }, /* Ec */
|
|
|
|
{ NULL, 0, 0 }, /* Ef */
|
|
|
|
{ NULL, TYPE_Em, 0 }, /* Em */
|
|
|
|
{ NULL, 0, 0 }, /* Eo */
|
|
|
|
{ NULL, TYPE_Fx, NODE_NOSRC }, /* Fx */
|
|
|
|
{ NULL, TYPE_Ms, 0 }, /* Ms */
|
|
|
|
{ NULL, 0, 0 }, /* No */
|
|
|
|
{ NULL, 0, 0 }, /* Ns */
|
|
|
|
{ NULL, TYPE_Nx, NODE_NOSRC }, /* Nx */
|
|
|
|
{ NULL, TYPE_Ox, NODE_NOSRC }, /* Ox */
|
|
|
|
{ NULL, 0, 0 }, /* Pc */
|
|
|
|
{ NULL, 0, 0 }, /* Pf */
|
|
|
|
{ NULL, 0, 0 }, /* Po */
|
|
|
|
{ NULL, 0, 0 }, /* Pq */
|
|
|
|
{ NULL, 0, 0 }, /* Qc */
|
|
|
|
{ NULL, 0, 0 }, /* Ql */
|
|
|
|
{ NULL, 0, 0 }, /* Qo */
|
|
|
|
{ NULL, 0, 0 }, /* Qq */
|
|
|
|
{ NULL, 0, 0 }, /* Re */
|
|
|
|
{ NULL, 0, 0 }, /* Rs */
|
|
|
|
{ NULL, 0, 0 }, /* Sc */
|
|
|
|
{ NULL, 0, 0 }, /* So */
|
|
|
|
{ NULL, 0, 0 }, /* Sq */
|
|
|
|
{ NULL, 0, 0 }, /* Sm */
|
|
|
|
{ NULL, 0, 0 }, /* Sx */
|
|
|
|
{ NULL, TYPE_Sy, 0 }, /* Sy */
|
|
|
|
{ NULL, TYPE_Tn, 0 }, /* Tn */
|
|
|
|
{ NULL, 0, NODE_NOSRC }, /* Ux */
|
|
|
|
{ NULL, 0, 0 }, /* Xc */
|
|
|
|
{ NULL, 0, 0 }, /* Xo */
|
|
|
|
{ parse_mdoc_Fo, 0, 0 }, /* Fo */
|
|
|
|
{ NULL, 0, 0 }, /* Fc */
|
|
|
|
{ NULL, 0, 0 }, /* Oo */
|
|
|
|
{ NULL, 0, 0 }, /* Oc */
|
|
|
|
{ NULL, 0, 0 }, /* Bk */
|
|
|
|
{ NULL, 0, 0 }, /* Ek */
|
|
|
|
{ NULL, 0, 0 }, /* Bt */
|
|
|
|
{ NULL, 0, 0 }, /* Hf */
|
|
|
|
{ NULL, 0, 0 }, /* Fr */
|
|
|
|
{ NULL, 0, 0 }, /* Ud */
|
|
|
|
{ NULL, TYPE_Lb, NODE_NOSRC }, /* Lb */
|
|
|
|
{ NULL, 0, 0 }, /* Lp */
|
|
|
|
{ NULL, TYPE_Lk, 0 }, /* Lk */
|
|
|
|
{ NULL, TYPE_Mt, NODE_NOSRC }, /* Mt */
|
|
|
|
{ NULL, 0, 0 }, /* Brq */
|
|
|
|
{ NULL, 0, 0 }, /* Bro */
|
|
|
|
{ NULL, 0, 0 }, /* Brc */
|
|
|
|
{ NULL, 0, 0 }, /* %C */
|
|
|
|
{ NULL, 0, 0 }, /* Es */
|
|
|
|
{ NULL, 0, 0 }, /* En */
|
|
|
|
{ NULL, TYPE_Dx, NODE_NOSRC }, /* Dx */
|
|
|
|
{ NULL, 0, 0 }, /* %Q */
|
|
|
|
{ NULL, 0, 0 }, /* br */
|
|
|
|
{ NULL, 0, 0 }, /* sp */
|
|
|
|
{ NULL, 0, 0 }, /* %U */
|
|
|
|
{ NULL, 0, 0 }, /* Ta */
|
|
|
|
{ NULL, 0, 0 }, /* ll */
|
2012-10-18 09:55:16 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
int
|
2014-12-25 21:33:25 +00:00
|
|
|
mandocdb(int argc, char *argv[])
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2016-01-15 23:08:59 +00:00
|
|
|
struct manconf conf;
|
2014-11-22 18:08:25 +00:00
|
|
|
struct mparse *mp;
|
2017-01-21 13:01:39 +00:00
|
|
|
struct dba *dba;
|
2016-01-15 23:08:59 +00:00
|
|
|
const char *path_arg, *progname;
|
|
|
|
size_t j, sz;
|
|
|
|
int ch, i;
|
2014-11-22 18:08:25 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
#if HAVE_PLEDGE
|
|
|
|
if (pledge("stdio rpath wpath cpath fattr flock proc exec", NULL) == -1) {
|
|
|
|
warn("pledge");
|
|
|
|
return (int)MANDOCLEVEL_SYSERR;
|
|
|
|
}
|
|
|
|
#endif
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-07-23 11:45:42 +00:00
|
|
|
#if HAVE_SANDBOX_INIT
|
|
|
|
if (sandbox_init(kSBXProfileNoInternet, SANDBOX_NAMED, NULL) == -1) {
|
|
|
|
warnx("sandbox_init");
|
|
|
|
return (int)MANDOCLEVEL_SYSERR;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
memset(&conf, 0, sizeof(conf));
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* We accept a few different invocations.
|
|
|
|
* The CHECKOP macro makes sure that invocation styles don't
|
|
|
|
* clobber each other.
|
|
|
|
*/
|
|
|
|
#define CHECKOP(_op, _ch) do \
|
|
|
|
if (OP_DEFAULT != (_op)) { \
|
2016-01-15 23:08:59 +00:00
|
|
|
warnx("-%c: Conflicting option", (_ch)); \
|
2014-11-22 18:08:25 +00:00
|
|
|
goto usage; \
|
|
|
|
} while (/*CONSTCOND*/0)
|
|
|
|
|
|
|
|
path_arg = NULL;
|
2012-10-18 09:55:16 +00:00
|
|
|
op = OP_DEFAULT;
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
while (-1 != (ch = getopt(argc, argv, "aC:Dd:npQT:tu:v")))
|
2012-10-18 09:55:16 +00:00
|
|
|
switch (ch) {
|
2014-11-22 18:08:25 +00:00
|
|
|
case 'a':
|
2012-10-18 09:55:16 +00:00
|
|
|
use_all = 1;
|
|
|
|
break;
|
2014-11-22 18:08:25 +00:00
|
|
|
case 'C':
|
|
|
|
CHECKOP(op, ch);
|
|
|
|
path_arg = optarg;
|
2012-10-18 09:55:16 +00:00
|
|
|
op = OP_CONFFILE;
|
|
|
|
break;
|
2014-11-22 18:08:25 +00:00
|
|
|
case 'D':
|
|
|
|
debug++;
|
|
|
|
break;
|
|
|
|
case 'd':
|
|
|
|
CHECKOP(op, ch);
|
|
|
|
path_arg = optarg;
|
2012-10-18 09:55:16 +00:00
|
|
|
op = OP_UPDATE;
|
|
|
|
break;
|
2014-11-22 18:08:25 +00:00
|
|
|
case 'n':
|
|
|
|
nodb = 1;
|
|
|
|
break;
|
|
|
|
case 'p':
|
2012-10-18 09:55:16 +00:00
|
|
|
warnings = 1;
|
|
|
|
break;
|
2014-11-22 18:08:25 +00:00
|
|
|
case 'Q':
|
|
|
|
mparse_options |= MPARSE_QUICK;
|
|
|
|
break;
|
|
|
|
case 'T':
|
|
|
|
if (strcmp(optarg, "utf8")) {
|
2016-01-15 23:08:59 +00:00
|
|
|
warnx("-T%s: Unsupported output format",
|
|
|
|
optarg);
|
2012-10-18 09:55:16 +00:00
|
|
|
goto usage;
|
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
write_utf8 = 1;
|
|
|
|
break;
|
|
|
|
case 't':
|
|
|
|
CHECKOP(op, ch);
|
|
|
|
dup2(STDOUT_FILENO, STDERR_FILENO);
|
|
|
|
op = OP_TEST;
|
|
|
|
nodb = warnings = 1;
|
2012-10-18 09:55:16 +00:00
|
|
|
break;
|
2014-11-22 18:08:25 +00:00
|
|
|
case 'u':
|
|
|
|
CHECKOP(op, ch);
|
|
|
|
path_arg = optarg;
|
|
|
|
op = OP_DELETE;
|
2012-10-18 09:55:16 +00:00
|
|
|
break;
|
2014-11-22 18:08:25 +00:00
|
|
|
case 'v':
|
|
|
|
/* Compatibility with espie@'s makewhatis. */
|
2012-10-18 09:55:16 +00:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
goto usage;
|
|
|
|
}
|
|
|
|
|
|
|
|
argc -= optind;
|
|
|
|
argv += optind;
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
#if HAVE_PLEDGE
|
|
|
|
if (nodb) {
|
|
|
|
if (pledge("stdio rpath", NULL) == -1) {
|
|
|
|
warn("pledge");
|
|
|
|
return (int)MANDOCLEVEL_SYSERR;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2012-10-18 09:55:16 +00:00
|
|
|
if (OP_CONFFILE == op && argc > 0) {
|
2016-01-15 23:08:59 +00:00
|
|
|
warnx("-C: Too many arguments");
|
2012-10-18 09:55:16 +00:00
|
|
|
goto usage;
|
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
exitcode = (int)MANDOCLEVEL_OK;
|
2016-01-15 23:08:59 +00:00
|
|
|
mchars_alloc();
|
|
|
|
mp = mparse_alloc(mparse_options, MANDOCLEVEL_BADARG, NULL, NULL);
|
|
|
|
mandoc_ohash_init(&mpages, 6, offsetof(struct mpage, inodev));
|
|
|
|
mandoc_ohash_init(&mlinks, 6, offsetof(struct mlink, file));
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (OP_UPDATE == op || OP_DELETE == op || OP_TEST == op) {
|
2012-10-18 09:55:16 +00:00
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* Most of these deal with a specific directory.
|
|
|
|
* Jump into that directory first.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
2014-12-02 07:34:06 +00:00
|
|
|
if (OP_TEST != op && 0 == set_basedir(path_arg, 1))
|
2014-11-22 18:08:25 +00:00
|
|
|
goto out;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
dba = nodb ? dba_new(128) : dba_read(MANDOC_DB);
|
|
|
|
if (dba != NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* The existing database is usable. Process
|
|
|
|
* all files specified on the command-line.
|
|
|
|
*/
|
2016-01-15 23:08:59 +00:00
|
|
|
#if HAVE_PLEDGE
|
|
|
|
if (!nodb) {
|
|
|
|
if (pledge("stdio rpath wpath cpath fattr flock", NULL) == -1) {
|
|
|
|
warn("pledge");
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif
|
2014-11-22 18:08:25 +00:00
|
|
|
use_all = 1;
|
|
|
|
for (i = 0; i < argc; i++)
|
|
|
|
filescan(argv[i]);
|
2017-01-21 13:01:39 +00:00
|
|
|
if (nodb == 0)
|
|
|
|
dbprune(dba);
|
2014-11-22 18:08:25 +00:00
|
|
|
} else {
|
2017-01-21 13:01:39 +00:00
|
|
|
/* Database missing or corrupt. */
|
|
|
|
if (op != OP_UPDATE || errno != ENOENT)
|
|
|
|
say(MANDOC_DB, "%s: Automatically recreating"
|
|
|
|
" from scratch", strerror(errno));
|
2014-11-22 18:08:25 +00:00
|
|
|
exitcode = (int)MANDOCLEVEL_OK;
|
|
|
|
op = OP_DEFAULT;
|
|
|
|
if (0 == treescan())
|
|
|
|
goto out;
|
2017-01-21 13:01:39 +00:00
|
|
|
dba = dba_new(128);
|
2014-01-31 19:59:03 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
if (OP_DELETE != op)
|
2017-01-21 13:01:39 +00:00
|
|
|
mpages_merge(dba, mp);
|
|
|
|
if (nodb == 0)
|
|
|
|
dbwrite(dba);
|
|
|
|
dba_free(dba);
|
2014-11-22 18:08:25 +00:00
|
|
|
} else {
|
2012-10-18 09:55:16 +00:00
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* If we have arguments, use them as our manpaths.
|
2017-01-21 13:01:39 +00:00
|
|
|
* If we don't, use man.conf(5).
