version 1.25, 2009/03/02 17:14:46 |
version 1.49, 2009/03/19 11:49:00 |
|
|
/* $Id$ */ |
/* $Id$ */ |
/* |
/* |
* Copyright (c) 2008 Kristaps Dzonsons <kristaps@kth.se> |
* Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@openbsd.org> |
* |
* |
* Permission to use, copy, modify, and distribute this software for any |
* Permission to use, copy, modify, and distribute this software for any |
* purpose with or without fee is hereby granted, provided that the |
* purpose with or without fee is hereby granted, provided that the |
|
|
* TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR |
* TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR |
* PERFORMANCE OF THIS SOFTWARE. |
* PERFORMANCE OF THIS SOFTWARE. |
*/ |
*/ |
#include <sys/utsname.h> |
#include <sys/types.h> |
|
#include <sys/stat.h> |
|
|
#include <assert.h> |
#include <assert.h> |
#include <ctype.h> |
|
#include <err.h> |
#include <err.h> |
|
#include <fcntl.h> |
#include <getopt.h> |
#include <getopt.h> |
#include <stdio.h> |
#include <stdio.h> |
#include <stdlib.h> |
#include <stdlib.h> |
#include <string.h> |
#include <string.h> |
#ifndef __OpenBSD__ |
#include <unistd.h> |
#include <time.h> |
|
#endif |
|
|
|
#include "mmain.h" |
|
#include "term.h" |
#include "term.h" |
|
|
static void body(struct termp *, |
#define WARN_WALL 0x03 /* All-warnings mask. */ |
struct termpair *, |
#define WARN_WCOMPAT (1 << 0) /* Compatibility warnings. */ |
|
#define WARN_WSYNTAX (1 << 1) /* Syntax warnings. */ |
|
#define WARN_WERR (1 << 2) /* Warnings->errors. */ |
|
|
|
enum termt { |
|
TERMT_ASCII, |
|
TERMT_LINT, |
|
TERMT_TREE |
|
}; |
|
|
|
extern char *__progname; |
|
|
|
__dead static void version(void); |
|
__dead static void usage(void); |
|
#if 0 |
|
__dead static void punt(struct cmdargs *, char **); |
|
#endif |
|
static int foptions(int *, char *); |
|
static int toptions(enum termt *, char *); |
|
static int woptions(int *, char *); |
|
static int merr(void *, int, int, const char *); |
|
static int mwarn(void *, int, int, |
|
enum mdoc_warn, const char *); |
|
static void body(struct termp *, struct termpair *, |
const struct mdoc_meta *, |
const struct mdoc_meta *, |
const struct mdoc_node *); |
const struct mdoc_node *); |
static void header(struct termp *, |
static void header(struct termp *, |
const struct mdoc_meta *); |
const struct mdoc_meta *); |
static void footer(struct termp *, |
static void footer(struct termp *, |
const struct mdoc_meta *); |
const struct mdoc_meta *); |
|
static int file(char **, size_t *, char **, size_t *, |
static void pword(struct termp *, const char *, size_t); |
const char *, struct mdoc *); |
static void pescape(struct termp *, const char *, |
static int fdesc(char **, size_t *, char **, size_t *, |
size_t *, size_t); |
const char *, int, struct mdoc *); |
static void style(struct termp *, enum tstyle); |
static void pword(struct termp *, const char *, int); |
|
static void pescape(struct termp *, |
|
const char *, int *, int); |
static void nescape(struct termp *, |
static void nescape(struct termp *, |
const char *, size_t); |
const char *, size_t); |
static void chara(struct termp *, char); |
static void chara(struct termp *, char); |
static void stringa(struct termp *, |
static void stringa(struct termp *, |
const char *, size_t); |
const char *, size_t); |
static void symbola(struct termp *, enum tsym); |
static void sanity(const struct mdoc_node *); |
|
|
#ifdef __linux__ |
|
extern size_t strlcat(char *, const char *, size_t); |
|
extern size_t strlcpy(char *, const char *, size_t); |
|
#endif |
|
|
|
static struct termsym termsym_ansi[] = { |
|
{ "]", 1 }, /* TERMSYM_RBRACK */ |
|
{ "[", 1 }, /* TERMSYM_LBRACK */ |
|
{ "<-", 2 }, /* TERMSYM_LARROW */ |
|
{ "->", 2 }, /* TERMSYM_RARROW */ |
|
{ "^", 1 }, /* TERMSYM_UARROW */ |
|
{ "v", 1 }, /* TERMSYM_DARROW */ |
|
{ "`", 1 }, /* TERMSYM_LSQUOTE */ |
|
{ "\'", 1 }, /* TERMSYM_RSQUOTE */ |
|
{ "\'", 1 }, /* TERMSYM_SQUOTE */ |
|
{ "``", 2 }, /* TERMSYM_LDQUOTE */ |
|
{ "\'\'", 2 }, /* TERMSYM_RDQUOTE */ |
|
{ "\"", 1 }, /* TERMSYM_DQUOTE */ |
|
{ "<", 1 }, /* TERMSYM_LT */ |
|
{ ">", 1 }, /* TERMSYM_GT */ |
|
{ "<=", 2 }, /* TERMSYM_LE */ |
|
{ ">=", 2 }, /* TERMSYM_GE */ |
|
{ "==", 2 }, /* TERMSYM_EQ */ |
|
{ "!=", 2 }, /* TERMSYM_NEQ */ |
|
{ "\'", 1 }, /* TERMSYM_ACUTE */ |
|
{ "`", 1 }, /* TERMSYM_GRAVE */ |
|
{ "pi", 2 }, /* TERMSYM_PI */ |
|
{ "+=", 2 }, /* TERMSYM_PLUSMINUS */ |
