-/* $Id: term.c,v 1.178 2011/03/15 16:23:51 kristaps Exp $ */
+/* $Id: term.c,v 1.273 2017/06/14 23:24:18 schwarze Exp $ */
/*
- * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
- * Copyright (c) 2010, 2011 Ingo Schwarze <schwarze@openbsd.org>
+ * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
+ * Copyright (c) 2010-2017 Ingo Schwarze <schwarze@openbsd.org>
*
* 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.
*
- * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
- * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR
* 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.
*/
-#ifdef HAVE_CONFIG_H
#include "config.h"
-#endif
#include <sys/types.h>
#include <assert.h>
#include <ctype.h>
-#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "mandoc.h"
-#include "chars.h"
+#include "mandoc_aux.h"
#include "out.h"
#include "term.h"
#include "main.h"
-static void spec(struct termp *, enum roffdeco,
- const char *, size_t);
-static void res(struct termp *, const char *, size_t);
-static void bufferc(struct termp *, char);
-static void adjbuf(struct termp *p, size_t);
-static void encode(struct termp *, const char *, size_t);
+static size_t cond_width(const struct termp *, int, int *);
+static void adjbuf(struct termp_col *, size_t);
+static void bufferc(struct termp *, char);
+static void encode(struct termp *, const char *, size_t);
+static void encode1(struct termp *, int);
+static void endline(struct termp *);
void
-term_free(struct termp *p)
+term_setcol(struct termp *p, size_t maxtcol)
{
+ if (maxtcol > p->maxtcol) {
+ p->tcols = mandoc_recallocarray(p->tcols,
+ p->maxtcol, maxtcol, sizeof(*p->tcols));
+ p->maxtcol = maxtcol;
+ }
+ p->lasttcol = maxtcol - 1;
+ p->tcol = p->tcols;
+}
- if (p->buf)
- free(p->buf);
- if (p->symtab)
- chars_free(p->symtab);
-
+void
+term_free(struct termp *p)
+{
+ for (p->tcol = p->tcols; p->tcol < p->tcols + p->maxtcol; p->tcol++)
+ free(p->tcol->buf);
+ free(p->tcols);
+ free(p->fontq);
free(p);
}
-
void
-term_begin(struct termp *p, term_margin head,
- term_margin foot, const void *arg)
+term_begin(struct termp *p, term_margin head,
+ term_margin foot, const struct roff_meta *arg)
{
p->headf = head;
(*p->begin)(p);
}
-
void
term_end(struct termp *p)
{
(*p->end)(p);
}
-
-struct termp *
-term_alloc(enum termenc enc)
-{
- struct termp *p;
-
- p = calloc(1, sizeof(struct termp));
- if (NULL == p) {
- perror(NULL);
- exit((int)MANDOCLEVEL_SYSERR);
- }
-
- p->enc = enc;
- return(p);
-}
-
-
/*
- * 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
- * broken apart by newlines getting there. A line can also be a
- * fragment of a columnar list (`Bl -tag' or `Bl -column'), which does
- * not have a trailing newline.
- *
+ * Flush a chunk of text. By default, break the output line each time
+ * the right margin is reached, and continue output on the next line
+ * at the same offset as the chunk itself. By default, also break the
+ * output line at the end of the chunk.
* The following flags may be specified:
*
- * - TERMP_NOLPAD: when beginning to write the line, don't left-pad the
- * offset value. This is useful when doing columnar lists where the
- * prior column has right-padded.
- *
- * - 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
- * right margin. Used in conjunction with TERMP_NOLPAD.
- *
- * - TERMP_TWOSPACE: when padding, make sure there are at least two
- * space characters of padding. Otherwise, rather break the line.
- *
- * - TERMP_DANGLE: don't newline when TERMP_NOBREAK is specified and
- * the line is overrun, and don't pad-right if it's underrun.
- *
- * - TERMP_HANG: like TERMP_DANGLE, but doesn't newline when
- * overruning, instead save the position and continue at that point
- * when the next invocation.
- *
- * In-line line breaking:
- *
- * If TERMP_NOBREAK is specified and the line overruns the right
- * margin, it will break and pad-right to the right margin after
- * writing. If maxrmargin is violated, it will break and continue
- * writing from the right-margin, which will lead to the above scenario
- * upon exit. Otherwise, the line will break at the right margin.
+ * - TERMP_NOBREAK: Do not break the output line at the right margin,
+ * but only at the max right margin. Also, do not break the output
+ * line at the end of the chunk, such that the next call can pad to
+ * the next column. However, if less than p->trailspace blanks,
+ * which can be 0, 1, or 2, remain to the right margin, the line
+ * will be broken.
+ * - TERMP_BRTRSP: Consider trailing whitespace significant
+ * when deciding whether the chunk fits or not.
+ * - TERMP_BRIND: If the chunk does not fit and the output line has
+ * to be broken, start the next line at the right margin instead
+ * of at the offset. Used together with TERMP_NOBREAK for the tags
+ * in various kinds of tagged lists.
+ * - TERMP_HANG: Do not break the output line at the right margin,
+ * append the next chunk after it even if this one is too long.
+ * To be used together with TERMP_NOBREAK.
+ * - TERMP_NOPAD: Start writing at the current position,
+ * do not pad with blank characters up to the offset.