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
2014-11-22 18:08:25 +00:00
|
|
|
if (argc > 0) {
|
2016-01-15 23:08:59 +00:00
|
|
|
conf.manpath.paths = mandoc_reallocarray(NULL,
|
2014-11-22 18:08:25 +00:00
|
|
|
argc, sizeof(char *));
|
2016-01-15 23:08:59 +00:00
|
|
|
conf.manpath.sz = (size_t)argc;
|
2014-11-22 18:08:25 +00:00
|
|
|
for (i = 0; i < argc; i++)
|
2016-01-15 23:08:59 +00:00
|
|
|
conf.manpath.paths[i] = mandoc_strdup(argv[i]);
|
2014-11-22 18:08:25 +00:00
|
|
|
} else
|
2016-01-15 23:08:59 +00:00
|
|
|
manconf_parse(&conf, path_arg, NULL, NULL);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if (conf.manpath.sz == 0) {
|
2014-11-22 18:08:25 +00:00
|
|
|
exitcode = (int)MANDOCLEVEL_BADARG;
|
|
|
|
say("", "Empty manpath");
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* First scan the tree rooted at a base directory, then
|
|
|
|
* build a new database and finally move it into place.
|
|
|
|
* Ignore zero-length directories and strip trailing
|
|
|
|
* slashes.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
2016-01-15 23:08:59 +00:00
|
|
|
for (j = 0; j < conf.manpath.sz; j++) {
|
|
|
|
sz = strlen(conf.manpath.paths[j]);
|
|
|
|
if (sz && conf.manpath.paths[j][sz - 1] == '/')
|
|
|
|
conf.manpath.paths[j][--sz] = '\0';
|
2014-11-22 18:08:25 +00:00
|
|
|
if (0 == sz)
|
|
|
|
continue;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (j) {
|
2016-01-15 23:08:59 +00:00
|
|
|
mandoc_ohash_init(&mpages, 6,
|
|
|
|
offsetof(struct mpage, inodev));
|
|
|
|
mandoc_ohash_init(&mlinks, 6,
|
|
|
|
offsetof(struct mlink, file));
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if ( ! set_basedir(conf.manpath.paths[j], argc > 0))
|
2014-12-02 07:34:06 +00:00
|
|
|
continue;
|
2014-11-22 18:08:25 +00:00
|
|
|
if (0 == treescan())
|
2014-12-02 07:34:06 +00:00
|
|
|
continue;
|
2017-01-21 13:01:39 +00:00
|
|
|
dba = dba_new(128);
|
|
|
|
mpages_merge(dba, mp);
|
|
|
|
if (nodb == 0)
|
|
|
|
dbwrite(dba);
|
|
|
|
dba_free(dba);
|
2014-01-31 19:59:03 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if (j + 1 < conf.manpath.sz) {
|
2014-11-22 18:08:25 +00:00
|
|
|
mpages_free();
|
|
|
|
ohash_delete(&mpages);
|
|
|
|
ohash_delete(&mlinks);
|
|
|
|
}
|
2014-01-31 19:59:03 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
out:
|
2016-01-15 23:08:59 +00:00
|
|
|
manconf_free(&conf);
|
2014-11-22 18:08:25 +00:00
|
|
|
mparse_free(mp);
|
2016-01-15 23:08:59 +00:00
|
|
|
mchars_free();
|
2014-11-22 18:08:25 +00:00
|
|
|
mpages_free();
|
|
|
|
ohash_delete(&mpages);
|
|
|
|
ohash_delete(&mlinks);
|
2016-01-15 23:08:59 +00:00
|
|
|
return exitcode;
|
2012-10-18 09:55:16 +00:00
|
|
|
usage:
|
2016-01-15 23:08:59 +00:00
|
|
|
progname = getprogname();
|
2014-11-22 18:08:25 +00:00
|
|
|
fprintf(stderr, "usage: %s [-aDnpQ] [-C file] [-Tutf8]\n"
|
|
|
|
" %s [-aDnpQ] [-Tutf8] dir ...\n"
|
|
|
|
" %s [-DnpQ] [-Tutf8] -d dir [file ...]\n"
|
|
|
|
" %s [-Dnp] -u dir [file ...]\n"
|
|
|
|
" %s [-Q] -t file ...\n",
|
2016-01-15 23:08:59 +00:00
|
|
|
progname, progname, progname, progname, progname);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
return (int)MANDOCLEVEL_BADARG;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
/*
|
|
|
|
* To get a singly linked list in alpha order while inserting entries
|
|
|
|
* at the beginning, process directory entries in reverse alpha order.
|
|
|
|
*/
|
|
|
|
static int
|
|
|
|
#if HAVE_FTS_COMPARE_CONST
|
|
|
|
fts_compare(const FTSENT *const *a, const FTSENT *const *b)
|
|
|
|
#else
|
|
|
|
fts_compare(const FTSENT **a, const FTSENT **b)
|
|
|
|
#endif
|
|
|
|
{
|
|
|
|
return -strcmp((*a)->fts_name, (*b)->fts_name);
|
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Scan a directory tree rooted at "basedir" for manpages.
|
|
|
|
* We use fts(), scanning directory parts along the way for clues to our
|
|
|
|
* section and architecture.
|
|
|
|
*
|
|
|
|
* If use_all has been specified, grok all files.
|
|
|
|
* If not, sanitise paths to the following:
|
|
|
|
*
|
|
|
|
* [./]man*[/<arch>]/<name>.<section>
|
|
|
|
* or
|
|
|
|
* [./]cat<section>[/<arch>]/<name>.0
|
|
|
|
*
|
2016-07-23 11:45:42 +00:00
|
|
|
* TODO: accommodate for multi-language directories.
|
2014-11-22 18:08:25 +00:00
|
|
|
*/
|
|
|
|
static int
|
|
|
|
treescan(void)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
char buf[PATH_MAX];
|
|
|
|
FTS *f;
|
|
|
|
FTSENT *ff;
|
|
|
|
struct mlink *mlink;
|
2017-01-21 13:01:39 +00:00
|
|
|
int gzip;
|
|
|
|
enum form dform;
|
2014-11-22 18:08:25 +00:00
|
|
|
char *dsec, *arch, *fsec, *cp;
|
|
|
|
const char *path;
|
|
|
|
const char *argv[2];
|
|
|
|
|
|
|
|
argv[0] = ".";
|
|
|
|
argv[1] = (char *)NULL;
|
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
f = fts_open((char * const *)argv, FTS_PHYSICAL | FTS_NOCHDIR,
|
|
|
|
fts_compare);
|
2016-01-15 23:08:59 +00:00
|
|
|
if (f == NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "&fts_open");
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
dsec = arch = NULL;
|
|
|
|
dform = FORM_NONE;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
while ((ff = fts_read(f)) != NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
path = ff->fts_path + 2;
|
|
|
|
switch (ff->fts_info) {
|
2012-10-18 09:55:16 +00:00
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* Symbolic links require various sanity checks,
|
|
|
|
* then get handled just like regular files.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
2014-11-22 18:08:25 +00:00
|
|
|
case FTS_SL:
|
2016-01-15 23:08:59 +00:00
|
|
|
if (realpath(path, buf) == NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings)
|
|
|
|
say(path, "&realpath");
|
|
|
|
continue;
|
|
|
|
}
|
2015-03-02 16:45:41 +00:00
|
|
|
if (strstr(buf, basedir) != buf
|
|
|
|
#ifdef HOMEBREWDIR
|
|
|
|
&& strstr(buf, HOMEBREWDIR) != buf
|
|
|
|
#endif
|
|
|
|
) {
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings) say("",
|
|
|
|
"%s: outside base directory", buf);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
/* Use logical inode to avoid mpages dupe. */
|
2016-01-15 23:08:59 +00:00
|
|
|
if (stat(path, ff->fts_statp) == -1) {
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings)
|
|
|
|
say(path, "&stat");
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
/* FALLTHROUGH */
|
2012-10-18 09:55:16 +00:00
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* If we're a regular file, add an mlink by using the
|
|
|
|
* stored directory data and handling the filename.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
2014-11-22 18:08:25 +00:00
|
|
|
case FTS_F:
|
2016-01-15 23:08:59 +00:00
|
|
|
if ( ! strcmp(path, MANDOC_DB))
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
|
|
|
if ( ! use_all && ff->fts_level < 2) {
|
|
|
|
if (warnings)
|
|
|
|
say(path, "Extraneous file");
|
|
|
|
continue;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
gzip = 0;
|
|
|
|
fsec = NULL;
|
2016-01-15 23:08:59 +00:00
|
|
|
while (fsec == NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
fsec = strrchr(ff->fts_name, '.');
|
2016-01-15 23:08:59 +00:00
|
|
|
if (fsec == NULL || strcmp(fsec+1, "gz"))
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
|
|
|
gzip = 1;
|
|
|
|
*fsec = '\0';
|
|
|
|
fsec = NULL;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2016-01-15 23:08:59 +00:00
|
|
|
if (fsec == NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
if ( ! use_all) {
|
|
|
|
if (warnings)
|
|
|
|
say(path,
|
|
|
|
"No filename suffix");
|
|
|
|
continue;
|
|
|
|
}
|
2016-01-15 23:08:59 +00:00
|
|
|
} else if ( ! strcmp(++fsec, "html")) {
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings)
|
|
|
|
say(path, "Skip html");
|
|
|
|
continue;
|
2016-01-15 23:08:59 +00:00
|
|
|
} else if ( ! strcmp(fsec, "ps")) {
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings)
|
|
|
|
say(path, "Skip ps");
|
|
|
|
continue;
|
2016-01-15 23:08:59 +00:00
|
|
|
} else if ( ! strcmp(fsec, "pdf")) {
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings)
|
|
|
|
say(path, "Skip pdf");
|
|
|
|
continue;
|
|
|
|
} else if ( ! use_all &&
|
2016-01-15 23:08:59 +00:00
|
|
|
((dform == FORM_SRC &&
|
2015-03-02 16:45:41 +00:00
|
|
|
strncmp(fsec, dsec, strlen(dsec))) ||
|
2016-01-15 23:08:59 +00:00
|
|
|
(dform == FORM_CAT && strcmp(fsec, "0")))) {
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings)
|
|
|
|
say(path, "Wrong filename suffix");
|
|
|
|
continue;
|
|
|
|
} else
|
|
|
|
fsec[-1] = '\0';
|
|
|
|
|
|
|
|
mlink = mandoc_calloc(1, sizeof(struct mlink));
|
|
|
|
if (strlcpy(mlink->file, path,
|
|
|
|
sizeof(mlink->file)) >=
|
|
|
|
sizeof(mlink->file)) {
|
|
|
|
say(path, "Filename too long");
|
|
|
|
free(mlink);
|
|
|
|
continue;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
mlink->dform = dform;
|
|
|
|
mlink->dsec = dsec;
|
|
|
|
mlink->arch = arch;
|
|
|
|
mlink->name = ff->fts_name;
|
|
|
|
mlink->fsec = fsec;
|
|
|
|
mlink->gzip = gzip;
|
|
|
|
mlink_add(mlink, ff->fts_statp);
|
|
|
|
continue;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
case FTS_D:
|
|
|
|
case FTS_DP:
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
if (warnings)
|
|
|
|
say(path, "Not a regular file");
|
2012-10-18 09:55:16 +00:00
|
|
|
continue;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
switch (ff->fts_level) {
|
|
|
|
case 0:
|
|
|
|
/* Ignore the root directory. */
|
|
|
|
break;
|
|
|
|
case 1:
|
|
|
|
/*
|
|
|
|
* This might contain manX/ or catX/.
