minix/usr.bin/tic/tic.c

592 lines
13 KiB
C
Raw Normal View History

/* $NetBSD: tic.c,v 1.19 2012/06/01 12:08:40 joerg Exp $ */
/*
* Copyright (c) 2009, 2010 The NetBSD Foundation, Inc.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Roy Marples.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
#include <sys/cdefs.h>
__RCSID("$NetBSD: tic.c,v 1.19 2012/06/01 12:08:40 joerg Exp $");
#include <sys/types.h>
#include <sys/queue.h>
#if !HAVE_NBTOOL_CONFIG_H || HAVE_SYS_ENDIAN_H
#include <sys/endian.h>
#endif
#include <ctype.h>
#include <err.h>
#include <errno.h>
#include <getopt.h>
#include <limits.h>
#include <fcntl.h>
#include <ndbm.h>
#include <search.h>
#include <stdarg.h>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <term_private.h>
#include <term.h>
#include <util.h>
#define HASH_SIZE 16384 /* 2012-06-01: 3600 entries */
/* We store the full list of terminals we have instead of iterating
through the database as the sequential iterator doesn't work
the the data size stored changes N amount which ours will. */
typedef struct term {
SLIST_ENTRY(term) next;
char *name;
char type;
TIC *tic;
} TERM;
static SLIST_HEAD(, term) terms = SLIST_HEAD_INITIALIZER(terms);
static int error_exit;
static int Sflag;
static char *dbname;
static size_t nterm, nalias;
static void
do_unlink(void)
{
if (dbname != NULL)
unlink(dbname);
}
static void __printflike(1, 2)
dowarn(const char *fmt, ...)
{
va_list va;
error_exit = 1;
va_start(va, fmt);
vwarnx(fmt, va);
va_end(va);
}
static char *
grow_tbuf(TBUF *tbuf, size_t len)
{
char *buf;
buf = _ti_grow_tbuf(tbuf, len);
if (buf == NULL)
err(1, "_ti_grow_tbuf");
return buf;
}
static int
save_term(DBM *db, TERM *term)
{
uint8_t *buf;
ssize_t len;
datum key, value;
len = _ti_flatten(&buf, term->tic);
if (len == -1)
return -1;
key.dptr = term->name;
key.dsize = strlen(term->name);
value.dptr = buf;
value.dsize = len;
if (dbm_store(db, key, value, DBM_REPLACE) == -1)
err(1, "dbm_store");
free(buf);
return 0;
}
static TERM *
find_term(const char *name)
{
ENTRY elem, *elemp;
elem.key = __UNCONST(name);
elem.data = NULL;
elemp = hsearch(elem, FIND);
return elemp ? (TERM *)elemp->data : NULL;
}
static TERM *
store_term(const char *name, char type)
{
TERM *term;
ENTRY elem;
term = ecalloc(1, sizeof(*term));
term->name = estrdup(name);
term->type = type;
SLIST_INSERT_HEAD(&terms, term, next);
elem.key = estrdup(name);
elem.data = term;
hsearch(elem, ENTER);
if (type == 'a')
nalias++;
else
nterm++;
return term;
}
static int
process_entry(TBUF *buf, int flags)
{
char *p, *e, *alias;
TERM *term;
TIC *tic;
if (buf->bufpos == 0)
return 0;
/* Terminate the string */
buf->buf[buf->bufpos - 1] = '\0';
/* First rewind the buffer for new entries */
buf->bufpos = 0;
if (isspace((unsigned char)*buf->buf))
return 0;
tic = _ti_compile(buf->buf, flags);
if (tic == NULL)
return 0;
if (find_term(tic->name) != NULL) {
dowarn("%s: duplicate entry", tic->name);
_ti_freetic(tic);
return 0;
}
term = store_term(tic->name, 't');
term->tic = tic;
/* Create aliased terms */
if (tic->alias != NULL) {
alias = p = estrdup(tic->alias);
while (p != NULL && *p != '\0') {
e = strchr(p, '|');
if (e != NULL)
*e++ = '\0';
if (find_term(p) != NULL) {
dowarn("%s: has alias for already assigned"
" term %s", tic->name, p);
} else {
term = store_term(p, 'a');
term->tic = ecalloc(sizeof(*term->tic), 1);
term->tic->name = estrdup(tic->name);
}
p = e;
}
free(alias);
}
return 0;
}
static void
merge(TIC *rtic, TIC *utic, int flags)
{
char *cap, flag, *code, type, *str;
short ind, num;
size_t n;
cap = utic->flags.buf;
for (n = utic->flags.entries; n > 0; n--) {
ind = le16dec(cap);
cap += sizeof(uint16_t);
flag = *cap++;