*/
void
term_flushln(struct termp *p)
{
- int i; /* current input position in p->buf */
size_t vis; /* current visual position on output */
size_t vbl; /* number of blanks to prepend to output */
size_t vend; /* end of word visual position on output */
size_t bp; /* visual right border position */
size_t dv; /* temporary for visual pos calculations */
- int j; /* temporary loop index for p->buf */
- int jhy; /* last hyph before overflow w/r/t j */
+ size_t j; /* temporary loop index for p->tcol->buf */
+ size_t jhy; /* last hyph before overflow w/r/t j */
size_t maxvis; /* output position of visible boundary */
- size_t mmax; /* used in calculating bp */
-
- /*
- * First, establish the maximum columns of "visible" content.
- * This is usually the difference between the right-margin and
- * an indentation, but can be, for tagged lists or columns, a
- * small set of values.
- */
- assert (p->rmargin >= p->offset);
- dv = p->rmargin - p->offset;
- maxvis = (int)dv > p->overstep ? dv - (size_t)p->overstep : 0;
- dv = p->maxrmargin - p->offset;
- mmax = (int)dv > p->overstep ? dv - (size_t)p->overstep : 0;
-
- bp = TERMP_NOBREAK & p->flags ? mmax : maxvis;
-
- /*
- * Indent the first line of a paragraph.
- */
- vbl = p->flags & TERMP_NOLPAD ? (size_t)0 : p->offset;
-
+ int ntab; /* number of tabs to prepend */
+ int breakline; /* after this word */
+
+ vbl = (p->flags & TERMP_NOPAD) || p->tcol->offset < p->viscol ?
+ 0 : p->tcol->offset - p->viscol;
+ if (p->minbl && vbl < p->minbl)
+ vbl = p->minbl;
+ maxvis = p->tcol->rmargin > p->viscol + vbl ?
+ p->tcol->rmargin - p->viscol - vbl : 0;
+ bp = !(p->flags & TERMP_NOBREAK) ? maxvis :
+ p->maxrmargin > p->viscol + vbl ?
+ p->maxrmargin - p->viscol - vbl : 0;
vis = vend = 0;
- i = 0;
- while (i < (int)p->col) {
+ if ((p->flags && TERMP_MULTICOL) == 0)
+ p->tcol->col = 0;
+ while (p->tcol->col < p->tcol->lastcol) {
+
/*
* Handle literal tab characters: collapse all
* subsequent tabs into a single huge set of spaces.
*/
- while (i < (int)p->col && '\t' == p->buf[i]) {
- vend = (vis / p->tabwidth + 1) * p->tabwidth;
+
+ ntab = 0;
+ while (p->tcol->col < p->tcol->lastcol &&
+ p->tcol->buf[p->tcol->col] == '\t') {
+ vend = term_tab_next(vis);
vbl += vend - vis;
vis = vend;
- i++;
+ ntab++;
+ p->tcol->col++;
}
/*
* space is printed according to regular spacing rules).
*/
- for (j = i, jhy = 0; j < (int)p->col; j++) {
- if ((j && ' ' == p->buf[j]) || '\t' == p->buf[j])
+ jhy = 0;
+ breakline = 0;
+ for (j = p->tcol->col; j < p->tcol->lastcol; j++) {
+ if (p->tcol->buf[j] == '\n') {
+ if ((p->flags & TERMP_BRIND) == 0)
+ breakline = 1;
+ continue;
+ }
+ if (p->tcol->buf[j] == ' ' || p->tcol->buf[j] == '\t')
break;
- /* Back over the the last printed character. */
- if (8 == p->buf[j]) {
+ /* Back over the last printed character. */
+ if (p->tcol->buf[j] == '\b') {
assert(j);
- vend -= (*p->width)(p, p->buf[j - 1]);
+ vend -= (*p->width)(p, p->tcol->buf[j - 1]);
continue;
}
/* Regular word. */
/* Break at the hyphen point if we overrun. */
- if (vend > vis && vend < bp &&
- ASCII_HYPH == p->buf[j])
+ if (vend > vis && vend < bp &&
+ (p->tcol->buf[j] == ASCII_HYPH||
+ p->tcol->buf[j] == ASCII_BREAK))
jhy = j;
- vend += (*p->width)(p, p->buf[j]);
+ /*
+ * Hyphenation now decided, put back a real
+ * hyphen such that we get the correct width.
+ */
+ if (p->tcol->buf[j] == ASCII_HYPH)
+ p->tcol->buf[j] = '-';
+
+ vend += (*p->width)(p, p->tcol->buf[j]);
}
/*
* Find out whether we would exceed the right margin.
* If so, break to the next line.
*/
- if (vend > bp && 0 == jhy && vis > 0) {
+
+ if (vend > bp && jhy == 0 && vis > 0 &&
+ (p->flags & TERMP_BRNEVER) == 0) {
+ if (p->flags & TERMP_MULTICOL)
+ return;
+
+ endline(p);
vend -= vis;
- (*p->endline)(p);
- if (TERMP_NOBREAK & p->flags) {
- p->viscol = p->rmargin;
- (*p->advance)(p, p->rmargin);
- vend += p->rmargin - p->offset;
- } else {
- p->viscol = 0;
- vbl = p->offset;
- }
- /* Remove the p->overstep width. */
+ /* Use pending tabs on the new line. */
- bp += (size_t)p->overstep;
- p->overstep = 0;
+ vbl = 0;
+ while (ntab--)
+ vbl = term_tab_next(vbl);
+
+ /* Re-establish indentation. */
+
+ if (p->flags & TERMP_BRIND)
+ vbl += p->tcol->rmargin;
+ else
+ vbl += p->tcol->offset;
+ maxvis = p->tcol->rmargin > vbl ?