|
{ "oo", 2 }, /* TERMSYM_INF */ |
|
{ "infinity", 8 }, /* TERMSYM_INF2 */ |
|
{ "NaN", 3 }, /* TERMSYM_NAN */ |
|
{ "|", 1 }, /* TERMSYM_BAR */ |
|
{ "o", 1 }, /* TERMSYM_BULLET */ |
|
{ "&", 1 }, /* TERMSYM_AND */ |
|
{ "|", 1 }, /* TERMSYM_OR */ |
|
}; |
|
|
|
static const char ansi_clear[] = { 27, '[', '0', 'm' }; |
|
static const char ansi_bold[] = { 27, '[', '1', 'm' }; |
|
static const char ansi_under[] = { 27, '[', '4', 'm' }; |
|
|
|
static struct termsym termstyle_ansi[] = { |
|
{ ansi_clear, 4 }, |
|
{ ansi_bold, 4 }, |
|
{ ansi_under, 4 } |
|
}; |
|
|
|
|
|
int |
int |
main(int argc, char *argv[]) |
main(int argc, char *argv[]) |
{ |
{ |
struct mmain *p; |
|
const struct mdoc *mdoc; |
|
struct termp termp; |
struct termp termp; |
|
int c, fflags, wflags; |
|
struct mdoc_cb cb; |
|
struct mdoc *mdoc; |
|
char *buf, *line; |
|
size_t bufsz, linesz; |
|
enum termt termt; |
|
|
p = mmain_alloc(); |
bzero(&termp, sizeof(struct termp)); |
|
bzero(&cb, sizeof(struct mdoc_cb)); |
|
|
if ( ! mmain_getopt(p, argc, argv, NULL, NULL, NULL, NULL)) |
termt = TERMT_ASCII; |
mmain_exit(p, 1); |
fflags = wflags = 0; |
|
|
if (NULL == (mdoc = mmain_mdoc(p))) |
/* LINTED */ |
mmain_exit(p, 1); |
while (-1 != (c = getopt(argc, argv, "f:VW:T:"))) |
|
switch (c) { |
|
case ('f'): |
|
if ( ! foptions(&fflags, optarg)) |
|
return(0); |
|
break; |
|
case ('T'): |
|
if ( ! toptions(&termt, optarg)) |
|
return(0); |
|
break; |
|
case ('W'): |
|
if ( ! woptions(&wflags, optarg)) |
|
return(0); |
|
break; |
|
case ('V'): |
|
version(); |
|
/* NOTREACHED */ |
|
default: |
|
usage(); |
|
/* NOTREACHED */ |
|
} |
|
|
termp.maxrmargin = 78; /* XXX */ |
argc -= optind; |
termp.rmargin = termp.maxrmargin; |
argv += optind; |
termp.maxcols = 1024; |
|
termp.offset = termp.col = 0; |
|
termp.flags = TERMP_NOSPACE; |
|
termp.symtab = termsym_ansi; |
|
termp.styletab = termstyle_ansi; |
|
|
|
if (NULL == (termp.buf = malloc(termp.maxcols))) |
termp.maxrmargin = 78; /* FIXME */ |
err(1, "malloc"); |
|
|
|
header(&termp, mdoc_meta(mdoc)); |
cb.mdoc_err = merr; |
body(&termp, NULL, mdoc_meta(mdoc), mdoc_node(mdoc)); |
cb.mdoc_warn = mwarn; |
footer(&termp, mdoc_meta(mdoc)); |
|
|
|
free(termp.buf); |
/* Line and block buffers persist between parses. */ |
|
|
mmain_exit(p, 0); |
buf = line = NULL; |
|
bufsz = linesz = 0; |
|
|
|
/* Overall mdoc persists between parses. */ |
|
|
|
mdoc = mdoc_alloc(&wflags, fflags, &cb); |
|
|
|
while (*argv) { |
|
if ( ! file(&line, &linesz, &buf, &bufsz, *argv, mdoc)) |
|
break; |
|
|
|
switch (termt) { |
|
case (TERMT_ASCII): |
|
if (NULL == termp.symtab) |
|
termp.symtab = ascii2htab(); |
|
header(&termp, mdoc_meta(mdoc)); |
|
body(&termp, NULL, mdoc_meta(mdoc), |
|
mdoc_node(mdoc)); |
|
footer(&termp, mdoc_meta(mdoc)); |
|
break; |
|
default: |
|
break; |
|
} |
|
|
|
mdoc_reset(mdoc); |
|
argv++; |
|
} |
|
|
|
if (buf) |
|
free(buf); |
|
if (line) |
|
free(line); |
|
if (termp.buf) |
|
free(termp.buf); |
|
if (termp.symtab) |
|
asciifree(termp.symtab); |
|
|
|
mdoc_free(mdoc); |
|
|
|
return(0); |
|
} |
|
|
|
|
|
__dead static void |
|
version(void) |
|
{ |
|
|
|
(void)printf("%s %s\n", __progname, VERSION); |
|
exit(0); |
/* NOTREACHED */ |
/* NOTREACHED */ |
} |
} |
|
|
|
|
|
__dead static void |
|
usage(void) |
|
{ |
|
|
|
(void)fprintf(stderr, "usage: %s\n", __progname); |
|
exit(1); |
|
/* NOTREACHED */ |
|
} |
|
|
|
|
|
static int |
|
file(char **ln, size_t *lnsz, char **buf, size_t *bufsz, |
|
const char *file, struct mdoc *mdoc) |
|
{ |
|
int fd, c; |
|
|
|
if (-1 == (fd = open(file, O_RDONLY, 0))) { |
|
warn("%s", file); |
|
return(0); |
|
} |
|
|
|
c = fdesc(ln, lnsz, buf, bufsz, file, fd, mdoc); |
|
|
|
if (-1 == close(fd)) |
|
warn("%s", file); |
|
|
|
return(c); |
|
} |
|
|
|
|
|
static int |
|
fdesc(char **lnp, size_t *lnsz, char **bufp, size_t *bufsz, |
|
const char *f, int fd, struct mdoc *mdoc) |
|
{ |
|
size_t sz; |
|
ssize_t ssz; |
|
struct stat st; |
|
int j, i, pos, lnn; |
|
char *ln, *buf; |
|
|
|
buf = *bufp; |
|
ln = *lnp; |
|
|
|
/* |
|
* Two buffers: ln and buf. buf is the input buffer, optimised |
|
* for each file's block size. ln is a line buffer. Both |
|
* growable, hence passed in by ptr-ptr. |
|
*/ |
|
|
|
if (-1 == fstat(fd, &st)) { |
|
warnx("%s", f); |
|
sz = BUFSIZ; |
|
} else |
|
sz = (unsigned)BUFSIZ > st.st_blksize ? |
|
(size_t)BUFSIZ : st.st_blksize; |
|
|
|