if (VALID_BOOLEAN(flag) &&
_ti_find_cap(&rtic->flags, 'f', ind) == NULL)
{
_ti_grow_tbuf(&rtic->flags, sizeof(uint16_t) + 1);
le16enc(rtic->flags.buf + rtic->flags.bufpos, ind);
rtic->flags.bufpos += sizeof(uint16_t);
rtic->flags.buf[rtic->flags.bufpos++] = flag;
rtic->flags.entries++;
}
}
cap = utic->nums.buf;
for (n = utic->nums.entries; n > 0; n--) {
ind = le16dec(cap);
cap += sizeof(uint16_t);
num = le16dec(cap);
cap += sizeof(uint16_t);
if (VALID_NUMERIC(num) &&
_ti_find_cap(&rtic->nums, 'n', ind) == NULL)
{
grow_tbuf(&rtic->nums, sizeof(uint16_t) * 2);
le16enc(rtic->nums.buf + rtic->nums.bufpos, ind);
rtic->nums.bufpos += sizeof(uint16_t);
le16enc(rtic->nums.buf + rtic->nums.bufpos, num);
rtic->nums.bufpos += sizeof(uint16_t);
rtic->nums.entries++;
}
}
cap = utic->strs.buf;
for (n = utic->strs.entries; n > 0; n--) {
ind = le16dec(cap);
cap += sizeof(uint16_t);
num = le16dec(cap);
cap += sizeof(uint16_t);
if (num > 0 &&
_ti_find_cap(&rtic->strs, 's', ind) == NULL)
{
grow_tbuf(&rtic->strs, (sizeof(uint16_t) * 2) + num);
le16enc(rtic->strs.buf + rtic->strs.bufpos, ind);
rtic->strs.bufpos += sizeof(uint16_t);
le16enc(rtic->strs.buf + rtic->strs.bufpos, num);
rtic->strs.bufpos += sizeof(uint16_t);
memcpy(rtic->strs.buf + rtic->strs.bufpos,
cap, num);
rtic->strs.bufpos += num;
rtic->strs.entries++;
}
cap += num;
}
cap = utic->extras.buf;
for (n = utic->extras.entries; n > 0; n--) {
num = le16dec(cap);
cap += sizeof(uint16_t);
code = cap;
cap += num;
type = *cap++;
flag = 0;
str = NULL;
switch (type) {
case 'f':
flag = *cap++;
if (!VALID_BOOLEAN(flag))
continue;
break;
case 'n':
num = le16dec(cap);
cap += sizeof(uint16_t);
if (!VALID_NUMERIC(num))
continue;
break;
case 's':
num = le16dec(cap);
cap += sizeof(uint16_t);
str = cap;
cap += num;
if (num == 0)
continue;
break;
}
_ti_store_extra(rtic, 0, code, type, flag, num, str, num,
flags);
}
}
static size_t
merge_use(int flags)
{
size_t skipped, merged, memn;
char *cap, *scap;
uint16_t num;
TIC *rtic, *utic;
TERM *term, *uterm;;
skipped = merged = 0;
SLIST_FOREACH(term, &terms, next) {
if (term->type == 'a')
continue;
rtic = term->tic;
while ((cap = _ti_find_extra(&rtic->extras, "use")) != NULL) {
if (*cap++ != 's') {
dowarn("%s: use is not string", rtic->name);
break;
}
cap += sizeof(uint16_t);
if (strcmp(rtic->name, cap) == 0) {
dowarn("%s: uses itself", rtic->name);
goto remove;
}
uterm = find_term(cap);
if (uterm != NULL && uterm->type == 'a')
uterm = find_term(uterm->tic->name);
if (uterm == NULL) {
dowarn("%s: no use record for %s",
rtic->name, cap);
goto remove;
}
utic = uterm->tic;
if (strcmp(utic->name, rtic->name) == 0) {
dowarn("%s: uses itself", rtic->name);
goto remove;
}
if (_ti_find_extra(&utic->extras, "use") != NULL) {
skipped++;
break;
}
cap = _ti_find_extra(&rtic->extras, "use");
merge(rtic, utic, flags);
remove:
/* The pointers may have changed, find the use again */
cap = _ti_find_extra(&rtic->extras, "use");
if (cap == NULL)
dowarn("%s: use no longer exists - impossible",
rtic->name);
else {
scap = cap - (4 + sizeof(uint16_t));
cap++;
num = le16dec(cap);
cap += sizeof(uint16_t) + num;
memn = rtic->extras.bufpos -
(cap - rtic->extras.buf);
memmove(scap, cap, memn);
rtic->extras.bufpos -= cap - scap;
cap = scap;
rtic->extras.entries--;
merged++;
}
}
}
if (merged == 0 && skipped != 0)
dowarn("circular use detected");
return merged;
}
static int
print_dump(int argc, char **argv)
{
TERM *term;
uint8_t *buf;
int i, n;
size_t j, col;
ssize_t len;
printf("struct compiled_term {\n");
printf("\tconst char *name;\n");
printf("\tconst char *cap;\n");
printf("\tsize_t caplen;\n");
printf("};\n\n");
printf("const struct compiled_term compiled_terms[] = {\n");