+ p->tcol->rmargin - vbl : 0;
+ bp = !(p->flags & TERMP_NOBREAK) ? maxvis :
+ p->maxrmargin > vbl ? p->maxrmargin - vbl : 0;
}
- /* Write out the [remaining] word. */
- for ( ; i < (int)p->col; i++) {
- if (vend > bp && jhy > 0 && i > jhy)
+ /*
+ * Write out the rest of the word.
+ */
+
+ for ( ; p->tcol->col < p->tcol->lastcol; p->tcol->col++) {
+ if (vend > bp && jhy > 0 && p->tcol->col > jhy)
break;
- if ('\t' == p->buf[i])
+ if (p->tcol->buf[p->tcol->col] == '\n')
+ continue;
+ if (p->tcol->buf[p->tcol->col] == '\t')
break;
- if (' ' == p->buf[i]) {
- j = i;
- while (' ' == p->buf[i])
- i++;
- dv = (size_t)(i - j) * (*p->width)(p, ' ');
+ if (p->tcol->buf[p->tcol->col] == ' ') {
+ j = p->tcol->col;
+ while (p->tcol->col < p->tcol->lastcol &&
+ p->tcol->buf[p->tcol->col] == ' ')
+ p->tcol->col++;
+ dv = (p->tcol->col - j) * (*p->width)(p, ' ');
vbl += dv;
vend += dv;
break;
}
- if (ASCII_NBRSP == p->buf[i]) {
+ if (p->tcol->buf[p->tcol->col] == ASCII_NBRSP) {
vbl += (*p->width)(p, ' ');
continue;
}
+ if (p->tcol->buf[p->tcol->col] == ASCII_BREAK)
+ continue;
/*
* Now we definitely know there will be
vbl = 0;
}
- if (ASCII_HYPH == p->buf[i]) {
- (*p->letter)(p, '-');
- p->viscol += (*p->width)(p, '-');
- } else {
- (*p->letter)(p, p->buf[i]);
- p->viscol += (*p->width)(p, p->buf[i]);
- }
+ (*p->letter)(p, p->tcol->buf[p->tcol->col]);
+ if (p->tcol->buf[p->tcol->col] == '\b')
+ p->viscol -= (*p->width)(p,
+ p->tcol->buf[p->tcol->col - 1]);
+ else
+ p->viscol += (*p->width)(p,
+ p->tcol->buf[p->tcol->col]);
}
vis = vend;
+
+ if (breakline == 0)
+ continue;
+
+ /* Explicitly requested output line break. */
+
+ if (p->flags & TERMP_MULTICOL)
+ return;
+
+ endline(p);
+ breakline = 0;
+ vis = vend = 0;
+
+ /* Re-establish indentation. */
+
+ vbl = p->tcol->offset;
+ maxvis = p->tcol->rmargin > vbl ?
+ p->tcol->rmargin - vbl : 0;
+ bp = !(p->flags & TERMP_NOBREAK) ? maxvis :
+ p->maxrmargin > vbl ? p->maxrmargin - vbl : 0;
}
/*
* If there was trailing white space, it was not printed;
* so reset the cursor position accordingly.
*/
- vis -= vbl;
- p->col = 0;
- p->overstep = 0;
+ if (vis > vbl)
+ vis -= vbl;
+ else
+ vis = 0;
- if ( ! (TERMP_NOBREAK & p->flags)) {
- p->viscol = 0;
- (*p->endline)(p);
+ p->col = p->tcol->col = p->tcol->lastcol = 0;
+ p->minbl = p->trailspace;
+ p->flags &= ~(TERMP_BACKAFTER | TERMP_BACKBEFORE | TERMP_NOPAD);
+
+ if (p->flags & TERMP_MULTICOL)
return;
- }
- if (TERMP_HANG & p->flags) {
- /* We need one blank after the tag. */
- p->overstep = (int)(vis - maxvis + (*p->width)(p, ' '));
+ /* Trailing whitespace is significant in some columns. */
- /*
- * Behave exactly the same way as groff:
- * If we have overstepped the margin, temporarily move
- * it to the right and flag the rest of the line to be
- * shorter.
- * If we landed right at the margin, be happy.
- * If we are one step before the margin, temporarily
- * move it one step LEFT and flag the rest of the line
- * to be longer.