|
|
|
|
* Try to infer this from the name.
|
|
|
|
* If we're not in use_all, enforce it.
|
|
|
|
*/
|
|
|
|
cp = ff->fts_name;
|
2016-01-15 23:08:59 +00:00
|
|
|
if (ff->fts_info == FTS_DP) {
|
|
|
|
dform = FORM_NONE;
|
|
|
|
dsec = NULL;
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
2016-01-15 23:08:59 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if ( ! strncmp(cp, "man", 3)) {
|
2014-11-22 18:08:25 +00:00
|
|
|
dform = FORM_SRC;
|
|
|
|
dsec = cp + 3;
|
2016-01-15 23:08:59 +00:00
|
|
|
} else if ( ! strncmp(cp, "cat", 3)) {
|
2014-11-22 18:08:25 +00:00
|
|
|
dform = FORM_CAT;
|
|
|
|
dsec = cp + 3;
|
|
|
|
} else {
|
|
|
|
dform = FORM_NONE;
|
|
|
|
dsec = NULL;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if (dsec != NULL || use_all)
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings)
|
|
|
|
say(path, "Unknown directory part");
|
|
|
|
fts_set(f, ff, FTS_SKIP);
|
|
|
|
break;
|
|
|
|
case 2:
|
|
|
|
/*
|
|
|
|
* Possibly our architecture.
|
|
|
|
* If we're descending, keep tabs on it.
|
|
|
|
*/
|
2016-01-15 23:08:59 +00:00
|
|
|
if (ff->fts_info != FTS_DP && dsec != NULL)
|
2014-11-22 18:08:25 +00:00
|
|
|
arch = ff->fts_name;
|
|
|
|
else
|
|
|
|
arch = NULL;
|
|
|
|
break;
|
|
|
|
default:
|
2016-01-15 23:08:59 +00:00
|
|
|
if (ff->fts_info == FTS_DP || use_all)
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
|
|
|
if (warnings)
|
|
|
|
say(path, "Extraneous directory part");
|
|
|
|
fts_set(f, ff, FTS_SKIP);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
fts_close(f);
|
2016-01-15 23:08:59 +00:00
|
|
|
return 1;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Add a file to the mlinks table.
|
|
|
|
* Do not verify that it's a "valid" looking manpage (we'll do that
|
|
|
|
* later).
|
|
|
|
*
|
|
|
|
* Try to infer the manual section, architecture, and page name from the
|
|
|
|
* path, assuming it looks like
|
|
|
|
*
|
|
|
|
* [./]man*[/<arch>]/<name>.<section>
|
|
|
|
* or
|
|
|
|
* [./]cat<section>[/<arch>]/<name>.0
|
|
|
|
*
|
|
|
|
* See treescan() for the fts(3) version of this.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
filescan(const char *file)
|
|
|
|
{
|
|
|
|
char buf[PATH_MAX];
|
|
|
|
struct stat st;
|
|
|
|
struct mlink *mlink;
|
|
|
|
char *p, *start;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
assert(use_all);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (0 == strncmp(file, "./", 2))
|
|
|
|
file += 2;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* We have to do lstat(2) before realpath(3) loses
|
|
|
|
* the information whether this is a symbolic link.
|
|
|
|
* We need to know that because for symbolic links,
|
|
|
|
* we want to use the orginal file name, while for
|
|
|
|
* regular files, we want to use the real path.
|
|
|
|
*/
|
|
|
|
if (-1 == lstat(file, &st)) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_BADARG;
|
|
|
|
say(file, "&lstat");
|
|
|
|
return;
|
|
|
|
} else if (0 == ((S_IFREG | S_IFLNK) & st.st_mode)) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_BADARG;
|
|
|
|
say(file, "Not a regular file");
|
|
|
|
return;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* We have to resolve the file name to the real path
|
|
|
|
* in any case for the base directory check.
|
|
|
|
*/
|
|
|
|
if (NULL == realpath(file, buf)) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_BADARG;
|
|
|
|
say(file, "&realpath");
|
|
|
|
return;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (OP_TEST == op)
|
|
|
|
start = buf;
|
|
|
|
else if (strstr(buf, basedir) == buf)
|
|
|
|
start = buf + strlen(basedir);
|
2015-03-02 16:45:41 +00:00
|
|
|
#ifdef HOMEBREWDIR
|
|
|
|
else if (strstr(buf, HOMEBREWDIR) == buf)
|
|
|
|
start = buf;
|
|
|
|
#endif
|
2014-11-22 18:08:25 +00:00
|
|
|
else {
|
|
|
|
exitcode = (int)MANDOCLEVEL_BADARG;
|
|
|
|
say("", "%s: outside base directory", buf);
|
|
|
|
return;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Now we are sure the file is inside our tree.
|
|
|
|
* If it is a symbolic link, ignore the real path
|
|
|
|
* and use the original name.
|
|
|
|
* This implies passing stuff like "cat1/../man1/foo.1"
|
|
|
|
* on the command line won't work. So don't do that.
|
|
|
|
* Note the stat(2) can still fail if the link target
|
|
|
|
* doesn't exist.
|
|
|
|
*/
|
|
|
|
if (S_IFLNK & st.st_mode) {
|
|
|
|
if (-1 == stat(buf, &st)) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_BADARG;
|
|
|
|
say(file, "&stat");
|
|
|
|
return;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
if (strlcpy(buf, file, sizeof(buf)) >= sizeof(buf)) {
|
|
|
|
say(file, "Filename too long");
|
|
|
|
return;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
start = buf;
|
|
|
|
if (OP_TEST != op && strstr(buf, basedir) == buf)
|
|
|
|
start += strlen(basedir);
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
mlink = mandoc_calloc(1, sizeof(struct mlink));
|
2014-12-02 07:34:06 +00:00
|
|
|
mlink->dform = FORM_NONE;
|
2014-11-22 18:08:25 +00:00
|
|
|
if (strlcpy(mlink->file, start, sizeof(mlink->file)) >=
|
|
|
|
sizeof(mlink->file)) {
|
|
|
|
say(start, "Filename too long");
|
2015-03-02 16:45:41 +00:00
|
|
|
free(mlink);
|
2014-11-22 18:08:25 +00:00
|
|
|
return;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* First try to guess our directory structure.
|
|
|
|
* If we find a separator, try to look for man* or cat*.
|
|
|
|
* If we find one of these and what's underneath is a directory,
|
|
|
|
* assume it's an architecture.
|
|
|
|
*/
|
|
|
|
if (NULL != (p = strchr(start, '/'))) {
|
|
|
|
*p++ = '\0';
|
|
|
|
if (0 == strncmp(start, "man", 3)) {
|
|
|
|
mlink->dform = FORM_SRC;
|
|
|
|
mlink->dsec = start + 3;
|
|
|
|
} else if (0 == strncmp(start, "cat", 3)) {
|
|
|
|
mlink->dform = FORM_CAT;
|
|
|
|
mlink->dsec = start + 3;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
start = p;
|
|
|
|
if (NULL != mlink->dsec && NULL != (p = strchr(start, '/'))) {
|
|
|
|
*p++ = '\0';
|
|
|
|
mlink->arch = start;
|
|
|
|
start = p;
|
|
|
|
}
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Now check the file suffix.
|
|
|
|
* Suffix of `.0' indicates a catpage, `.1-9' is a manpage.
|
|
|
|
*/
|
|
|
|
p = strrchr(start, '\0');
|
|
|
|
while (p-- > start && '/' != *p && '.' != *p)
|
|
|
|
/* Loop. */ ;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if ('.' == *p) {
|
|
|
|
*p++ = '\0';
|
|
|
|
mlink->fsec = p;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* Now try to parse the name.