if (sz > *bufsz) { |
|
if (NULL == (buf = realloc(buf, sz))) |
|
err(1, "realloc"); |
|
*bufp = buf; |
|
*bufsz = sz; |
|
} |
|
|
|
/* |
|
* Fill buf with file blocksize and parse newlines into ln. |
|
*/ |
|
|
|
for (lnn = 1, pos = 0; ; ) { |
|
if (-1 == (ssz = read(fd, buf, sz))) { |
|
warn("%s", f); |
|
return(0); |
|
} else if (0 == ssz) |
|
break; |
|
|
|
for (i = 0; i < (int)ssz; i++) { |
|
if (pos >= (int)*lnsz) { |
|
*lnsz += 256; /* Step-size. */ |
|
ln = realloc(ln, *lnsz); |
|
if (NULL == ln) |
|
err(1, "realloc"); |
|
*lnp = ln; |
|
} |
|
|
|
if ('\n' != buf[i]) { |
|
ln[pos++] = buf[i]; |
|
continue; |
|
} |
|
|
|
/* Check for CPP-escaped newline. */ |
|
|
|
if (pos > 0 && '\\' == ln[pos - 1]) { |
|
for (j = pos - 1; j >= 0; j--) |
|
if ('\\' != ln[j]) |
|
break; |
|
|
|
if ( ! ((pos - j) % 2)) { |
|
pos--; |
|
lnn++; |
|
continue; |
|
} |
|
} |
|
|
|
ln[pos] = 0; |
|
if ( ! mdoc_parseln(mdoc, lnn, ln)) |
|
return(0); |
|
lnn++; |
|
pos = 0; |
|
} |
|
} |
|
|
|
return(mdoc_endparse(mdoc)); |
|
} |
|
|
|
|
|
static int |
|
toptions(enum termt *tflags, char *arg) |
|
{ |
|
|
|
if (0 == strcmp(arg, "ascii")) |
|
*tflags = TERMT_ASCII; |
|
else if (0 == strcmp(arg, "lint")) |
|
*tflags = TERMT_LINT; |
|
else if (0 == strcmp(arg, "tree")) |
|
*tflags = TERMT_TREE; |
|
else { |
|
warnx("bad argument: -T%s", arg); |
|
return(0); |
|
} |
|
|
|
return(1); |
|
} |
|
|
|
|
/* |
/* |
|
* Parse out the options for [-fopt...] setting compiler options. These |
|
* can be comma-delimited or called again. |
|
*/ |
|
static int |
|
foptions(int *fflags, char *arg) |
|
{ |
|
char *v; |
|
char *toks[] = { "ign-scope", "ign-escape", |
|
"ign-macro", NULL }; |
|
|
|
while (*arg) |
|
switch (getsubopt(&arg, toks, &v)) { |
|
case (0): |
|
*fflags |= MDOC_IGN_SCOPE; |
|
break; |
|
case (1): |
|
*fflags |= MDOC_IGN_ESCAPE; |
|
break; |
|
case (2): |
|
*fflags |= MDOC_IGN_MACRO; |
|
break; |
|
default: |
|
warnx("bad argument: -f%s", arg); |
|
return(0); |
|
} |
|
|
|
return(1); |
|
} |
|
|
|
|
|
/* |
|
* Parse out the options for [-Werr...], which sets warning modes. |
|
* These can be comma-delimited or called again. XXX - should this be |
|
* using -w like troff? |
|
*/ |
|
static int |
|
woptions(int *wflags, char *arg) |
|
{ |
|
char *v; |
|
char *toks[] = { "all", "compat", |
|
"syntax", "error", NULL }; |
|
|
|
while (*arg) |
|
switch (getsubopt(&arg, toks, &v)) { |
|
case (0): |
|
*wflags |= WARN_WALL; |
|
break; |
|
case (1): |
|
*wflags |= WARN_WCOMPAT; |
|
break; |
|
case (2): |
|
*wflags |= WARN_WSYNTAX; |
|
break; |
|
case (3): |
|
*wflags |= WARN_WERR; |
|
break; |
|
default: |
|
warnx("bad argument: -W%s", arg); |
|
return(0); |
|
} |
|
|
|
return(1); |
|
} |
|
|
|
|
|
/* |
* Flush a line of text. A "line" is loosely defined as being something |
* Flush a line of text. A "line" is loosely defined as being something |
* that should be followed by a newline, regardless of whether it's |
* that should be followed by a newline, regardless of whether it's |
* broken apart by newlines getting there. A line can also be a |
* broken apart by newlines getting there. A line can also be a |
Line 153 main(int argc, char *argv[]) |
|
Line 394 main(int argc, char *argv[]) |
|
* offset value. This is useful when doing columnar lists where the |
* offset value. This is useful when doing columnar lists where the |
* prior column has right-padded. |
* prior column has right-padded. |
* |
* |
* - TERMP_LITERAL: don't break apart words. Note that a long literal |
|
* word will violate the right margin. |
|
* |
|
* - TERMP_NOBREAK: this is the most important and is used when making |
* - TERMP_NOBREAK: this is the most important and is used when making |
* columns. In short: don't print a newline and instead pad to the |
* columns. In short: don't print a newline and instead pad to the |
* right margin. Used in conjunction with TERMP_NOLPAD. |
* right margin. Used in conjunction with TERMP_NOLPAD. |
* |
* |
|
* - TERMP_NONOBREAK: don't newline when TERMP_NOBREAK is specified. |
|
* |
* In-line line breaking: |
* In-line line breaking: |
* |
* |
* If TERMP_NOBREAK is specified and the line overruns the right |
* If TERMP_NOBREAK is specified and the line overruns the right |
Line 175 main(int argc, char *argv[]) |
|
Line 415 main(int argc, char *argv[]) |
|
void |
void |
flushln(struct termp *p) |
flushln(struct termp *p) |
{ |
{ |
size_t i, j, vsz, vis, maxvis, mmax, bp; |