- */
- if (p->overstep >= -1) {
- assert((int)maxvis + p->overstep >= 0);
- maxvis += (size_t)p->overstep;
- } else
- p->overstep = 0;
+ if (vis && vbl && (TERMP_BRTRSP & p->flags))
+ vis += vbl;
- } else if (TERMP_DANGLE & p->flags)
- return;
+ /* If the column was overrun, break the line. */
+ if ((p->flags & TERMP_NOBREAK) == 0 ||
+ ((p->flags & TERMP_HANG) == 0 &&
+ vis + p->trailspace * (*p->width)(p, ' ') > maxvis))
+ endline(p);
+}
- /* Right-pad. */
- if (maxvis > vis +
- ((TERMP_TWOSPACE & p->flags) ? (*p->width)(p, ' ') : 0)) {
- p->viscol += maxvis - vis;
- (*p->advance)(p, maxvis - vis);
- vis += (maxvis - vis);
- } else { /* ...or newline break. */
- (*p->endline)(p);
- p->viscol = p->rmargin;
- (*p->advance)(p, p->rmargin);
+static void
+endline(struct termp *p)
+{
+ if ((p->flags & (TERMP_NEWMC | TERMP_ENDMC)) == TERMP_ENDMC) {
+ p->mc = NULL;
+ p->flags &= ~TERMP_ENDMC;
}
+ if (p->mc != NULL) {
+ if (p->viscol && p->maxrmargin >= p->viscol)
+ (*p->advance)(p, p->maxrmargin - p->viscol + 1);
+ p->flags |= TERMP_NOBUF | TERMP_NOSPACE;
+ term_word(p, p->mc);
+ p->flags &= ~(TERMP_NOBUF | TERMP_NEWMC);
+ }
+ p->viscol = 0;
+ p->minbl = 0;
+ (*p->endline)(p);
}
-
-/*
+/*
* A newline only breaks an existing line; it won't assert vertical
* space. All data in the output buffer is flushed prior to the newline
* assertion.
{
p->flags |= TERMP_NOSPACE;
- if (0 == p->col && 0 == p->viscol) {
- p->flags &= ~TERMP_NOLPAD;
- return;
- }
- term_flushln(p);
- p->flags &= ~TERMP_NOLPAD;
+ if (p->tcol->lastcol || p->viscol)
+ term_flushln(p);
}
-
/*
* Asserts a vertical space (a full, empty line-break between lines).
* Note that if used twice, this will cause two blank spaces and so on.
term_newln(p);
p->viscol = 0;
- (*p->endline)(p);
-}
-
-
-static void
-numbered(struct termp *p, const char *word, size_t len)
-{
- const char *rhs;
-
- rhs = chars_num2char(word, len);
- if (rhs)
- encode(p, rhs, 1);
-}
-
-
-static void
-spec(struct termp *p, enum roffdeco d, const char *word, size_t len)
-{
- const char *rhs;
- size_t sz;
-
- rhs = chars_spec2str(p->symtab, word, len, &sz);
- if (rhs)
- encode(p, rhs, sz);
- else if (DECO_SSPECIAL == d)
- encode(p, word, len);
-}
-
-
-static void
-res(struct termp *p, const char *word, size_t len)
-{
- const char *rhs;
- size_t sz;
-
- rhs = chars_res2str(p->symtab, word, len, &sz);
- if (rhs)
- encode(p, rhs, sz);
+ p->minbl = 0;
+ if (0 < p->skipvsp)
+ p->skipvsp--;
+ else
+ (*p->endline)(p);
}
-
+/* Swap current and previous font; for \fP and .ft P */
void
term_fontlast(struct termp *p)
{
p->fontq[p->fonti] = f;
}
-
+/* Set font, save current, discard previous; for \f, .ft, .B etc. */
void
term_fontrepl(struct termp *p, enum termfont f)
{
p->fontq[p->fonti] = f;
}
-
+/* Set font, save previous. */
void
term_fontpush(struct termp *p, enum termfont f)
{
- assert(p->fonti + 1 < 10);
p->fontl = p->fontq[p->fonti];
- p->fontq[++p->fonti] = f;
-}
-
-
-const void *
-term_fontq(struct termp *p)
-{
-
- return(&p->fontq[p->fonti]);
-}
-
-
-enum termfont
-term_fonttop(struct termp *p)
-{
-
- return(p->fontq[p->fonti]);
+ if (++p->fonti == p->fontsz) {
+ p->fontsz += 8;
+ p->fontq = mandoc_reallocarray(p->fontq,
+ p->fontsz, sizeof(*p->fontq));
+ }
+ p->fontq[p->fonti] = f;
}
-
+/* Flush to make the saved pointer current again. */
void
-term_fontpopq(struct termp *p, const void *key)
+term_fontpopq(struct termp *p, int i)
{
- while (p->fonti >= 0 && key != &p->fontq[p->fonti])
- p->fonti--;
- assert(p->fonti >= 0);
+ assert(i >= 0);
+ if (p->fonti > i)
+ p->fonti = i;
}
-
+/* Pop one font off the stack. */
void
term_fontpop(struct termp *p)