|
|
|
|
* Use the filename portion of the path.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
2014-11-22 18:08:25 +00:00
|
|
|
mlink->name = start;
|
|
|
|
if (NULL != (p = strrchr(start, '/'))) {
|
|
|
|
mlink->name = p + 1;
|
|
|
|
*p = '\0';
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
mlink_add(mlink, &st);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2014-11-22 18:08:25 +00:00
|
|
|
mlink_add(struct mlink *mlink, const struct stat *st)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
struct inodev inodev;
|
|
|
|
struct mpage *mpage;
|
|
|
|
unsigned int slot;
|
|
|
|
|
|
|
|
assert(NULL != mlink->file);
|
|
|
|
|
|
|
|
mlink->dsec = mandoc_strdup(mlink->dsec ? mlink->dsec : "");
|
|
|
|
mlink->arch = mandoc_strdup(mlink->arch ? mlink->arch : "");
|
|
|
|
mlink->name = mandoc_strdup(mlink->name ? mlink->name : "");
|
|
|
|
mlink->fsec = mandoc_strdup(mlink->fsec ? mlink->fsec : "");
|
|
|
|
|
|
|
|
if ('0' == *mlink->fsec) {
|
|
|
|
free(mlink->fsec);
|
|
|
|
mlink->fsec = mandoc_strdup(mlink->dsec);
|
|
|
|
mlink->fform = FORM_CAT;
|
|
|
|
} else if ('1' <= *mlink->fsec && '9' >= *mlink->fsec)
|
|
|
|
mlink->fform = FORM_SRC;
|
|
|
|
else
|
|
|
|
mlink->fform = FORM_NONE;
|
|
|
|
|
|
|
|
slot = ohash_qlookup(&mlinks, mlink->file);
|
|
|
|
assert(NULL == ohash_find(&mlinks, slot));
|
|
|
|
ohash_insert(&mlinks, slot, mlink);
|
|
|
|
|
2015-03-15 11:28:48 +00:00
|
|
|
memset(&inodev, 0, sizeof(inodev)); /* Clear padding. */
|
2014-11-22 18:08:25 +00:00
|
|
|
inodev.st_ino = st->st_ino;
|
|
|
|
inodev.st_dev = st->st_dev;
|
|
|
|
slot = ohash_lookup_memory(&mpages, (char *)&inodev,
|
|
|
|
sizeof(struct inodev), inodev.st_ino);
|
|
|
|
mpage = ohash_find(&mpages, slot);
|
|
|
|
if (NULL == mpage) {
|
|
|
|
mpage = mandoc_calloc(1, sizeof(struct mpage));
|
|
|
|
mpage->inodev.st_ino = inodev.st_ino;
|
|
|
|
mpage->inodev.st_dev = inodev.st_dev;
|
2017-01-21 13:01:39 +00:00
|
|
|
mpage->form = FORM_NONE;
|
|
|
|
mpage->next = mpage_head;
|
|
|
|
mpage_head = mpage;
|
2014-11-22 18:08:25 +00:00
|
|
|
ohash_insert(&mpages, slot, mpage);
|
|
|
|
} else
|
|
|
|
mlink->next = mpage->mlinks;
|
|
|
|
mpage->mlinks = mlink;
|
|
|
|
mlink->mpage = mpage;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static void
|
|
|
|
mlink_free(struct mlink *mlink)
|
|
|
|
{
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
free(mlink->dsec);
|
|
|
|
free(mlink->arch);
|
|
|
|
free(mlink->name);
|
|
|
|
free(mlink->fsec);
|
|
|
|
free(mlink);
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static void
|
|
|
|
mpages_free(void)
|
|
|
|
{
|
|
|
|
struct mpage *mpage;
|
|
|
|
struct mlink *mlink;
|
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
while ((mpage = mpage_head) != NULL) {
|
|
|
|
while ((mlink = mpage->mlinks) != NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
mpage->mlinks = mlink->next;
|
|
|
|
mlink_free(mlink);
|
|
|
|
}
|
2017-01-21 13:01:39 +00:00
|
|
|
mpage_head = mpage->next;
|
2014-11-22 18:08:25 +00:00
|
|
|
free(mpage->sec);
|
|
|
|
free(mpage->arch);
|
|
|
|
free(mpage->title);
|
|
|
|
free(mpage->desc);
|
|
|
|
free(mpage);
|
|
|
|
}
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* For each mlink to the mpage, check whether the path looks like
|
|
|
|
* it is formatted, and if it does, check whether a source manual
|
|
|
|
* exists by the same name, ignoring the suffix.
|
|
|
|
* If both conditions hold, drop the mlink.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
mlinks_undupe(struct mpage *mpage)
|
|
|
|
{
|
|
|
|
char buf[PATH_MAX];
|
|
|
|
struct mlink **prev;
|
|
|
|
struct mlink *mlink;
|
|
|
|
char *bufp;
|
|
|
|
|
|
|
|
mpage->form = FORM_CAT;
|
|
|
|
prev = &mpage->mlinks;
|
|
|
|
while (NULL != (mlink = *prev)) {
|
|
|
|
if (FORM_CAT != mlink->dform) {
|
|
|
|
mpage->form = FORM_NONE;
|
|
|
|
goto nextlink;
|
|
|
|
}
|
|
|
|
(void)strlcpy(buf, mlink->file, sizeof(buf));
|
|
|
|
bufp = strstr(buf, "cat");
|
|
|
|
assert(NULL != bufp);
|
|
|
|
memcpy(bufp, "man", 3);
|
|
|
|
if (NULL != (bufp = strrchr(buf, '.')))
|
|
|
|
*++bufp = '\0';
|
|
|
|
(void)strlcat(buf, mlink->dsec, sizeof(buf));
|
|
|
|
if (NULL == ohash_find(&mlinks,
|
|
|
|
ohash_qlookup(&mlinks, buf)))
|
|
|
|
goto nextlink;
|
|
|
|
if (warnings)
|
|
|
|
say(mlink->file, "Man source exists: %s", buf);
|
|
|
|
if (use_all)
|
|
|
|
goto nextlink;
|
|
|
|
*prev = mlink->next;
|
|
|
|
mlink_free(mlink);
|
|
|
|
continue;
|
|
|
|
nextlink:
|
|
|
|
prev = &(*prev)->next;
|
|
|
|
}
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static void
|
|
|
|
mlink_check(struct mpage *mpage, struct mlink *mlink)
|
|
|
|
{
|
|
|
|
struct str *str;
|
|
|
|
unsigned int slot;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Check whether the manual section given in a file
|
|
|
|
* agrees with the directory where the file is located.
|
|
|
|
* Some manuals have suffixes like (3p) on their
|
|
|
|
* section number either inside the file or in the
|
|
|
|
* directory name, some are linked into more than one
|
|
|
|
* section, like encrypt(1) = makekey(8).
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (FORM_SRC == mpage->form &&
|
|
|
|
strcasecmp(mpage->sec, mlink->dsec))
|
|
|
|
say(mlink->file, "Section \"%s\" manual in %s directory",
|
|
|
|
mpage->sec, mlink->dsec);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Manual page directories exist for each kernel
|
|
|
|
* architecture as returned by machine(1).
|
|
|
|
* However, many manuals only depend on the
|
|
|
|
* application architecture as returned by arch(1).
|
|
|
|
* For example, some (2/ARM) manuals are shared
|
|
|
|
* across the "armish" and "zaurus" kernel
|
|
|
|
* architectures.
|
|
|
|
* A few manuals are even shared across completely
|
|
|
|
* different architectures, for example fdformat(1)
|
2017-01-21 13:01:39 +00:00
|
|
|
* on amd64, i386, and sparc64.
|
2014-11-22 18:08:25 +00:00
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (strcasecmp(mpage->arch, mlink->arch))
|
|
|
|
say(mlink->file, "Architecture \"%s\" manual in "
|
|
|
|
"\"%s\" directory", mpage->arch, mlink->arch);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* XXX
|
|
|
|
* parse_cat() doesn't set NAME_TITLE yet.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (FORM_CAT == mpage->form)
|
|
|
|
return;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Check whether this mlink
|
|
|
|
* appears as a name in the NAME section.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
slot = ohash_qlookup(&names, mlink->name);
|
|
|
|
str = ohash_find(&names, slot);
|
|
|
|
assert(NULL != str);
|
|
|
|
if ( ! (NAME_TITLE & str->mask))
|
|
|
|
say(mlink->file, "Name missing in NAME section");
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* Run through the files in the global vector "mpages"
|
|
|
|
* and add them to the database specified in "basedir".
|
|
|
|
*
|
|
|
|
* This handles the parsing scheme itself, using the cues of directory
|
|
|
|
* and filename to determine whether the file is parsable or not.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
|
|
|
static void
|
2017-01-21 13:01:39 +00:00
|
|
|
mpages_merge(struct dba *dba, struct mparse *mp)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
struct mpage *mpage, *mpage_dest;
|
|
|
|
struct mlink *mlink, *mlink_dest;
|
2016-01-15 23:08:59 +00:00
|
|
|
struct roff_man *man;
|
2014-11-22 18:08:25 +00:00
|
|
|
char *sodest;
|
|
|
|
char *cp;
|
2014-12-02 07:34:06 +00:00
|
|
|
int fd;
|
2014-11-22 18:08:25 +00:00
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
for (mpage = mpage_head; mpage != NULL; mpage = mpage->next) {
|
2014-11-22 18:08:25 +00:00
|
|
|
mlinks_undupe(mpage);
|
2017-01-21 13:01:39 +00:00
|
|
|
if ((mlink = mpage->mlinks) == NULL)
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
name_mask = NAME_MASK;
|
2016-01-15 23:08:59 +00:00
|
|
|
mandoc_ohash_init(&names, 4, offsetof(struct str, key));
|
|
|
|
mandoc_ohash_init(&strings, 6, offsetof(struct str, key));
|
2014-11-22 18:08:25 +00:00
|
|
|
mparse_reset(mp);
|
|
|
|
man = NULL;
|
|
|
|
sodest = NULL;
|
2014-12-02 07:34:06 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if ((fd = mparse_open(mp, mlink->file)) == -1) {
|
2015-03-02 16:45:41 +00:00
|
|
|
say(mlink->file, "&open");
|
2014-12-02 07:34:06 +00:00
|
|
|
goto nextpage;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
2015-03-02 16:45:41 +00:00
|
|
|
* Interpret the file as mdoc(7) or man(7) source
|
|
|
|
* code, unless it is known to be formatted.
|
2014-11-22 18:08:25 +00:00
|
|
|
*/
|
2015-03-02 16:45:41 +00:00
|
|
|
if (mlink->dform != FORM_CAT || mlink->fform != FORM_CAT) {
|
|
|
|
mparse_readfd(mp, fd, mlink->file);
|
2016-01-15 23:08:59 +00:00
|
|
|
close(fd);
|
|
|
|
mparse_result(mp, &man, &sodest);
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-02 07:34:06 +00:00
|
|
|
if (sodest != NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
mlink_dest = ohash_find(&mlinks,
|
|
|
|
ohash_qlookup(&mlinks, sodest));
|
2014-12-02 07:34:06 +00:00
|
|
|
if (mlink_dest == NULL) {
|
|
|
|
mandoc_asprintf(&cp, "%s.gz", sodest);
|
|
|
|
mlink_dest = ohash_find(&mlinks,
|
|
|
|
ohash_qlookup(&mlinks, cp));
|
|
|
|
free(cp);
|
|
|
|
}
|
|
|
|
if (mlink_dest != NULL) {
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/* The .so target exists. */
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
mpage_dest = mlink_dest->mpage;
|
|
|
|
while (1) {
|
|
|
|
mlink->mpage = mpage_dest;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* If the target was already
|
|
|
|
* processed, add the links
|
|
|
|
* to the database now.