int i, j; |
|
size_t vsz, vis, maxvis, mmax, bp; |
|
|
/* |
/* |
* First, establish the maximum columns of "visible" content. |
* First, establish the maximum columns of "visible" content. |
Line 198 flushln(struct termp *p) |
|
Line 439 flushln(struct termp *p) |
|
|
|
if ( ! (p->flags & TERMP_NOLPAD)) |
if ( ! (p->flags & TERMP_NOLPAD)) |
/* LINTED */ |
/* LINTED */ |
for (j = 0; j < p->offset; j++) |
for (j = 0; j < (int)p->offset; j++) |
putchar(' '); |
putchar(' '); |
|
|
for (i = 0; i < p->col; i++) { |
for (i = 0; i < (int)p->col; i++) { |
/* |
/* |
* Count up visible word characters. Control sequences |
* Count up visible word characters. Control sequences |
* (starting with the CSI) aren't counted. A space |
* (starting with the CSI) aren't counted. A space |
Line 209 flushln(struct termp *p) |
|
Line 450 flushln(struct termp *p) |
|
* space is printed according to regular spacing rules). |
* space is printed according to regular spacing rules). |
*/ |
*/ |
|
|
/* FIXME: make non-ANSI friendly. */ |
|
|
|
/* LINTED */ |
/* LINTED */ |
for (j = i, vsz = 0; j < p->col; j++) { |
for (j = i, vsz = 0; j < (int)p->col; j++) { |
if (isspace((int)p->buf[j])) |
if (' ' == p->buf[j]) |
break; |
break; |
else if (27 == p->buf[j]) { |
else if (8 == p->buf[j]) |
assert(j + 4 <= p->col); |
j += 1; |
j += 3; |
else |
} else |
|
vsz++; |
vsz++; |
} |
} |
|
|
Line 233 flushln(struct termp *p) |
|
Line 471 flushln(struct termp *p) |
|
if ( ! (TERMP_NOBREAK & p->flags)) { |
if ( ! (TERMP_NOBREAK & p->flags)) { |
if (vis && vis + vsz > bp) { |
if (vis && vis + vsz > bp) { |
putchar('\n'); |
putchar('\n'); |
for (j = 0; j < p->offset; j++) |
for (j = 0; j < (int)p->offset; j++) |
putchar(' '); |
putchar(' '); |
vis = 0; |
vis = 0; |
} else if (vis + vsz > bp) |
} else if (vis + vsz > bp) |
Line 244 flushln(struct termp *p) |
|
Line 482 flushln(struct termp *p) |
|
} else { |
} else { |
if (vis && vis + vsz > bp) { |
if (vis && vis + vsz > bp) { |
putchar('\n'); |
putchar('\n'); |
for (j = 0; j < p->rmargin; j++) |
for (j = 0; j < (int)p->rmargin; j++) |
putchar(' '); |
putchar(' '); |
vis = p->rmargin; |
vis = p->rmargin - p->offset; |
} else if (vis + vsz > bp) |
} else if (vis + vsz > bp) |
warnx("word breaks right margin"); |
warnx("word breaks right margin"); |
|
|
Line 259 flushln(struct termp *p) |
|
Line 497 flushln(struct termp *p) |
|
* our breakpoint. |
* our breakpoint. |
*/ |
*/ |
|
|
for ( ; i < p->col; i++) { |
for ( ; i < (int)p->col; i++) { |
if (isspace((int)p->buf[i])) |
if (' ' == p->buf[i]) |
break; |
break; |
putchar(p->buf[i]); |
putchar(p->buf[i]); |
} |
} |
vis += vsz; |
vis += vsz; |
if (i < p->col && vis <= bp) { |
if (i < (int)p->col && vis <= bp) { |
putchar(' '); |
putchar(' '); |
vis++; |
vis++; |
} |
} |
Line 277 flushln(struct termp *p) |
|
Line 515 flushln(struct termp *p) |
|
*/ |
*/ |
|
|
if ((TERMP_NOBREAK & p->flags) && vis >= maxvis) { |
if ((TERMP_NOBREAK & p->flags) && vis >= maxvis) { |
putchar('\n'); |
if ( ! (TERMP_NONOBREAK & p->flags)) { |
for (i = 0; i < p->rmargin; i++) |
putchar('\n'); |
putchar(' '); |
for (i = 0; i < (int)p->rmargin; i++) |
|
putchar(' '); |
|
} |
p->col = 0; |
p->col = 0; |
return; |
return; |
} |
} |
Line 290 flushln(struct termp *p) |
|
Line 530 flushln(struct termp *p) |
|
*/ |
*/ |
|
|
if (p->flags & TERMP_NOBREAK) { |
if (p->flags & TERMP_NOBREAK) { |
for ( ; vis < maxvis; vis++) |
if ( ! (TERMP_NONOBREAK & p->flags)) |
putchar(' '); |
for ( ; vis < maxvis; vis++) |
|
putchar(' '); |
} else |
} else |
putchar('\n'); |
putchar('\n'); |
|
|
Line 342 vspace(struct termp *p) |
|
Line 583 vspace(struct termp *p) |
|
void |
void |
word(struct termp *p, const char *word) |
word(struct termp *p, const char *word) |
{ |
{ |
size_t i, j, len; |
int i, j, len; |
|
|
if (p->flags & TERMP_LITERAL) { |
if (p->flags & TERMP_LITERAL) { |
pword(p, word, strlen(word)); |
pword(p, word, (int)strlen(word)); |
return; |
return; |
} |
} |
|
|
len = strlen(word); |
if (0 == (len = (int)strlen(word))) |
assert(len > 0); |
errx(1, "blank line not in literal context"); |
|
|
if (mdoc_isdelim(word)) { |
if (mdoc_isdelim(word)) { |
if ( ! (p->flags & TERMP_IGNDELIM)) |
if ( ! (p->flags & TERMP_IGNDELIM)) |
Line 360 word(struct termp *p, const char *word) |
|
Line 601 word(struct termp *p, const char *word) |
|
|
|
/* LINTED */ |
/* LINTED */ |
for (j = i = 0; i < len; i++) { |