{
p->fonti--;
}
-
/*
* Handle pwords, partial words, which may be either a single word or a
* phrase that cannot be broken down (such as a literal string). This
void
term_word(struct termp *p, const char *word)
{
- const char *sv, *seq;
- size_t ssz;
- enum roffdeco deco;
-
- sv = word;
-
- if (word[0] && '\0' == word[1])
- switch (word[0]) {
- case('.'):
- /* FALLTHROUGH */
- case(','):
- /* FALLTHROUGH */
- case(';'):
- /* FALLTHROUGH */
- case(':'):
- /* FALLTHROUGH */
- case('?'):
- /* FALLTHROUGH */
- case('!'):
- /* FALLTHROUGH */
- case(')'):
- /* FALLTHROUGH */
- case(']'):
- if ( ! (TERMP_IGNDELIM & p->flags))
- p->flags |= TERMP_NOSPACE;
- break;
- default:
- break;
- }
-
- if ( ! (TERMP_NOSPACE & p->flags)) {
- if ( ! (TERMP_KEEP & p->flags)) {
- if (TERMP_PREKEEP & p->flags)
- p->flags |= TERMP_KEEP;
- bufferc(p, ' ');
- if (TERMP_SENTENCE & p->flags)
+ struct roffsu su;
+ const char nbrsp[2] = { ASCII_NBRSP, 0 };
+ const char *seq, *cp;
+ int sz, uc;
+ size_t csz, lsz, ssz;
+ enum mandoc_esc esc;
+
+ if ((p->flags & TERMP_NOBUF) == 0) {
+ if ((p->flags & TERMP_NOSPACE) == 0) {
+ if ((p->flags & TERMP_KEEP) == 0) {
bufferc(p, ' ');
- } else
- bufferc(p, ASCII_NBRSP);
+ if (p->flags & TERMP_SENTENCE)
+ bufferc(p, ' ');
+ } else
+ bufferc(p, ASCII_NBRSP);
+ }
+ if (p->flags & TERMP_PREKEEP)
+ p->flags |= TERMP_KEEP;
+ if (p->flags & TERMP_NONOSPACE)
+ p->flags |= TERMP_NOSPACE;
+ else
+ p->flags &= ~TERMP_NOSPACE;
+ p->flags &= ~(TERMP_SENTENCE | TERMP_NONEWLINE);
+ p->skipvsp = 0;
}
- if ( ! (p->flags & TERMP_NONOSPACE))
- p->flags &= ~TERMP_NOSPACE;
- else
- p->flags |= TERMP_NOSPACE;
-
- p->flags &= ~(TERMP_SENTENCE | TERMP_IGNDELIM);
-
- while (*word) {
- if ((ssz = strcspn(word, "\\")) > 0)
+ while ('\0' != *word) {
+ if ('\\' != *word) {
+ if (TERMP_NBRWORD & p->flags) {
+ if (' ' == *word) {
+ encode(p, nbrsp, 1);
+ word++;
+ continue;
+ }
+ ssz = strcspn(word, "\\ ");
+ } else
+ ssz = strcspn(word, "\\");
encode(p, word, ssz);
-
- word += (int)ssz;
- if ('\\' != *word)
+ word += (int)ssz;
continue;
+ }
- seq = ++word;
- word += a2roffdeco(&deco, &seq, &ssz);
+ word++;
+ esc = mandoc_escape(&word, &seq, &sz);
+ if (ESCAPE_ERROR == esc)
+ continue;
- switch (deco) {
- case (DECO_NUMBERED):
- numbered(p, seq, ssz);
+ switch (esc) {
+ case ESCAPE_UNICODE:
+ uc = mchars_num2uc(seq + 1, sz - 1);
break;
- case (DECO_RESERVED):
- res(p, seq, ssz);
- break;
- case (DECO_SPECIAL):
- /* FALLTHROUGH */
- case (DECO_SSPECIAL):
- spec(p, deco, seq, ssz);
+ case ESCAPE_NUMBERED:
+ uc = mchars_num2char(seq, sz);
+ if (uc < 0)
+ continue;
break;
- case (DECO_BOLD):
+ case ESCAPE_SPECIAL:
+ if (p->enc == TERMENC_ASCII) {
+ cp = mchars_spec2str(seq, sz, &ssz);
+ if (cp != NULL)
+ encode(p, cp, ssz);
+ } else {
+ uc = mchars_spec2cp(seq, sz);
+ if (uc > 0)
+ encode1(p, uc);
+ }
+ continue;
+ case ESCAPE_FONTBOLD:
term_fontrepl(p, TERMFONT_BOLD);
- break;
- case (DECO_ITALIC):
+ continue;
+ case ESCAPE_FONTITALIC:
term_fontrepl(p, TERMFONT_UNDER);
- break;
- case (DECO_ROMAN):
+ continue;
+ case ESCAPE_FONTBI:
+ term_fontrepl(p, TERMFONT_BI);
+ continue;
+ case ESCAPE_FONT:
+ case ESCAPE_FONTROMAN:
term_fontrepl(p, TERMFONT_NONE);
- break;
- case (DECO_PREVIOUS):
+ continue;
+ case ESCAPE_FONTPREV:
term_fontlast(p);
- break;
+ continue;
+ case ESCAPE_BREAK:
+ bufferc(p, '\n');
+ continue;
+ case ESCAPE_NOSPACE:
+ if (p->flags & TERMP_BACKAFTER)
+ p->flags &= ~TERMP_BACKAFTER;
+ else if (*word == '\0')
+ p->flags |= (TERMP_NOSPACE | TERMP_NONEWLINE);
+ continue;
+ case ESCAPE_HORIZ:
+ if (*seq == '|') {
+ seq++;
+ uc = -p->col;
+ } else
+ uc = 0;