|
|
|
|
* Otherwise, this will
|
|
|
|
* happen when we come
|
|
|
|
* to the target.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
if (mpage_dest->dba != NULL)
|
|
|
|
dbadd_mlink(mlink);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-02 07:34:06 +00:00
|
|
|
if (mlink->next == NULL)
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
|
|
|
mlink = mlink->next;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/* Move all links to the target. */
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
mlink->next = mlink_dest->next;
|
|
|
|
mlink_dest->next = mpage->mlinks;
|
|
|
|
mpage->mlinks = NULL;
|
|
|
|
}
|
|
|
|
goto nextpage;
|
2016-01-15 23:08:59 +00:00
|
|
|
} else if (man != NULL && man->macroset == MACROSET_MDOC) {
|
|
|
|
mdoc_validate(man);
|
2014-11-22 18:08:25 +00:00
|
|
|
mpage->form = FORM_SRC;
|
2016-01-15 23:08:59 +00:00
|
|
|
mpage->sec = man->meta.msec;
|
2014-11-22 18:08:25 +00:00
|
|
|
mpage->sec = mandoc_strdup(
|
2014-12-02 07:34:06 +00:00
|
|
|
mpage->sec == NULL ? "" : mpage->sec);
|
2016-01-15 23:08:59 +00:00
|
|
|
mpage->arch = man->meta.arch;
|
2014-11-22 18:08:25 +00:00
|
|
|
mpage->arch = mandoc_strdup(
|
2014-12-02 07:34:06 +00:00
|
|
|
mpage->arch == NULL ? "" : mpage->arch);
|
2016-01-15 23:08:59 +00:00
|
|
|
mpage->title = mandoc_strdup(man->meta.title);
|
|
|
|
} else if (man != NULL && man->macroset == MACROSET_MAN) {
|
|
|
|
man_validate(man);
|
2014-11-22 18:08:25 +00:00
|
|
|
mpage->form = FORM_SRC;
|
2016-01-15 23:08:59 +00:00
|
|
|
mpage->sec = mandoc_strdup(man->meta.msec);
|
2015-03-02 16:45:41 +00:00
|
|
|
mpage->arch = mandoc_strdup(mlink->arch);
|
2016-01-15 23:08:59 +00:00
|
|
|
mpage->title = mandoc_strdup(man->meta.title);
|
2014-11-22 18:08:25 +00:00
|
|
|
} else {
|
|
|
|
mpage->form = FORM_CAT;
|
2015-03-02 16:45:41 +00:00
|
|
|
mpage->sec = mandoc_strdup(mlink->dsec);
|
|
|
|
mpage->arch = mandoc_strdup(mlink->arch);
|
|
|
|
mpage->title = mandoc_strdup(mlink->name);
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
assert(mpage->desc == NULL);
|
2016-01-15 23:08:59 +00:00
|
|
|
if (man != NULL && man->macroset == MACROSET_MDOC)
|
|
|
|
parse_mdoc(mpage, &man->meta, man->first);
|
2014-12-25 21:33:25 +00:00
|
|
|
else if (man != NULL)
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_man(mpage, &man->meta, man->first);
|
2014-11-22 18:08:25 +00:00
|
|
|
else
|
2014-12-02 07:34:06 +00:00
|
|
|
parse_cat(mpage, fd);
|
2014-12-25 21:33:25 +00:00
|
|
|
if (mpage->desc == NULL)
|
2014-11-22 18:08:25 +00:00
|
|
|
mpage->desc = mandoc_strdup(mpage->mlinks->name);
|
|
|
|
|
|
|
|
if (warnings && !use_all)
|
|
|
|
for (mlink = mpage->mlinks; mlink;
|
|
|
|
mlink = mlink->next)
|
|
|
|
mlink_check(mpage, mlink);
|
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
dbadd(dba, mpage);
|
2015-03-02 16:45:41 +00:00
|
|
|
mlink = mpage->mlinks;
|
2014-11-22 18:08:25 +00:00
|
|
|
|
|
|
|
nextpage:
|
|
|
|
ohash_delete(&strings);
|
|
|
|
ohash_delete(&names);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static void
|
|
|
|
parse_cat(struct mpage *mpage, int fd)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
FILE *stream;
|
|
|
|
char *line, *p, *title;
|
2016-01-15 23:08:59 +00:00
|
|
|
size_t linesz, plen, titlesz;
|
|
|
|
ssize_t len;
|
|
|
|
int offs;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
stream = (-1 == fd) ?
|
|
|
|
fopen(mpage->mlinks->file, "r") :
|
|
|
|
fdopen(fd, "r");
|
|
|
|
if (NULL == stream) {
|
2014-12-02 07:34:06 +00:00
|
|
|
if (-1 != fd)
|
|
|
|
close(fd);
|
2014-11-22 18:08:25 +00:00
|
|
|
if (warnings)
|
|
|
|
say(mpage->mlinks->file, "&fopen");
|
|
|
|
return;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
line = NULL;
|
|
|
|
linesz = 0;
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/* Skip to first blank line. */
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
while (getline(&line, &linesz, stream) != -1)
|
|
|
|
if (*line == '\n')
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Assume the first line that is not indented
|
|
|
|
* is the first section header. Skip to it.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
while (getline(&line, &linesz, stream) != -1)
|
|
|
|
if (*line != '\n' && *line != ' ')
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* Read up until the next section into a buffer.
|
|
|
|
* Strip the leading and trailing newline from each read line,
|
|
|
|
* appending a trailing space.
|
|
|
|
* Ignore empty (whitespace-only) lines.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
titlesz = 0;
|
|
|
|
title = NULL;
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
while ((len = getline(&line, &linesz, stream)) != -1) {
|
|
|
|
if (*line != ' ')
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
2016-01-15 23:08:59 +00:00
|
|
|
offs = 0;
|
|
|
|
while (isspace((unsigned char)line[offs]))
|
|
|
|
offs++;
|
|
|
|
if (line[offs] == '\0')
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
2016-01-15 23:08:59 +00:00
|
|
|
title = mandoc_realloc(title, titlesz + len - offs);
|
|
|
|
memcpy(title + titlesz, line + offs, len - offs);
|
|
|
|
titlesz += len - offs;
|
2014-11-22 18:08:25 +00:00
|
|
|
title[titlesz - 1] = ' ';
|
|
|
|
}
|
2016-01-15 23:08:59 +00:00
|
|
|
free(line);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* If no page content can be found, or the input line
|
|
|
|
* is already the next section header, or there is no
|
|
|
|
* trailing newline, reuse the page title as the page
|
|
|
|
* description.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (NULL == title || '\0' == *title) {
|
|
|
|
if (warnings)
|
|
|
|
say(mpage->mlinks->file,
|
|
|
|
"Cannot find NAME section");
|
|
|
|
fclose(stream);
|
|
|
|
free(title);
|
|
|
|
return;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
title[titlesz - 1] = '\0';
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Skip to the first dash.
|
|
|
|
* Use the remaining line as the description (no more than 70
|
|
|
|
* bytes).
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (NULL != (p = strstr(title, "- "))) {
|
|
|
|
for (p += 2; ' ' == *p || '\b' == *p; p++)
|
|
|
|
/* Skip to next word. */ ;
|
|
|
|
} else {
|
|
|
|
if (warnings)
|
|
|
|
say(mpage->mlinks->file,
|
|
|
|
"No dash in title line");
|
|
|
|
p = title;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
plen = strlen(p);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/* Strip backspace-encoding from line. */
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
while (NULL != (line = memchr(p, '\b', plen))) {
|
|
|
|
len = line - p;
|
|
|
|
if (0 == len) {
|
|
|
|
memmove(line, line + 1, plen--);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
memmove(line - 1, line + 1, plen - len);
|
|
|
|
plen -= 2;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
mpage->desc = mandoc_strdup(p);
|
|
|
|
fclose(stream);
|
|
|
|
free(title);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Put a type/word pair into the word database for this particular file.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
putkey(const struct mpage *mpage, char *value, uint64_t type)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
putkeys(mpage, value, strlen(value), type);
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Grok all nodes at or below a certain mdoc node into putkey().
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
putmdockey(const struct mpage *mpage,
|
2017-01-21 13:01:39 +00:00
|
|
|
const struct roff_node *n, uint64_t m, int taboo)
|
2014-11-22 18:08:25 +00:00
|
|
|
{
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
for ( ; NULL != n; n = n->next) {
|
2017-01-21 13:01:39 +00:00
|
|
|
if (n->flags & taboo)
|
|
|
|
continue;
|
2014-11-22 18:08:25 +00:00
|
|
|
if (NULL != n->child)
|
2017-01-21 13:01:39 +00:00
|
|
|
putmdockey(mpage, n->child, m, taboo);
|
2016-01-15 23:08:59 +00:00
|
|
|
if (n->type == ROFFT_TEXT)
|
2014-11-22 18:08:25 +00:00
|
|
|
putkey(mpage, n->string, m);
|
|
|
|
}
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static void
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_man(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2014-11-22 18:08:25 +00:00
|
|
|
{
|
2016-01-15 23:08:59 +00:00
|
|
|
const struct roff_node *head, *body;
|
2014-11-22 18:08:25 +00:00
|
|
|
char *start, *title;
|
|
|
|
char byte;
|
|
|
|
size_t sz;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if (n == NULL)
|
2014-11-22 18:08:25 +00:00
|
|
|
return;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* We're only searching for one thing: the first text child in
|
|
|
|
* the BODY of a NAME section. Since we don't keep track of
|
|
|
|
* sections in -man, run some hoops to find out whether we're in
|
|
|
|
* the correct section or not.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if (n->type == ROFFT_BODY && n->tok == MAN_SH) {
|
2014-11-22 18:08:25 +00:00
|
|
|
body = n;
|
2016-01-15 23:08:59 +00:00
|
|
|
if ((head = body->parent->head) != NULL &&
|
|
|
|
(head = head->child) != NULL &&
|
|
|
|
head->next == NULL &&
|
|
|
|
head->type == ROFFT_TEXT &&
|
|
|
|
strcmp(head->string, "NAME") == 0 &&
|
|
|
|
body->child != NULL) {
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Suck the entire NAME section into memory.
|
|
|
|
* Yes, we might run away.
|
|
|
|
* But too many manuals have big, spread-out
|
|
|
|
* NAME sections over many lines.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
title = NULL;
|
2016-01-15 23:08:59 +00:00
|
|
|
deroff(&title, body);
|
2014-11-22 18:08:25 +00:00
|
|
|
if (NULL == title)
|
|
|
|
return;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Go through a special heuristic dance here.