for (j = i = 0; i < len; i++) { |
if ( ! isspace((int)word[i])) { |
if (' ' != word[i]) { |
j++; |
j++; |
continue; |
continue; |
} |
} |
|
|
/* Escaped spaces don't delimit... */ |
/* Escaped spaces don't delimit... */ |
if (i > 0 && isspace((int)word[i]) && |
if (i && ' ' == word[i] && '\\' == word[i - 1]) { |
'\\' == word[i - 1]) { |
|
j++; |
j++; |
continue; |
continue; |
} |
} |
Line 399 body(struct termp *p, struct termpair *ppair, |
|
Line 639 body(struct termp *p, struct termpair *ppair, |
|
int dochild; |
int dochild; |
struct termpair pair; |
struct termpair pair; |
|
|
|
/* Some quick sanity-checking. */ |
|
|
|
sanity(node); |
|
|
/* Pre-processing. */ |
/* Pre-processing. */ |
|
|
dochild = 1; |
dochild = 1; |
Line 413 body(struct termp *p, struct termpair *ppair, |
|
Line 657 body(struct termp *p, struct termpair *ppair, |
|
if ( ! (*termacts[node->tok].pre)(p, &pair, meta, node)) |
if ( ! (*termacts[node->tok].pre)(p, &pair, meta, node)) |
dochild = 0; |
dochild = 0; |
} else /* MDOC_TEXT == node->type */ |
} else /* MDOC_TEXT == node->type */ |
word(p, node->data.text.string); |
word(p, node->string); |
|
|
/* Children. */ |
/* Children. */ |
|
|
Line 493 footer(struct termp *p, const struct mdoc_meta *meta) |
|
Line 737 footer(struct termp *p, const struct mdoc_meta *meta) |
|
static void |
static void |
header(struct termp *p, const struct mdoc_meta *meta) |
header(struct termp *p, const struct mdoc_meta *meta) |
{ |
{ |
char *buf, *title, *bufp, *vbuf; |
char *buf, *title; |
const char *pp; |
|
struct utsname uts; |
|
|
|
p->rmargin = p->maxrmargin; |
p->rmargin = p->maxrmargin; |
p->offset = 0; |
p->offset = 0; |
Line 504 header(struct termp *p, const struct mdoc_meta *meta) |
|
Line 746 header(struct termp *p, const struct mdoc_meta *meta) |
|
err(1, "malloc"); |
err(1, "malloc"); |
if (NULL == (title = malloc(p->rmargin))) |
if (NULL == (title = malloc(p->rmargin))) |
err(1, "malloc"); |
err(1, "malloc"); |
if (NULL == (vbuf = malloc(p->rmargin))) |
|
err(1, "malloc"); |
|
|
|
if (NULL == (pp = mdoc_vol2a(meta->vol))) { |
|
switch (meta->msec) { |
|
case (MSEC_1): |
|
/* FALLTHROUGH */ |
|
case (MSEC_6): |
|
/* FALLTHROUGH */ |
|
case (MSEC_7): |
|
pp = mdoc_vol2a(VOL_URM); |
|
break; |
|
case (MSEC_8): |
|
pp = mdoc_vol2a(VOL_SMM); |
|
break; |
|
case (MSEC_2): |
|
/* FALLTHROUGH */ |
|
case (MSEC_3): |
|
/* FALLTHROUGH */ |
|
case (MSEC_4): |
|
/* FALLTHROUGH */ |
|
case (MSEC_5): |
|
pp = mdoc_vol2a(VOL_PRM); |
|
break; |
|
case (MSEC_9): |
|
pp = mdoc_vol2a(VOL_KM); |
|
break; |
|
default: |
|
break; |
|
} |
|
} |
|
vbuf[0] = 0; |
|
|
|
if (pp) { |
|
if (-1 == uname(&uts)) |
|
err(1, "uname"); |
|
(void)strlcat(vbuf, uts.sysname, p->rmargin); |
|
(void)strlcat(vbuf, " ", p->rmargin); |
|
} else if (NULL == (pp = mdoc_msec2a(meta->msec))) |
|
pp = mdoc_msec2a(MSEC_local); |
|
|
|
(void)strlcat(vbuf, pp, p->rmargin); |
|
|
|
/* |
/* |
* The header is strange. It has three components, which are |
* The header is strange. It has three components, which are |
* really two with the first duplicated. It goes like this: |
* really two with the first duplicated. It goes like this: |
Line 560 header(struct termp *p, const struct mdoc_meta *meta) |
|
Line 760 header(struct termp *p, const struct mdoc_meta *meta) |
|
* switches on the manual section. |
* switches on the manual section. |
*/ |
*/ |
|
|
if (mdoc_arch2a(meta->arch)) |
assert(meta->vol); |
(void)snprintf(buf, p->rmargin, "%s (%s)", |
(void)strlcpy(buf, meta->vol, p->rmargin); |
vbuf, mdoc_arch2a(meta->arch)); |
|
else |
|
(void)strlcpy(buf, vbuf, p->rmargin); |
|
|
|
pp = mdoc_msec2a(meta->msec); |
if (meta->arch) { |
|
(void)strlcat(buf, " (", p->rmargin); |
|
(void)strlcat(buf, meta->arch, p->rmargin); |
|
(void)strlcat(buf, ")", p->rmargin); |
|
} |
|
|
(void)snprintf(title, p->rmargin, "%s(%s)", |
(void)snprintf(title, p->rmargin, "%s(%d)", |
meta->title, pp ? pp : ""); |
meta->title, meta->msec); |
|
|
for (bufp = title; *bufp; bufp++) |
|
*bufp = toupper(*bufp); |
|
|
|
p->offset = 0; |
p->offset = 0; |
p->rmargin = (p->maxrmargin - strlen(buf)) / 2; |
p->rmargin = (p->maxrmargin - strlen(buf)) / 2; |
p->flags |= TERMP_NOBREAK | TERMP_NOSPACE; |
p->flags |= TERMP_NOBREAK | TERMP_NOSPACE; |
Line 601 header(struct termp *p, const struct mdoc_meta *meta) |
|
Line 799 header(struct termp *p, const struct mdoc_meta *meta) |
|
p->flags &= ~TERMP_NOSPACE; |