+ if (a2roffsu(seq, &su, SCALE_EM) == NULL)
+ continue;
+ uc += term_hen(p, &su);
+ if (uc > 0)
+ while (uc-- > 0)
+ bufferc(p, ASCII_NBRSP);
+ else if (p->col > (size_t)(-uc))
+ p->col += uc;
+ else {
+ uc += p->col;
+ p->col = 0;
+ if (p->tcol->offset > (size_t)(-uc)) {
+ p->ti += uc;
+ p->tcol->offset += uc;
+ } else {
+ p->ti -= p->tcol->offset;
+ p->tcol->offset = 0;
+ }
+ }
+ continue;
+ case ESCAPE_HLINE:
+ if ((cp = a2roffsu(seq, &su, SCALE_EM)) == NULL)
+ continue;
+ uc = term_hen(p, &su);
+ if (uc <= 0) {
+ if (p->tcol->rmargin <= p->tcol->offset)
+ continue;
+ lsz = p->tcol->rmargin - p->tcol->offset;
+ } else
+ lsz = uc;
+ if (*cp == seq[-1])
+ uc = -1;
+ else if (*cp == '\\') {
+ seq = cp + 1;
+ esc = mandoc_escape(&seq, &cp, &sz);
+ switch (esc) {
+ case ESCAPE_UNICODE:
+ uc = mchars_num2uc(cp + 1, sz - 1);
+ break;
+ case ESCAPE_NUMBERED:
+ uc = mchars_num2char(cp, sz);
+ break;
+ case ESCAPE_SPECIAL:
+ uc = mchars_spec2cp(cp, sz);
+ break;
+ default:
+ uc = -1;
+ break;
+ }
+ } else
+ uc = *cp;
+ if (uc < 0x20 || (uc > 0x7E && uc < 0xA0))
+ uc = '_';
+ if (p->enc == TERMENC_ASCII) {
+ cp = ascii_uc2str(uc);
+ csz = term_strlen(p, cp);
+ ssz = strlen(cp);
+ } else
+ csz = (*p->width)(p, uc);
+ while (lsz >= csz) {
+ if (p->enc == TERMENC_ASCII)
+ encode(p, cp, ssz);
+ else
+ encode1(p, uc);
+ lsz -= csz;
+ }
+ continue;
+ case ESCAPE_SKIPCHAR:
+ p->flags |= TERMP_BACKAFTER;
+ continue;
+ case ESCAPE_OVERSTRIKE:
+ cp = seq + sz;
+ while (seq < cp) {
+ if (*seq == '\\') {
+ mandoc_escape(&seq, NULL, NULL);
+ continue;
+ }
+ encode1(p, *seq++);
+ if (seq < cp) {
+ if (p->flags & TERMP_BACKBEFORE)
+ p->flags |= TERMP_BACKAFTER;
+ else
+ p->flags |= TERMP_BACKBEFORE;
+ }
+ }
+ /* Trim trailing backspace/blank pair. */
+ if (p->tcol->lastcol > 2 &&
+ (p->tcol->buf[p->tcol->lastcol - 1] == ' ' ||
+ p->tcol->buf[p->tcol->lastcol - 1] == '\t'))
+ p->tcol->lastcol -= 2;
+ if (p->col > p->tcol->lastcol)
+ p->col = p->tcol->lastcol;
+ continue;
default:
- break;
+ continue;
}
- if (DECO_NOSPACE == deco && '\0' == *word)
- p->flags |= TERMP_NOSPACE;
- }
+ /*
+ * Common handling for Unicode and numbered
+ * character escape sequences.
+ */
- /*
- * Note that we don't process the pipe: the parser sees it as
- * punctuation, but we don't in terms of typography.
- */
- if (sv[0] && '\0' == sv[1])
- switch (sv[0]) {
- case('('):
- /* FALLTHROUGH */
- case('['):
- p->flags |= TERMP_NOSPACE;
- break;
- default:
- break;
+ if (p->enc == TERMENC_ASCII) {
+ cp = ascii_uc2str(uc);
+ encode(p, cp, strlen(cp));
+ } else {
+ if ((uc < 0x20 && uc != 0x09) ||
+ (uc > 0x7E && uc < 0xA0))
+ uc = 0xFFFD;
+ encode1(p, uc);
}
+ }
+ p->flags &= ~TERMP_NBRWORD;
}
-
static void
-adjbuf(struct termp *p, size_t sz)
+adjbuf(struct termp_col *c, size_t sz)
{
+ if (c->maxcols == 0)
+ c->maxcols = 1024;
+ while (c->maxcols <= sz)
+ c->maxcols <<= 2;
+ c->buf = mandoc_reallocarray(c->buf, c->maxcols, sizeof(*c->buf));
+}
- if (0 == p->maxcols)
- p->maxcols = 1024;
- while (sz >= p->maxcols)
- p->maxcols <<= 2;
-
- p->buf = realloc(p->buf, p->maxcols);
- if (NULL == p->buf) {
- perror(NULL);
- exit((int)MANDOCLEVEL_SYSERR);
+static void
+bufferc(struct termp *p, char c)
+{
+ if (p->flags & TERMP_NOBUF) {
+ (*p->letter)(p, c);
+ return;
}
+ if (p->col + 1 >= p->tcol->maxcols)
+ adjbuf(p->tcol, p->col + 1);
+ if (p->tcol->lastcol <= p->col || (c != ' ' && c != ASCII_NBRSP))
+ p->tcol->buf[p->col] = c;
+ if (p->tcol->lastcol < ++p->col)
+ p->tcol->lastcol = p->col;
}
-
+/*
+ * See encode().
+ * Do this for a single (probably unicode) value.
+ * Does not check for non-decorated glyphs.