|
|
|
|
* Conventionally, one or more manual names are
|
|
|
|
* comma-specified prior to a whitespace, then a
|
|
|
|
* dash, then a description. Try to puzzle out
|
|
|
|
* the name parts here.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
start = title;
|
|
|
|
for ( ;; ) {
|
|
|
|
sz = strcspn(start, " ,");
|
|
|
|
if ('\0' == start[sz])
|
|
|
|
break;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
byte = start[sz];
|
|
|
|
start[sz] = '\0';
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Assume a stray trailing comma in the
|
|
|
|
* name list if a name begins with a dash.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if ('-' == start[0] ||
|
|
|
|
('\\' == start[0] && '-' == start[1]))
|
|
|
|
break;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
putkey(mpage, start, NAME_TITLE);
|
2014-12-25 21:33:25 +00:00
|
|
|
if ( ! (mpage->name_head_done ||
|
|
|
|
strcasecmp(start, meta->title))) {
|
|
|
|
putkey(mpage, start, NAME_HEAD);
|
|
|
|
mpage->name_head_done = 1;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (' ' == byte) {
|
|
|
|
start += sz + 1;
|
|
|
|
break;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
assert(',' == byte);
|
|
|
|
start += sz + 1;
|
|
|
|
while (' ' == *start)
|
|
|
|
start++;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (start == title) {
|
|
|
|
putkey(mpage, start, NAME_TITLE);
|
2014-12-25 21:33:25 +00:00
|
|
|
if ( ! (mpage->name_head_done ||
|
|
|
|
strcasecmp(start, meta->title))) {
|
|
|
|
putkey(mpage, start, NAME_HEAD);
|
|
|
|
mpage->name_head_done = 1;
|
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
free(title);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
while (isspace((unsigned char)*start))
|
|
|
|
start++;
|
|
|
|
|
|
|
|
if (0 == strncmp(start, "-", 1))
|
|
|
|
start += 1;
|
|
|
|
else if (0 == strncmp(start, "\\-\\-", 4))
|
|
|
|
start += 4;
|
|
|
|
else if (0 == strncmp(start, "\\-", 2))
|
|
|
|
start += 2;
|
|
|
|
else if (0 == strncmp(start, "\\(en", 4))
|
|
|
|
start += 4;
|
|
|
|
else if (0 == strncmp(start, "\\(em", 4))
|
|
|
|
start += 4;
|
|
|
|
|
|
|
|
while (' ' == *start)
|
|
|
|
start++;
|
|
|
|
|
|
|
|
mpage->desc = mandoc_strdup(start);
|
|
|
|
free(title);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for (n = n->child; n; n = n->next) {
|
|
|
|
if (NULL != mpage->desc)
|
|
|
|
break;
|
2014-12-25 21:33:25 +00:00
|
|
|
parse_man(mpage, meta, n);
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2014-11-22 18:08:25 +00:00
|
|
|
{
|
|
|
|
|
|
|
|
assert(NULL != n);
|
|
|
|
for (n = n->child; NULL != n; n = n->next) {
|
2017-01-21 13:01:39 +00:00
|
|
|
if (n->flags & mdocs[n->tok].taboo)
|
|
|
|
continue;
|
2014-11-22 18:08:25 +00:00
|
|
|
switch (n->type) {
|
2016-01-15 23:08:59 +00:00
|
|
|
case ROFFT_ELEM:
|
|
|
|
case ROFFT_BLOCK:
|
|
|
|
case ROFFT_HEAD:
|
|
|
|
case ROFFT_BODY:
|
|
|
|
case ROFFT_TAIL:
|
2014-11-22 18:08:25 +00:00
|
|
|
if (NULL != mdocs[n->tok].fp)
|
2014-12-25 21:33:25 +00:00
|
|
|
if (0 == (*mdocs[n->tok].fp)(mpage, meta, n))
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
|
|
|
if (mdocs[n->tok].mask)
|
|
|
|
putmdockey(mpage, n->child,
|
2017-01-21 13:01:39 +00:00
|
|
|
mdocs[n->tok].mask, mdocs[n->tok].taboo);
|
2014-11-22 18:08:25 +00:00
|
|
|
break;
|
|
|
|
default:
|
2016-01-15 23:08:59 +00:00
|
|
|
assert(n->type != ROFFT_ROOT);
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (NULL != n->child)
|
2014-12-25 21:33:25 +00:00
|
|
|
parse_mdoc(mpage, meta, n);
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_Fd(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-12-25 21:33:25 +00:00
|
|
|
char *start, *end;
|
2014-11-22 18:08:25 +00:00
|
|
|
size_t sz;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (SEC_SYNOPSIS != n->sec ||
|
|
|
|
NULL == (n = n->child) ||
|
2016-01-15 23:08:59 +00:00
|
|
|
n->type != ROFFT_TEXT)
|
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Only consider those `Fd' macro fields that begin with an
|
|
|
|
* "inclusion" token (versus, e.g., #define).
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (strcmp("#include", n->string))
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if ((n = n->next) == NULL || n->type != ROFFT_TEXT)
|
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Strip away the enclosing angle brackets and make sure we're
|
|
|
|
* not zero-length.
|
|
|
|
*/
|
|
|
|
|
|
|
|
start = n->string;
|
|
|
|
if ('<' == *start || '"' == *start)
|
|
|
|
start++;
|
|
|
|
|
|
|
|
if (0 == (sz = strlen(start)))
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
|
|
|
|
end = &start[(int)sz - 1];
|
|
|
|
if ('>' == *end || '"' == *end)
|
|
|
|
end--;
|
|
|
|
|
|
|
|
if (end > start)
|
|
|
|
putkeys(mpage, start, end - start + 1, TYPE_In);
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
static void
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_fname(struct mpage *mpage, const struct roff_node *n)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
char *cp;
|
2014-12-25 21:33:25 +00:00
|
|
|
size_t sz;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if (n->type != ROFFT_TEXT)
|
2014-12-25 21:33:25 +00:00
|
|
|
return;
|
2014-11-22 18:08:25 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
/* Skip function pointer punctuation. */
|
2014-11-22 18:08:25 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
cp = n->string;
|
|
|
|
while (*cp == '(' || *cp == '*')
|
2014-11-22 18:08:25 +00:00
|
|
|
cp++;
|
2014-12-25 21:33:25 +00:00
|
|
|
sz = strcspn(cp, "()");
|
|
|
|
|
|
|
|
putkeys(mpage, cp, sz, TYPE_Fn);
|
|
|
|
if (n->sec == SEC_SYNOPSIS)
|
|
|
|
putkeys(mpage, cp, sz, NAME_SYN);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_Fn(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2014-12-25 21:33:25 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
if (n->child == NULL)
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
parse_mdoc_fname(mpage, n->child);
|
2014-11-22 18:08:25 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
for (n = n->child->next; n != NULL; n = n->next)
|
2016-01-15 23:08:59 +00:00
|
|
|
if (n->type == ROFFT_TEXT)
|
2014-11-22 18:08:25 +00:00
|
|
|
putkey(mpage, n->string, TYPE_Fa);
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_Fo(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2014-12-25 21:33:25 +00:00
|
|
|
{
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if (n->type != ROFFT_HEAD)
|
|
|
|
return 1;
|
2014-12-25 21:33:25 +00:00
|
|
|
|
|
|
|
if (n->child != NULL)
|
|
|
|
parse_mdoc_fname(mpage, n->child);
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-12-25 21:33:25 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_Va(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
|
|
|
{
|
|
|
|
char *cp;
|
|
|
|
|
|
|
|
if (n->type != ROFFT_ELEM && n->type != ROFFT_BODY)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (n->child != NULL &&
|
|
|
|
n->child->next == NULL &&
|
|
|
|
n->child->type == ROFFT_TEXT)
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
cp = NULL;
|
|
|
|
deroff(&cp, n);
|
|
|
|
if (cp != NULL) {
|
|
|
|
putkey(mpage, cp, TYPE_Vt | (n->tok == MDOC_Va ||
|
|
|
|
n->type == ROFFT_BODY ? TYPE_Va : 0));
|
|
|
|
free(cp);
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
parse_mdoc_Xr(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
char *cp;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (NULL == (n = n->child))
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (NULL == n->next) {
|
|
|
|
putkey(mpage, n->string, TYPE_Xr);
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
mandoc_asprintf(&cp, "%s(%s)", n->string, n->next->string);
|
|
|
|
putkey(mpage, cp, TYPE_Xr);
|
|
|
|
free(cp);
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_Nd(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
if (n->type == ROFFT_BODY)
|
|
|
|
deroff(&mpage->desc, n);
|
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static int
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_Nm(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (SEC_NAME == n->sec)
|
2017-01-21 13:01:39 +00:00
|
|
|
putmdockey(mpage, n->child, NAME_TITLE, 0);
|
2016-01-15 23:08:59 +00:00
|
|
|
else if (n->sec == SEC_SYNOPSIS && n->type == ROFFT_HEAD) {
|
2014-12-25 21:33:25 +00:00
|
|
|
if (n->child == NULL)
|
|
|
|
putkey(mpage, meta->name, NAME_SYN);
|
|
|
|
else
|
2017-01-21 13:01:39 +00:00
|
|
|
putmdockey(mpage, n->child, NAME_SYN, 0);
|
2014-12-25 21:33:25 +00:00
|
|
|
}
|
|
|
|
if ( ! (mpage->name_head_done ||
|
|
|
|
n->child == NULL || n->child->string == NULL ||
|
|
|
|
strcasecmp(n->child->string, meta->title))) {
|
2017-01-21 13:01:39 +00:00
|
|
|
putkey(mpage, n->child->string, NAME_HEAD);
|
2014-12-25 21:33:25 +00:00
|
|
|
mpage->name_head_done = 1;
|
|
|
|
}
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static int
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_Sh(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2014-11-22 18:08:25 +00:00
|
|
|
{
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
return n->sec == SEC_CUSTOM && n->type == ROFFT_HEAD;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static int
|
2016-01-15 23:08:59 +00:00
|
|
|
parse_mdoc_head(struct mpage *mpage, const struct roff_meta *meta,
|
|
|
|
const struct roff_node *n)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
|
|
|
|
2016-01-15 23:08:59 +00:00
|
|
|
return n->type == ROFFT_HEAD;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* Add a string to the hash table for the current manual.
|
|
|
|
* Each string has a bitmask telling which macros it belongs to.
|
|
|
|
* When we finish the manual, we'll dump the table.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
|
|
|
static void
|
2014-12-25 21:33:25 +00:00
|
|
|
putkeys(const struct mpage *mpage, char *cp, size_t sz, uint64_t v)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
struct ohash *htab;
|
|
|
|
struct str *s;
|
|
|
|
const char *end;
|
|
|
|
unsigned int slot;
|
2014-12-25 21:33:25 +00:00
|
|
|
int i, mustfree;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (0 == sz)
|
2012-10-18 09:55:16 +00:00
|
|
|
return;
|
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
mustfree = render_string(&cp, &sz);
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (TYPE_Nm & v) {
|
|
|
|
htab = &names;
|
|
|
|
v &= name_mask;
|
2014-12-02 07:34:06 +00:00
|
|
|
if (v & NAME_FIRST)
|
|
|
|
name_mask &= ~NAME_FIRST;
|
2014-11-22 18:08:25 +00:00
|
|
|
if (debug > 1)
|
|
|
|
say(mpage->mlinks->file,
|
2017-01-21 13:01:39 +00:00
|
|
|
"Adding name %*s, bits=0x%llx", (int)sz, cp,
|
|
|
|
(unsigned long long)v);
|
2014-11-22 18:08:25 +00:00
|
|
|
} else {
|
|
|
|
htab = &strings;
|
|
|
|
if (debug > 1)
|
2017-01-21 13:01:39 +00:00
|
|
|
for (i = 0; i < KEY_MAX; i++)
|
2014-12-02 07:34:06 +00:00
|
|
|
if ((uint64_t)1 << i & v)
|
2014-11-22 18:08:25 +00:00
|
|
|
say(mpage->mlinks->file,
|
|
|
|
"Adding key %s=%*s",
|
2017-01-21 13:01:39 +00:00
|
|
|
mansearch_keynames[i], (int)sz, cp);
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
end = cp + sz;
|
|
|
|
slot = ohash_qlookupi(htab, cp, &end);
|
|
|
|
s = ohash_find(htab, slot);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (NULL != s && mpage == s->mpage) {
|
|
|
|
s->mask |= v;
|
|
|
|
return;
|
|
|
|
} else if (NULL == s) {
|
|
|
|
s = mandoc_calloc(1, sizeof(struct str) + sz + 1);
|
|
|
|
memcpy(s->key, cp, sz);
|
|
|
|
ohash_insert(htab, slot, s);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
s->mpage = mpage;
|
|
|
|
s->mask = v;
|
2014-12-25 21:33:25 +00:00
|
|
|
|
|
|
|
if (mustfree)
|
|
|
|
free(cp);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Take a Unicode codepoint and produce its UTF-8 encoding.
|
|
|
|
* This isn't the best way to do this, but it works.
|
|
|
|
* The magic numbers are from the UTF-8 packaging.
|
|
|
|
* They're not as scary as they seem: read the UTF-8 spec for details.