p->flags &= ~TERMP_NOSPACE; |
|
|
free(title); |
free(title); |
free(vbuf); |
|
free(buf); |
free(buf); |
} |
} |
|
|
Line 614 header(struct termp *p, const struct mdoc_meta *meta) |
|
Line 811 header(struct termp *p, const struct mdoc_meta *meta) |
|
static void |
static void |
nescape(struct termp *p, const char *word, size_t len) |
nescape(struct termp *p, const char *word, size_t len) |
{ |
{ |
|
const char *rhs; |
|
size_t sz; |
|
|
switch (len) { |
if (NULL == (rhs = a2ascii(p->symtab, word, len, &sz))) |
case (1): |
return; |
switch (word[0]) { |
stringa(p, rhs, sz); |
case ('\\'): |
|
/* FALLTHROUGH */ |
|
case ('\''): |
|
/* FALLTHROUGH */ |
|
case ('`'): |
|
/* FALLTHROUGH */ |
|
case ('-'): |
|
/* FALLTHROUGH */ |
|
case (' '): |
|
/* FALLTHROUGH */ |
|
case ('.'): |
|
chara(p, word[0]); /* FIXME */ |
|
break; |
|
case ('&'): |
|
break; |
|
case ('e'): |
|
chara(p, '\\'); /* FIXME */ |
|
break; |
|
case ('q'): |
|
symbola(p, TERMSYM_DQUOTE); |
|
break; |
|
default: |
|
warnx("escape sequence not supported: %c", |
|
word[0]); |
|
break; |
|
} |
|
break; |
|
|
|
case (2): |
|
if ('r' == word[0] && 'B' == word[1]) |
|
symbola(p, TERMSYM_RBRACK); |
|
else if ('l' == word[0] && 'B' == word[1]) |
|
symbola(p, TERMSYM_LBRACK); |
|
else if ('l' == word[0] && 'q' == word[1]) |
|
symbola(p, TERMSYM_LDQUOTE); |
|
else if ('r' == word[0] && 'q' == word[1]) |
|
symbola(p, TERMSYM_RDQUOTE); |
|
else if ('o' == word[0] && 'q' == word[1]) |
|
symbola(p, TERMSYM_LSQUOTE); |
|
else if ('a' == word[0] && 'q' == word[1]) |
|
symbola(p, TERMSYM_RSQUOTE); |
|
else if ('<' == word[0] && '-' == word[1]) |
|
symbola(p, TERMSYM_LARROW); |
|
else if ('-' == word[0] && '>' == word[1]) |
|
symbola(p, TERMSYM_RARROW); |
|
else if ('b' == word[0] && 'u' == word[1]) |
|
symbola(p, TERMSYM_BULLET); |
|
else if ('<' == word[0] && '=' == word[1]) |
|
symbola(p, TERMSYM_LE); |
|
else if ('>' == word[0] && '=' == word[1]) |
|
symbola(p, TERMSYM_GE); |
|
else if ('=' == word[0] && '=' == word[1]) |
|
symbola(p, TERMSYM_EQ); |
|
else if ('+' == word[0] && '-' == word[1]) |
|
symbola(p, TERMSYM_PLUSMINUS); |
|
else if ('u' == word[0] && 'a' == word[1]) |
|
symbola(p, TERMSYM_UARROW); |
|
else if ('d' == word[0] && 'a' == word[1]) |
|
symbola(p, TERMSYM_DARROW); |
|
else if ('a' == word[0] && 'a' == word[1]) |
|
symbola(p, TERMSYM_ACUTE); |
|
else if ('g' == word[0] && 'a' == word[1]) |
|
symbola(p, TERMSYM_GRAVE); |
|
else if ('!' == word[0] && '=' == word[1]) |
|
symbola(p, TERMSYM_NEQ); |
|
else if ('i' == word[0] && 'f' == word[1]) |
|
symbola(p, TERMSYM_INF); |
|
else if ('n' == word[0] && 'a' == word[1]) |
|
symbola(p, TERMSYM_NAN); |
|
else if ('b' == word[0] && 'a' == word[1]) |
|
symbola(p, TERMSYM_BAR); |
|
|
|
/* Deprecated forms. */ |
|
else if ('A' == word[0] && 'm' == word[1]) |
|
symbola(p, TERMSYM_AMP); |
|
else if ('B' == word[0] && 'a' == word[1]) |
|
symbola(p, TERMSYM_BAR); |
|
else if ('I' == word[0] && 'f' == word[1]) |
|
symbola(p, TERMSYM_INF2); |
|
else if ('G' == word[0] && 'e' == word[1]) |
|
symbola(p, TERMSYM_GE); |
|
else if ('G' == word[0] && 't' == word[1]) |
|
symbola(p, TERMSYM_GT); |
|
else if ('L' == word[0] && 'e' == word[1]) |
|
symbola(p, TERMSYM_LE); |
|
else if ('L' == word[0] && 'q' == word[1]) |
|
symbola(p, TERMSYM_LDQUOTE); |
|
else if ('L' == word[0] && 't' == word[1]) |
|
symbola(p, TERMSYM_LT); |
|
else if ('N' == word[0] && 'a' == word[1]) |
|
symbola(p, TERMSYM_NAN); |
|
else if ('N' == word[0] && 'e' == word[1]) |
|
symbola(p, TERMSYM_NEQ); |
|
else if ('P' == word[0] && 'i' == word[1]) |
|
symbola(p, TERMSYM_PI); |
|
else if ('P' == word[0] && 'm' == word[1]) |
|
symbola(p, TERMSYM_PLUSMINUS); |
|
else if ('R' == word[0] && 'q' == word[1]) |
|
symbola(p, TERMSYM_RDQUOTE); |
|
else |
|
warnx("escape sequence not supported: %c%c", |
|
word[0], word[1]); |
|
break; |
|
|
|
default: |
|
warnx("escape sequence not supported"); |
|
break; |
|
} |
|
} |
} |
|
|
|
|
/* |
/* |
* Apply a style to the output buffer. This is looked up by means of |
|
* the styletab. |
|
*/ |
|
static void |
|
style(struct termp *p, enum tstyle esc) |
|
{ |
|
|
|
if (p->col + 4 >= p->maxcols) |
|
errx(1, "line overrun"); |
|
|
|
p->buf[(p->col)++] = 27; |
|
p->buf[(p->col)++] = '['; |
|
switch (esc) { |
|
case (TERMSTYLE_CLEAR): |
|
p->buf[(p->col)++] = '0'; |
|
break; |
|
case (TERMSTYLE_BOLD): |
|
p->buf[(p->col)++] = '1'; |
|
break; |
|
case (TERMSTYLE_UNDER): |