n = 0;
for (i = 0; i < argc; i++) {
term = find_term(argv[i]);
if (term == NULL) {
warnx("%s: no description for terminal", argv[i]);
continue;
}
if (term->type == 'a') {
warnx("%s: cannot dump alias", argv[i]);
continue;
}
/* Don't compile the aliases in, save space */
free(term->tic->alias);
term->tic->alias = NULL;
len = _ti_flatten(&buf, term->tic);
if (len == 0 || len == -1)
continue;
printf("\t{\n");
printf("\t\t\"%s\",\n", argv[i]);
n++;
for (j = 0, col = 0; j < (size_t)len; j++) {
if (col == 0) {
printf("\t\t\"");
col = 16;
}
col += printf("\\%03o", (uint8_t)buf[j]);
if (col > 75) {
printf("\"%s\n",
j + 1 == (size_t)len ? "," : "");
col = 0;
}
}
if (col != 0)
printf("\",\n");
printf("\t\t%zu\n", len);
printf("\t}");
if (i + 1 < argc)
printf(",");
printf("\n");
free(buf);
}
printf("};\n");
return n;
}
int
main(int argc, char **argv)
{
int ch, cflag, sflag, flags;
char *source, *p, *buf, *ofile;
FILE *f;
DBM *db;
size_t buflen;
ssize_t len;
TBUF tbuf;
TERM *term;
cflag = sflag = 0;
ofile = NULL;
flags = TIC_ALIAS | TIC_DESCRIPTION | TIC_WARNING;
while ((ch = getopt(argc, argv, "Saco:sx")) != -1)
switch (ch) {
case 'S':
Sflag = 1;
/* We still compile aliases so that use= works.
* However, it's removed before we flatten to save space. */
flags &= ~TIC_DESCRIPTION;
break;
case 'a':
flags |= TIC_COMMENT;
break;
case 'c':
cflag = 1;
break;
case 'o':
ofile = optarg;
break;
case 's':
sflag = 1;
break;
case 'x':
flags |= TIC_EXTRA;
break;
case '?': /* FALLTHROUGH */
default:
fprintf(stderr, "usage: %s [-acSsx] [-o file] source\n",
getprogname());
return EXIT_FAILURE;
}
if (optind == argc)
errx(1, "No source file given");
source = argv[optind++];
f = fopen(source, "r");
if (f == NULL)
err(1, "fopen: %s", source);
if (!cflag && !Sflag) {
if (ofile == NULL)
ofile = source;
len = strlen(ofile) + 9;
dbname = emalloc(len + 4); /* For adding .db after open */
snprintf(dbname, len, "%s.tmp", ofile);
db = dbm_open(dbname, O_CREAT | O_RDWR | O_TRUNC, DEFFILEMODE);
if (db == NULL)
err(1, "dbopen: %s", source);
p = dbname + strlen(dbname);
*p++ = '.';
*p++ = 'd';
*p++ = 'b';
*p++ = '\0';
atexit(do_unlink);
} else
db = NULL; /* satisfy gcc warning */
hcreate(HASH_SIZE);
buf = tbuf.buf = NULL;
buflen = tbuf.buflen = tbuf.bufpos = 0;
while ((len = getline(&buf, &buflen, f)) != -1) {
/* Skip comments */
if (*buf == '#')
continue;
if (buf[len - 1] != '\n') {
process_entry(&tbuf, flags);
dowarn("last line is not a comment"
" and does not end with a newline");
continue;
}
/*
If the first char is space not a space then we have a
new entry, so process it.
*/
if (!isspace((unsigned char)*buf) && tbuf.bufpos != 0)
process_entry(&tbuf, flags);
/* Grow the buffer if needed */
grow_tbuf(&tbuf, len);
/* Append the string */
memcpy(tbuf.buf + tbuf.bufpos, buf, len);
tbuf.bufpos += len;
}
free(buf);
/* Process the last entry if not done already */
process_entry(&tbuf, flags);
free(tbuf.buf);
/* Merge use entries until we have merged all we can */
while (merge_use(flags) != 0)
;
if (Sflag) {
print_dump(argc - optind, argv + optind);
return error_exit;
}
if (cflag)
return error_exit;
/* Save the terms */
SLIST_FOREACH(term, &terms, next)
save_term(db, term);
/* done! */
dbm_close(db);
/* Rename the tmp db to the real one now */
easprintf(&p, "%s.db", ofile);
if (rename(dbname, p) == -1)
err(1, "rename");
free(dbname);
dbname = NULL;
if (sflag != 0)
fprintf(stderr, "%zu entries and %zu aliases written to %s\n",
nterm, nalias, p);
#ifdef __VALGRIND__
free(p);
while ((term = SLIST_FIRST(&terms)) != NULL) {
SLIST_REMOVE_HEAD(&terms, next);
_ti_freetic(term->tic);
free(term->name);
free(term);
}
hdestroy();
#endif
return EXIT_SUCCESS;
}