|
|
|
|
*/
|
|
|
|
static size_t
|
|
|
|
utf8(unsigned int cp, char out[7])
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
size_t rc;
|
|
|
|
|
|
|
|
rc = 0;
|
|
|
|
if (cp <= 0x0000007F) {
|
|
|
|
rc = 1;
|
|
|
|
out[0] = (char)cp;
|
|
|
|
} else if (cp <= 0x000007FF) {
|
|
|
|
rc = 2;
|
|
|
|
out[0] = (cp >> 6 & 31) | 192;
|
|
|
|
out[1] = (cp & 63) | 128;
|
|
|
|
} else if (cp <= 0x0000FFFF) {
|
|
|
|
rc = 3;
|
|
|
|
out[0] = (cp >> 12 & 15) | 224;
|
|
|
|
out[1] = (cp >> 6 & 63) | 128;
|
|
|
|
out[2] = (cp & 63) | 128;
|
|
|
|
} else if (cp <= 0x001FFFFF) {
|
|
|
|
rc = 4;
|
|
|
|
out[0] = (cp >> 18 & 7) | 240;
|
|
|
|
out[1] = (cp >> 12 & 63) | 128;
|
|
|
|
out[2] = (cp >> 6 & 63) | 128;
|
|
|
|
out[3] = (cp & 63) | 128;
|
|
|
|
} else if (cp <= 0x03FFFFFF) {
|
|
|
|
rc = 5;
|
|
|
|
out[0] = (cp >> 24 & 3) | 248;
|
|
|
|
out[1] = (cp >> 18 & 63) | 128;
|
|
|
|
out[2] = (cp >> 12 & 63) | 128;
|
|
|
|
out[3] = (cp >> 6 & 63) | 128;
|
|
|
|
out[4] = (cp & 63) | 128;
|
|
|
|
} else if (cp <= 0x7FFFFFFF) {
|
|
|
|
rc = 6;
|
|
|
|
out[0] = (cp >> 30 & 1) | 252;
|
|
|
|
out[1] = (cp >> 24 & 63) | 128;
|
|
|
|
out[2] = (cp >> 18 & 63) | 128;
|
|
|
|
out[3] = (cp >> 12 & 63) | 128;
|
|
|
|
out[4] = (cp >> 6 & 63) | 128;
|
|
|
|
out[5] = (cp & 63) | 128;
|
|
|
|
} else
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
out[rc] = '\0';
|
2016-01-15 23:08:59 +00:00
|
|
|
return rc;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
2014-12-25 21:33:25 +00:00
|
|
|
* If the string contains escape sequences,
|
|
|
|
* replace it with an allocated rendering and return 1,
|
|
|
|
* such that the caller can free it after use.
|
|
|
|
* Otherwise, do nothing and return 0.
|
2014-11-22 18:08:25 +00:00
|
|
|
*/
|
2014-12-25 21:33:25 +00:00
|
|
|
static int
|
|
|
|
render_string(char **public, size_t *psz)
|
2014-11-22 18:08:25 +00:00
|
|
|
{
|
2014-12-25 21:33:25 +00:00
|
|
|
const char *src, *scp, *addcp, *seq;
|
|
|
|
char *dst;
|
|
|
|
size_t ssz, dsz, addsz;
|
2014-11-22 18:08:25 +00:00
|
|
|
char utfbuf[7], res[6];
|
2014-12-25 21:33:25 +00:00
|
|
|
int seqlen, unicode;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
res[0] = '\\';
|
|
|
|
res[1] = '\t';
|
|
|
|
res[2] = ASCII_NBRSP;
|
|
|
|
res[3] = ASCII_HYPH;
|
|
|
|
res[4] = ASCII_BREAK;
|
|
|
|
res[5] = '\0';
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
src = scp = *public;
|
|
|
|
ssz = *psz;
|
|
|
|
dst = NULL;
|
|
|
|
dsz = 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
while (scp < src + *psz) {
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
/* Leave normal characters unchanged. */
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
if (strchr(res, *scp) == NULL) {
|
|
|
|
if (dst != NULL)
|
|
|
|
dst[dsz++] = *scp;
|
|
|
|
scp++;
|
|
|
|
continue;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
2014-12-25 21:33:25 +00:00
|
|
|
* Found something that requires replacing,
|
|
|
|
* make sure we have a destination buffer.
|
2014-11-22 18:08:25 +00:00
|
|
|
*/
|
2014-12-25 21:33:25 +00:00
|
|
|
|
|
|
|
if (dst == NULL) {
|
|
|
|
dst = mandoc_malloc(ssz + 1);
|
|
|
|
dsz = scp - src;
|
|
|
|
memcpy(dst, src, dsz);
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
/* Handle single-char special characters. */
|
|
|
|
|
|
|
|
switch (*scp) {
|
|
|
|
case '\\':
|
|
|
|
break;
|
2014-11-22 18:08:25 +00:00
|
|
|
case '\t':
|
|
|
|
case ASCII_NBRSP:
|
2014-12-25 21:33:25 +00:00
|
|
|
dst[dsz++] = ' ';
|
|
|
|
scp++;
|
|
|
|
continue;
|
|
|
|
case ASCII_HYPH:
|
|
|
|
dst[dsz++] = '-';
|
2014-11-22 18:08:25 +00:00
|
|
|
/* FALLTHROUGH */
|
|
|
|
case ASCII_BREAK:
|
2014-12-25 21:33:25 +00:00
|
|
|
scp++;
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
|
|
|
default:
|
2014-12-25 21:33:25 +00:00
|
|
|
abort();
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
2014-12-25 21:33:25 +00:00
|
|
|
* Found an escape sequence.
|
|
|
|
* Read past the slash, then parse it.
|
|
|
|
* Ignore everything except characters.
|
2014-11-22 18:08:25 +00:00
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
scp++;
|
|
|
|
if (mandoc_escape(&scp, &seq, &seqlen) != ESCAPE_SPECIAL)
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Render the special character
|
|
|
|
* as either UTF-8 or ASCII.
|
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (write_utf8) {
|
2016-01-15 23:08:59 +00:00
|
|
|
unicode = mchars_spec2cp(seq, seqlen);
|
2014-12-25 21:33:25 +00:00
|
|
|
if (unicode <= 0)
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
2014-12-25 21:33:25 +00:00
|
|
|
addsz = utf8(unicode, utfbuf);
|
|
|
|
if (addsz == 0)
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
2014-12-25 21:33:25 +00:00
|
|
|
addcp = utfbuf;
|
2014-11-22 18:08:25 +00:00
|
|
|
} else {
|
2016-01-15 23:08:59 +00:00
|
|
|
addcp = mchars_spec2str(seq, seqlen, &addsz);
|
2014-12-25 21:33:25 +00:00
|
|
|
if (addcp == NULL)
|
2014-11-22 18:08:25 +00:00
|
|
|
continue;
|
2014-12-25 21:33:25 +00:00
|
|
|
if (*addcp == ASCII_NBRSP) {
|
|
|
|
addcp = " ";
|
|
|
|
addsz = 1;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/* Copy the rendered glyph into the stream. */
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
ssz += addsz;
|
|
|
|
dst = mandoc_realloc(dst, ssz + 1);
|
|
|
|
memcpy(dst + dsz, addcp, addsz);
|
|
|
|
dsz += addsz;
|
|
|
|
}
|
|
|
|
if (dst != NULL) {
|
|
|
|
*public = dst;
|
|
|
|
*psz = dsz;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
/* Trim trailing whitespace and NUL-terminate. */
|
|
|
|
|
|
|
|
while (*psz > 0 && (*public)[*psz - 1] == ' ')
|
|
|
|
--*psz;
|
|
|
|
if (dst != NULL) {
|
|
|
|
(*public)[*psz] = '\0';
|
2016-01-15 23:08:59 +00:00
|
|
|
return 1;
|
2014-12-25 21:33:25 +00:00
|
|
|
} else
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static void
|
|
|
|
dbadd_mlink(const struct mlink *mlink)
|
|
|
|
{
|
2017-01-21 13:01:39 +00:00
|
|
|
dba_page_alias(mlink->mpage->dba, mlink->name, NAME_FILE);
|
|
|
|
dba_page_add(mlink->mpage->dba, DBP_SECT, mlink->dsec);
|
|
|
|
dba_page_add(mlink->mpage->dba, DBP_SECT, mlink->fsec);
|
|
|
|
dba_page_add(mlink->mpage->dba, DBP_ARCH, mlink->arch);
|
|
|
|
dba_page_add(mlink->mpage->dba, DBP_FILE, mlink->file);
|
2014-12-02 07:34:06 +00:00
|
|
|
}
|
|
|
|
|
2012-10-18 09:55:16 +00:00
|
|
|
/*
|
2014-11-22 18:08:25 +00:00
|
|
|
* Flush the current page's terms (and their bits) into the database.
|
|
|
|
* Also, handle escape sequences at the last possible moment.