|
p->buf[(p->col)++] = '4'; |
|
break; |
|
default: |
|
abort(); |
|
/* NOTREACHED */ |
|
} |
|
p->buf[(p->col)++] = 'm'; |
|
} |
|
|
|
|
|
/* |
|
* Handle an escape sequence: determine its length and pass it to the |
* Handle an escape sequence: determine its length and pass it to the |
* escape-symbol look table. Note that we assume mdoc(3) has validated |
* escape-symbol look table. Note that we assume mdoc(3) has validated |
* the escape sequence (we assert upon badly-formed escape sequences). |
* the escape sequence (we assert upon badly-formed escape sequences). |
*/ |
*/ |
static void |
static void |
pescape(struct termp *p, const char *word, size_t *i, size_t len) |
pescape(struct termp *p, const char *word, int *i, int len) |
{ |
{ |
size_t j; |
int j; |
|
|
(*i)++; |
if (++(*i) >= len) |
assert(*i < len); |
return; |
|
|
if ('(' == word[*i]) { |
if ('(' == word[*i]) { |
(*i)++; |
(*i)++; |
assert(*i + 1 < len); |
if (*i + 1 >= len) |
|
return; |
|
|
nescape(p, &word[*i], 2); |
nescape(p, &word[*i], 2); |
(*i)++; |
(*i)++; |
return; |
return; |
|
|
} else if ('*' == word[*i]) { |
} else if ('*' == word[*i]) { |
/* XXX - deprecated! */ |
|
(*i)++; |
(*i)++; |
assert(*i < len); |
if (*i >= len) |
|
return; |
|
|
switch (word[*i]) { |
switch (word[*i]) { |
case ('('): |
case ('('): |
(*i)++; |
(*i)++; |
assert(*i + 1 < len); |
if (*i + 1 >= len) |
|
return; |
|
|
nescape(p, &word[*i], 2); |
nescape(p, &word[*i], 2); |
(*i)++; |
(*i)++; |
return; |
return; |
Line 807 pescape(struct termp *p, const char *word, size_t *i, |
|
Line 872 pescape(struct termp *p, const char *word, size_t *i, |
|
for (j = 0; word[*i] && ']' != word[*i]; (*i)++, j++) |
for (j = 0; word[*i] && ']' != word[*i]; (*i)++, j++) |
/* Loop... */ ; |
/* Loop... */ ; |
|
|
assert(word[*i]); |
if (0 == word[*i]) |
nescape(p, &word[*i - j], j); |
return; |
|
|
|
nescape(p, &word[*i - j], (size_t)j); |
} |
} |
|
|
|
|
Line 818 pescape(struct termp *p, const char *word, size_t *i, |
|
Line 885 pescape(struct termp *p, const char *word, size_t *i, |
|
* handles word styling. |
* handles word styling. |
*/ |
*/ |
static void |
static void |
pword(struct termp *p, const char *word, size_t len) |
pword(struct termp *p, const char *word, int len) |
{ |
{ |
size_t i; |
int i; |
|
|
if ( ! (TERMP_NOSPACE & p->flags) && |
if ( ! (TERMP_NOSPACE & p->flags) && |
! (TERMP_LITERAL & p->flags)) |
! (TERMP_LITERAL & p->flags)) |
Line 830 pword(struct termp *p, const char *word, size_t len) |
|
Line 897 pword(struct termp *p, const char *word, size_t len) |
|
p->flags &= ~TERMP_NOSPACE; |
p->flags &= ~TERMP_NOSPACE; |
|
|
/* |
/* |
* XXX - if literal and underlining, this will underline the |
* If ANSI (word-length styling), then apply our style now, |
* spaces between literal words. |
* before the word. |
*/ |
*/ |
|
|
if (p->flags & TERMP_BOLD) |
|
style(p, TERMSTYLE_BOLD); |
|
if (p->flags & TERMP_UNDERLINE) |
|
style(p, TERMSTYLE_UNDER); |
|
|
|
for (i = 0; i < len; i++) { |
for (i = 0; i < len; i++) { |
if ('\\' == word[i]) { |
if ('\\' == word[i]) { |
pescape(p, word, &i, len); |
pescape(p, word, &i, len); |
continue; |
continue; |
} |
} |
|
|
|
if (TERMP_STYLE & p->flags) { |
|
if (TERMP_BOLD & p->flags) { |
|
chara(p, word[i]); |
|
chara(p, 8); |
|
} |
|
if (TERMP_UNDER & p->flags) { |
|
chara(p, '_'); |
|
chara(p, 8); |
|
} |
|
} |
|
|
chara(p, word[i]); |
chara(p, word[i]); |
} |
} |
|
|
if (p->flags & TERMP_BOLD || |
|
p->flags & TERMP_UNDERLINE) |
|
style(p, TERMSTYLE_CLEAR); |
|
} |
} |
|
|
|
|
/* |
/* |
* Add a symbol to the output line buffer. |
|
*/ |
|
static void |
|
symbola(struct termp *p, enum tsym sym) |
|
{ |
|
|
|
assert(p->symtab[sym].sym); |
|
stringa(p, p->symtab[sym].sym, p->symtab[sym].sz); |
|
} |
|
|
|
|
|
/* |
|
* Like chara() but for arbitrary-length buffers. Resize the buffer by |
* Like chara() but for arbitrary-length buffers. Resize the buffer by |
* a factor of two (if the buffer is less than that) or the buffer's |
* a factor of two (if the buffer is less than that) or the buffer's |
* size. |
* size. |
Line 875 stringa(struct termp *p, const char *c, size_t sz) |
|
Line 933 stringa(struct termp *p, const char *c, size_t sz) |
|
{ |
{ |
size_t s; |
size_t s; |
|
|
s = sz > p->maxcols * 2 ? sz : p->maxcols * 2; |
if (0 == sz) |
|
return; |
|
|
assert(c); |
assert(c); |
if (p->col + sz >= p->maxcols) { |
if (p->col + sz >= p->maxcols) { |