+ */
static void
-bufferc(struct termp *p, char c)
+encode1(struct termp *p, int c)
{
+ enum termfont f;
- if (p->col + 1 >= p->maxcols)
- adjbuf(p, p->col + 1);
+ if (p->flags & TERMP_NOBUF) {
+ (*p->letter)(p, c);
+ return;
+ }
- p->buf[(int)p->col++] = c;
-}
+ if (p->col + 7 >= p->tcol->maxcols)
+ adjbuf(p->tcol, p->col + 7);
+
+ f = (c == ASCII_HYPH || c > 127 || isgraph(c)) ?
+ p->fontq[p->fonti] : TERMFONT_NONE;
+ if (p->flags & TERMP_BACKBEFORE) {
+ if (p->tcol->buf[p->col - 1] == ' ' ||
+ p->tcol->buf[p->col - 1] == '\t')
+ p->col--;
+ else
+ p->tcol->buf[p->col++] = '\b';
+ p->flags &= ~TERMP_BACKBEFORE;
+ }
+ if (f == TERMFONT_UNDER || f == TERMFONT_BI) {
+ p->tcol->buf[p->col++] = '_';
+ p->tcol->buf[p->col++] = '\b';
+ }
+ if (f == TERMFONT_BOLD || f == TERMFONT_BI) {
+ if (c == ASCII_HYPH)
+ p->tcol->buf[p->col++] = '-';
+ else
+ p->tcol->buf[p->col++] = c;
+ p->tcol->buf[p->col++] = '\b';
+ }
+ if (p->tcol->lastcol <= p->col || (c != ' ' && c != ASCII_NBRSP))
+ p->tcol->buf[p->col] = c;
+ if (p->tcol->lastcol < ++p->col)
+ p->tcol->lastcol = p->col;
+ if (p->flags & TERMP_BACKAFTER) {
+ p->flags |= TERMP_BACKBEFORE;
+ p->flags &= ~TERMP_BACKAFTER;
+ }
+}
static void
encode(struct termp *p, const char *word, size_t sz)
{
- enum termfont f;
- int i;
-
- /*
- * Encode and buffer a string of characters. If the current
- * font mode is unset, buffer directly, else encode then buffer
- * character by character.
- */
+ size_t i;
- if (TERMFONT_NONE == (f = term_fonttop(p))) {
- if (p->col + sz >= p->maxcols)
- adjbuf(p, p->col + sz);
- memcpy(&p->buf[(int)p->col], word, sz);
- p->col += sz;
+ if (p->flags & TERMP_NOBUF) {
+ for (i = 0; i < sz; i++)
+ (*p->letter)(p, word[i]);
return;
}
- /* Pre-buffer, assuming worst-case. */
+ if (p->col + 2 + (sz * 5) >= p->tcol->maxcols)
+ adjbuf(p->tcol, p->col + 2 + (sz * 5));
- if (p->col + 1 + (sz * 3) >= p->maxcols)
- adjbuf(p, p->col + 1 + (sz * 3));
+ for (i = 0; i < sz; i++) {
+ if (ASCII_HYPH == word[i] ||
+ isgraph((unsigned char)word[i]))
+ encode1(p, word[i]);
+ else {
+ if (p->tcol->lastcol <= p->col ||
+ (word[i] != ' ' && word[i] != ASCII_NBRSP))
+ p->tcol->buf[p->col] = word[i];
+ p->col++;
- for (i = 0; i < (int)sz; i++) {
- if ( ! isgraph((u_char)word[i])) {
- p->buf[(int)p->col++] = word[i];
- continue;
+ /*
+ * Postpone the effect of \z while handling
+ * an overstrike sequence from ascii_uc2str().
+ */
+
+ if (word[i] == '\b' &&
+ (p->flags & TERMP_BACKBEFORE)) {
+ p->flags &= ~TERMP_BACKBEFORE;
+ p->flags |= TERMP_BACKAFTER;
+ }
}
+ }
+ if (p->tcol->lastcol < p->col)
+ p->tcol->lastcol = p->col;
+}
- if (TERMFONT_UNDER == f)
- p->buf[(int)p->col++] = '_';
+void
+term_setwidth(struct termp *p, const char *wstr)
+{
+ struct roffsu su;
+ int iop, width;
+
+ iop = 0;
+ width = 0;
+ if (NULL != wstr) {
+ switch (*wstr) {
+ case '+':
+ iop = 1;
+ wstr++;
+ break;
+ case '-':
+ iop = -1;
+ wstr++;
+ break;
+ default:
+ break;
+ }
+ if (a2roffsu(wstr, &su, SCALE_MAX) != NULL)
+ width = term_hspan(p, &su);
else
- p->buf[(int)p->col++] = word[i];
-
- p->buf[(int)p->col++] = 8;
- p->buf[(int)p->col++] = word[i];
+ iop = 0;
}
+ (*p->setwidth)(p, iop, width);
}
-
size_t
term_len(const struct termp *p, size_t sz)
{
- return((*p->width)(p, ' ') * sz);
+ return (*p->width)(p, ' ') * sz;
}
+static size_t
+cond_width(const struct termp *p, int c, int *skip)
+{
+
+ if (*skip) {
+ (*skip) = 0;
+ return 0;
+ } else
+ return (*p->width)(p, c);
+}
size_t
term_strlen(const struct termp *p, const char *cp)
{
- size_t sz, ssz, rsz, i;
- enum roffdeco d;
+ size_t sz, rsz, i;
+ int ssz, skip, uc;
const char *seq, *rhs;
+ enum mandoc_esc esc;
+ static const char rej[] = { '\\', ASCII_NBRSP, ASCII_HYPH,
+ ASCII_BREAK, '\0' };
- for (sz = 0; '\0' != *cp; )
- /*
- * Account for escaped sequences within string length
- * calculations. This follows the logic in term_word()
- * as we must calculate the width of produced strings.