|
2012-10-18 09:55:16 +00:00
|
|
|
*/
|
|
|
|
static void
|
2017-01-21 13:01:39 +00:00
|
|
|
dbadd(struct dba *dba, struct mpage *mpage)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
struct mlink *mlink;
|
|
|
|
struct str *key;
|
2014-12-25 21:33:25 +00:00
|
|
|
char *cp;
|
2017-01-21 13:01:39 +00:00
|
|
|
uint64_t mask;
|
2014-11-22 18:08:25 +00:00
|
|
|
size_t i;
|
|
|
|
unsigned int slot;
|
2014-12-25 21:33:25 +00:00
|
|
|
int mustfree;
|
2014-11-22 18:08:25 +00:00
|
|
|
|
|
|
|
mlink = mpage->mlinks;
|
|
|
|
|
|
|
|
if (nodb) {
|
|
|
|
for (key = ohash_first(&names, &slot); NULL != key;
|
2014-12-25 21:33:25 +00:00
|
|
|
key = ohash_next(&names, &slot))
|
2014-11-22 18:08:25 +00:00
|
|
|
free(key);
|
|
|
|
for (key = ohash_first(&strings, &slot); NULL != key;
|
2014-12-25 21:33:25 +00:00
|
|
|
key = ohash_next(&strings, &slot))
|
2014-11-22 18:08:25 +00:00
|
|
|
free(key);
|
|
|
|
if (0 == debug)
|
|
|
|
return;
|
|
|
|
while (NULL != mlink) {
|
|
|
|
fputs(mlink->name, stdout);
|
|
|
|
if (NULL == mlink->next ||
|
|
|
|
strcmp(mlink->dsec, mlink->next->dsec) ||
|
|
|
|
strcmp(mlink->fsec, mlink->next->fsec) ||
|
|
|
|
strcmp(mlink->arch, mlink->next->arch)) {
|
|
|
|
putchar('(');
|
|
|
|
if ('\0' == *mlink->dsec)
|
|
|
|
fputs(mlink->fsec, stdout);
|
|
|
|
else
|
|
|
|
fputs(mlink->dsec, stdout);
|
|
|
|
if ('\0' != *mlink->arch)
|
|
|
|
printf("/%s", mlink->arch);
|
|
|
|
putchar(')');
|
|
|
|
}
|
|
|
|
mlink = mlink->next;
|
|
|
|
if (NULL != mlink)
|
|
|
|
fputs(", ", stdout);
|
|
|
|
}
|
|
|
|
printf(" - %s\n", mpage->desc);
|
2012-10-18 09:55:16 +00:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
if (debug)
|
|
|
|
say(mlink->file, "Adding to database");
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-12-25 21:33:25 +00:00
|
|
|
cp = mpage->desc;
|
|
|
|
i = strlen(cp);
|
|
|
|
mustfree = render_string(&cp, &i);
|
2017-01-21 13:01:39 +00:00
|
|
|
mpage->dba = dba_page_new(dba->pages,
|
|
|
|
*mpage->arch == '\0' ? mlink->arch : mpage->arch,
|
|
|
|
cp, mlink->file, mpage->form);
|
2014-12-25 21:33:25 +00:00
|
|
|
if (mustfree)
|
|
|
|
free(cp);
|
2017-01-21 13:01:39 +00:00
|
|
|
dba_page_add(mpage->dba, DBP_SECT, mpage->sec);
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
while (mlink != NULL) {
|
2014-11-22 18:08:25 +00:00
|
|
|
dbadd_mlink(mlink);
|
|
|
|
mlink = mlink->next;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
|
|
|
|
for (key = ohash_first(&names, &slot); NULL != key;
|
|
|
|
key = ohash_next(&names, &slot)) {
|
|
|
|
assert(key->mpage == mpage);
|
2017-01-21 13:01:39 +00:00
|
|
|
dba_page_alias(mpage->dba, key->key, key->mask);
|
2014-11-22 18:08:25 +00:00
|
|
|
free(key);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
for (key = ohash_first(&strings, &slot); NULL != key;
|
|
|
|
key = ohash_next(&strings, &slot)) {
|
|
|
|
assert(key->mpage == mpage);
|
2017-01-21 13:01:39 +00:00
|
|
|
i = 0;
|
|
|
|
for (mask = TYPE_Xr; mask <= TYPE_Lb; mask *= 2) {
|
|
|
|
if (key->mask & mask)
|
|
|
|
dba_macro_add(dba->macros, i,
|
|
|
|
key->key, mpage->dba);
|
|
|
|
i++;
|
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
free(key);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static void
|
2017-01-21 13:01:39 +00:00
|
|
|
dbprune(struct dba *dba)
|
2014-11-22 18:08:25 +00:00
|
|
|
{
|
2017-01-21 13:01:39 +00:00
|
|
|
struct dba_array *page, *files;
|
|
|
|
char *file;
|
|
|
|
|
|
|
|
dba_array_FOREACH(dba->pages, page) {
|
|
|
|
files = dba_array_get(page, DBP_FILE);
|
|
|
|
dba_array_FOREACH(files, file) {
|
|
|
|
if (*file < ' ')
|
|
|
|
file++;
|
|
|
|
if (ohash_find(&mlinks, ohash_qlookup(&mlinks,
|
|
|
|
file)) != NULL) {
|
|
|
|
if (debug)
|
|
|
|
say(file, "Deleting from database");
|
|
|
|
dba_array_del(dba->pages);
|
|
|
|
break;
|
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
2017-01-21 13:01:39 +00:00
|
|
|
* Write the database from memory to disk.
|
2014-11-22 18:08:25 +00:00
|
|
|
*/
|
2012-10-18 09:55:16 +00:00
|
|
|
static void
|
2017-01-21 13:01:39 +00:00
|
|
|
dbwrite(struct dba *dba)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2017-01-21 13:01:39 +00:00
|
|
|
char tfn[32];
|
2014-11-22 18:08:25 +00:00
|
|
|
int status;
|
|
|
|
pid_t child;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
if (dba_write(MANDOC_DB "~", dba) != -1) {
|
|
|
|
if (rename(MANDOC_DB "~", MANDOC_DB) == -1) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say(MANDOC_DB, "&rename");
|
|
|
|
unlink(MANDOC_DB "~");
|
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
return;
|
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
(void)strlcpy(tfn, "/tmp/mandocdb.XXXXXXXX", sizeof(tfn));
|
|
|
|
if (mkdtemp(tfn) == NULL) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "&%s", tfn);
|
2014-11-22 18:08:25 +00:00
|
|
|
return;
|
2017-01-21 13:01:39 +00:00
|
|
|
}
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
(void)strlcat(tfn, "/" MANDOC_DB, sizeof(tfn));
|
|
|
|
if (dba_write(tfn, dba) == -1) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say(tfn, "&dba_write");
|
|
|
|
goto out;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
switch (child = fork()) {
|
|
|
|
case -1:
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "&fork cmp");
|
|
|
|
return;
|
|
|
|
case 0:
|
2017-01-21 13:01:39 +00:00
|
|
|
execlp("cmp", "cmp", "-s", tfn, MANDOC_DB, (char *)NULL);
|
2014-11-22 18:08:25 +00:00
|
|
|
say("", "&exec cmp");
|
|
|
|
exit(0);
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
2017-01-21 13:01:39 +00:00
|
|
|
if (waitpid(child, &status, 0) == -1) {
|
2014-11-22 18:08:25 +00:00
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "&wait cmp");
|
|
|
|
} else if (WIFSIGNALED(status)) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "cmp died from signal %d", WTERMSIG(status));
|
|
|
|
} else if (WEXITSTATUS(status)) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say(MANDOC_DB,
|
|
|
|
"Data changed, but cannot replace database");
|
|
|
|
}
|
|
|
|
|
2017-01-21 13:01:39 +00:00
|
|
|
out:
|
|
|
|
*strrchr(tfn, '/') = '\0';
|
2014-11-22 18:08:25 +00:00
|
|
|
switch (child = fork()) {
|
|
|
|
case -1:
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "&fork rm");
|
|
|
|
return;
|
|
|
|
case 0:
|
2017-01-21 13:01:39 +00:00
|
|
|
execlp("rm", "rm", "-rf", tfn, (char *)NULL);
|
2014-11-22 18:08:25 +00:00
|
|
|
say("", "&exec rm");
|
|
|
|
exit((int)MANDOCLEVEL_SYSERR);
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
2017-01-21 13:01:39 +00:00
|
|
|
if (waitpid(child, &status, 0) == -1) {
|
2014-11-22 18:08:25 +00:00
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "&wait rm");
|
|
|
|
} else if (WIFSIGNALED(status) || WEXITSTATUS(status)) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
2017-01-21 13:01:39 +00:00
|
|
|
say("", "%s: Cannot remove temporary directory", tfn);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
static int
|
2014-12-02 07:34:06 +00:00
|
|
|
set_basedir(const char *targetdir, int report_baddir)
|
2014-11-22 18:08:25 +00:00
|
|
|
{
|
|
|
|
static char startdir[PATH_MAX];
|
|
|
|
static int getcwd_status; /* 1 = ok, 2 = failure */
|
|
|
|
static int chdir_status; /* 1 = changed directory */
|
|
|
|
char *cp;
|
2012-10-18 09:55:16 +00:00
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Remember the original working directory, if possible.
|
|
|
|
* This will be needed if the second or a later directory
|
|
|
|
* on the command line is given as a relative path.
|
|
|
|
* Do not error out if the current directory is not
|
|
|
|
* searchable: Maybe it won't be needed after all.
|
|
|
|
*/
|
|
|
|
if (0 == getcwd_status) {
|
|
|
|
if (NULL == getcwd(startdir, sizeof(startdir))) {
|
|
|
|
getcwd_status = 2;
|
|
|
|
(void)strlcpy(startdir, strerror(errno),
|
|
|
|
sizeof(startdir));
|
|
|
|
} else
|
|
|
|
getcwd_status = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We are leaving the old base directory.
|
|
|
|
* Do not use it any longer, not even for messages.
|
|
|
|
*/
|
|
|
|
*basedir = '\0';
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If and only if the directory was changed earlier and
|
|
|
|
* the next directory to process is given as a relative path,
|
|
|
|
* first go back, or bail out if that is impossible.
|
|
|
|
*/
|
|
|
|
if (chdir_status && '/' != *targetdir) {
|
|
|
|
if (2 == getcwd_status) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "getcwd: %s", startdir);
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
|
|
|
if (-1 == chdir(startdir)) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "&chdir %s", startdir);
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2014-11-22 18:08:25 +00:00
|
|
|
/*
|
|
|
|
* Always resolve basedir to the canonicalized absolute
|
|
|
|
* pathname and append a trailing slash, such that
|
|
|
|
* we can reliably check whether files are inside.
|
|
|
|
*/
|
|
|
|
if (NULL == realpath(targetdir, basedir)) {
|
2014-12-02 07:34:06 +00:00
|
|
|
if (report_baddir || errno != ENOENT) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_BADARG;
|
|
|
|
say("", "&%s: realpath", targetdir);
|
|
|
|
}
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
} else if (-1 == chdir(basedir)) {
|
2014-12-02 07:34:06 +00:00
|
|
|
if (report_baddir || errno != ENOENT) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_BADARG;
|
|
|
|
say("", "&chdir");
|
|
|
|
}
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
|
|
|
chdir_status = 1;
|
|
|
|
cp = strchr(basedir, '\0');
|
|
|
|
if ('/' != cp[-1]) {
|
|
|
|
if (cp - basedir >= PATH_MAX - 1) {
|
|
|
|
exitcode = (int)MANDOCLEVEL_SYSERR;
|
|
|
|
say("", "Filename too long");
|
2016-01-15 23:08:59 +00:00
|
|
|
return 0;
|
2014-11-22 18:08:25 +00:00
|
|
|
}
|
|
|
|
*cp++ = '/';
|
|
|
|
*cp = '\0';
|
|
|
|
}
|
2016-01-15 23:08:59 +00:00
|
|
|
return 1;
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2014-11-22 18:08:25 +00:00
|
|
|
say(const char *file, const char *format, ...)
|
2012-10-18 09:55:16 +00:00
|
|
|
{
|
2014-11-22 18:08:25 +00:00
|
|
|
va_list ap;
|
|
|
|
int use_errno;
|
|
|
|
|
|
|
|
if ('\0' != *basedir)
|
|
|
|
fprintf(stderr, "%s", basedir);
|
|
|
|
if ('\0' != *basedir && '\0' != *file)
|
|
|
|
fputc('/', stderr);
|
|
|
|
if ('\0' != *file)
|
|
|
|
fprintf(stderr, "%s", file);
|
|
|
|
|
|
|
|
use_errno = 1;
|
|
|
|
if (NULL != format) {
|
|
|
|
switch (*format) {
|
|
|
|
case '&':
|
|
|
|
format++;
|
|
|
|
break;
|
|
|
|
case '\0':
|
|
|
|
format = NULL;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
use_errno = 0;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if (NULL != format) {
|
|
|
|
if ('\0' != *basedir || '\0' != *file)
|
|
|
|
fputs(": ", stderr);
|
|
|
|
va_start(ap, format);
|
|
|
|
vfprintf(stderr, format, ap);
|
|
|
|
va_end(ap);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|
2014-11-22 18:08:25 +00:00
|
|
|
if (use_errno) {
|
|
|
|
if ('\0' != *basedir || '\0' != *file || NULL != format)
|
|
|
|
fputs(": ", stderr);
|
|
|
|
perror(NULL);
|
|
|
|
} else
|
|
|
|
fputc('\n', stderr);
|
2012-10-18 09:55:16 +00:00
|
|
|
}
|