|
if (0 == p->maxcols) |
|
p->maxcols = 256; |
|
s = sz > p->maxcols * 2 ? sz : p->maxcols * 2; |
p->buf = realloc(p->buf, s); |
p->buf = realloc(p->buf, s); |
if (NULL == p->buf) |
if (NULL == p->buf) |
err(1, "realloc"); |
err(1, "realloc"); |
p->maxcols = s; |
p->maxcols = s; |
} |
} |
|
|
(void)memcpy(&p->buf[p->col], c, sz); |
(void)memcpy(&p->buf[(int)p->col], c, sz); |
p->col += sz; |
p->col += sz; |
} |
} |
|
|
Line 898 stringa(struct termp *p, const char *c, size_t sz) |
|
Line 960 stringa(struct termp *p, const char *c, size_t sz) |
|
static void |
static void |
chara(struct termp *p, char c) |
chara(struct termp *p, char c) |
{ |
{ |
|
size_t s; |
|
|
if (p->col + 1 >= p->maxcols) { |
if (p->col + 1 >= p->maxcols) { |
p->buf = realloc(p->buf, p->maxcols * 2); |
if (0 == p->maxcols) |
|
p->maxcols = 256; |
|
s = p->maxcols * 2; |
|
p->buf = realloc(p->buf, s); |
if (NULL == p->buf) |
if (NULL == p->buf) |
err(1, "malloc"); |
err(1, "realloc"); |
p->maxcols *= 2; |
p->maxcols = s; |
} |
} |
p->buf[(p->col)++] = c; |
p->buf[(int)(p->col)++] = c; |
} |
} |
|
|
|
|
|
static void |
|
sanity(const struct mdoc_node *n) |
|
{ |
|
|
|
switch (n->type) { |
|
case (MDOC_TEXT): |
|
if (n->child) |
|
errx(1, "regular form violated (1)"); |
|
if (NULL == n->parent) |
|
errx(1, "regular form violated (2)"); |
|
if (NULL == n->string) |
|
errx(1, "regular form violated (3)"); |
|
switch (n->parent->type) { |
|
case (MDOC_TEXT): |
|
/* FALLTHROUGH */ |
|
case (MDOC_ROOT): |
|
errx(1, "regular form violated (4)"); |
|
/* NOTREACHED */ |
|
default: |
|
break; |
|
} |
|
break; |
|
case (MDOC_ELEM): |
|
if (NULL == n->parent) |
|
errx(1, "regular form violated (5)"); |
|
switch (n->parent->type) { |
|
case (MDOC_TAIL): |
|
/* FALLTHROUGH */ |
|
case (MDOC_BODY): |
|
/* FALLTHROUGH */ |
|
case (MDOC_HEAD): |
|
break; |
|
default: |
|
errx(1, "regular form violated (6)"); |
|
/* NOTREACHED */ |
|
} |
|
if (n->child) switch (n->child->type) { |
|
case (MDOC_TEXT): |
|
break; |
|
default: |
|
errx(1, "regular form violated (7("); |
|
/* NOTREACHED */ |
|
} |
|
break; |
|
case (MDOC_HEAD): |
|
/* FALLTHROUGH */ |
|
case (MDOC_BODY): |
|
/* FALLTHROUGH */ |
|
case (MDOC_TAIL): |
|
if (NULL == n->parent) |
|
errx(1, "regular form violated (8)"); |
|
if (MDOC_BLOCK != n->parent->type) |
|
errx(1, "regular form violated (9)"); |
|
if (n->child) switch (n->child->type) { |
|
case (MDOC_BLOCK): |
|
/* FALLTHROUGH */ |
|
case (MDOC_ELEM): |
|
/* FALLTHROUGH */ |
|
case (MDOC_TEXT): |
|
break; |
|
default: |
|
errx(1, "regular form violated (a)"); |
|
/* NOTREACHED */ |
|
} |
|
break; |
|
case (MDOC_BLOCK): |
|
if (NULL == n->parent) |
|
errx(1, "regular form violated (b)"); |
|
if (NULL == n->child) |
|
errx(1, "regular form violated (c)"); |
|
switch (n->parent->type) { |
|
case (MDOC_ROOT): |
|
/* FALLTHROUGH */ |
|
case (MDOC_HEAD): |
|
/* FALLTHROUGH */ |
|
case (MDOC_BODY): |
|
/* FALLTHROUGH */ |
|
case (MDOC_TAIL): |
|
break; |
|
default: |
|
errx(1, "regular form violated (d)"); |
|
/* NOTREACHED */ |
|
} |
|
switch (n->child->type) { |
|
case (MDOC_ROOT): |
|
/* FALLTHROUGH */ |
|
case (MDOC_ELEM): |
|
errx(1, "regular form violated (e)"); |
|
/* NOTREACHED */ |
|
default: |
|
break; |
|
} |
|
break; |
|
case (MDOC_ROOT): |
|
if (n->parent) |
|
errx(1, "regular form violated (f)"); |
|
if (NULL == n->child) |
|
errx(1, "regular form violated (10)"); |
|
switch (n->child->type) { |
|
case (MDOC_BLOCK): |
|
break; |
|
default: |
|
errx(1, "regular form violated (11)"); |
|
/* NOTREACHED */ |
|
} |
|
break; |
|
} |
|
} |
|
|
|
|
|
static int |
|
merr(void *arg, int line, int col, const char *msg) |
|
{ |
|
|
|
warnx("error: %s (line %d, column %d)", msg, line, col); |
|
return(0); |
|
} |
|
|
|
|
|
static int |
|
mwarn(void *arg, int line, int col, |
|
enum mdoc_warn type, const char *msg) |
|
{ |
|
int flags; |
|
char *wtype; |
|
|
|
flags = *(int *)arg; |
|
wtype = NULL; |
|
|
|
switch (type) { |
|
case (WARN_COMPAT): |
|
wtype = "compat"; |
|
if (flags & WARN_WCOMPAT) |
|
break; |
|
return(1); |
|
case (WARN_SYNTAX): |
|
wtype = "syntax"; |
|
if (flags & WARN_WSYNTAX) |
|
break; |
|
return(1); |
|
} |
|
|
|
assert(wtype); |
|
warnx("%s warning: %s (line %d, column %d)", |
|
wtype, msg, line, col); |
|
|
|
if ( ! (flags & WARN_WERR)) |
|
return(1); |
|
|
|
warnx("%s: considering warnings as errors", |
|
__progname); |
|
return(0); |
|
} |
|
|
|
|