- */
- if ('\\' == *cp) {
- seq = ++cp;
- cp += a2roffdeco(&d, &seq, &ssz);
-
- switch (d) {
- case (DECO_RESERVED):
- rhs = chars_res2str
- (p->symtab, seq, ssz, &rsz);
- break;
- case (DECO_SPECIAL):
- /* FALLTHROUGH */
- case (DECO_SSPECIAL):
- rhs = chars_spec2str
- (p->symtab, seq, ssz, &rsz);
-
- /* Allow for one-char escapes. */
- if (DECO_SSPECIAL != d || rhs)
- break;
+ /*
+ * Account for escaped sequences within string length
+ * calculations. This follows the logic in term_word() as we
+ * must calculate the width of produced strings.
+ */
+
+ sz = 0;
+ skip = 0;
+ while ('\0' != *cp) {
+ rsz = strcspn(cp, rej);
+ for (i = 0; i < rsz; i++)
+ sz += cond_width(p, *cp++, &skip);
+
+ switch (*cp) {
+ case '\\':
+ cp++;
+ esc = mandoc_escape(&cp, &seq, &ssz);
+ if (ESCAPE_ERROR == esc)
+ continue;
- rhs = seq;
- rsz = ssz;
+ rhs = NULL;
+
+ switch (esc) {
+ case ESCAPE_UNICODE:
+ uc = mchars_num2uc(seq + 1, ssz - 1);
+ break;
+ case ESCAPE_NUMBERED:
+ uc = mchars_num2char(seq, ssz);
+ if (uc < 0)
+ continue;
break;
+ case ESCAPE_SPECIAL:
+ if (p->enc == TERMENC_ASCII) {
+ rhs = mchars_spec2str(seq, ssz, &rsz);
+ if (rhs != NULL)
+ break;
+ } else {
+ uc = mchars_spec2cp(seq, ssz);
+ if (uc > 0)
+ sz += cond_width(p, uc, &skip);
+ }
+ continue;
+ case ESCAPE_SKIPCHAR:
+ skip = 1;
+ continue;
+ case ESCAPE_OVERSTRIKE:
+ rsz = 0;
+ rhs = seq + ssz;
+ while (seq < rhs) {
+ if (*seq == '\\') {
+ mandoc_escape(&seq, NULL, NULL);
+ continue;
+ }
+ i = (*p->width)(p, *seq++);
+ if (rsz < i)
+ rsz = i;
+ }
+ sz += rsz;
+ continue;
default:
- rhs = NULL;
+ continue;
+ }
+
+ /*
+ * Common handling for Unicode and numbered
+ * character escape sequences.
+ */
+
+ if (rhs == NULL) {
+ if (p->enc == TERMENC_ASCII) {
+ rhs = ascii_uc2str(uc);
+ rsz = strlen(rhs);
+ } else {
+ if ((uc < 0x20 && uc != 0x09) ||
+ (uc > 0x7E && uc < 0xA0))
+ uc = 0xFFFD;
+ sz += cond_width(p, uc, &skip);
+ continue;
+ }
+ }
+
+ if (skip) {
+ skip = 0;
break;
}
- if (rhs)
- for (i = 0; i < rsz; i++)
- sz += (*p->width)(p, *rhs++);
- } else if (ASCII_NBRSP == *cp) {
- sz += (*p->width)(p, ' ');
+ /*
+ * Common handling for all escape sequences
+ * printing more than one character.
+ */
+
+ for (i = 0; i < rsz; i++)
+ sz += (*p->width)(p, *rhs++);
+ break;
+ case ASCII_NBRSP:
+ sz += cond_width(p, ' ', &skip);
cp++;
- } else if (ASCII_HYPH == *cp) {
- sz += (*p->width)(p, '-');
+ break;
+ case ASCII_HYPH:
+ sz += cond_width(p, '-', &skip);
cp++;
- } else
- sz += (*p->width)(p, *cp++);
+ break;
+ default:
+ break;
+ }
+ }
- return(sz);
+ return sz;
}
-
-/* ARGSUSED */
-size_t
+int
term_vspan(const struct termp *p, const struct roffsu *su)
{
double r;
+ int ri;
switch (su->unit) {
- case (SCALE_CM):
- r = su->scale * 2;
+ case SCALE_BU:
+ r = su->scale / 40.0;
break;
- case (SCALE_IN):
- r = su->scale * 6;
+ case SCALE_CM:
+ r = su->scale * 6.0 / 2.54;
break;
- case (SCALE_PC):
+ case SCALE_FS:
+ r = su->scale * 65536.0 / 40.0;
+ break;
+ case SCALE_IN:
+ r = su->scale * 6.0;
+ break;
+ case SCALE_MM:
+ r = su->scale * 0.006;
+ break;
+ case SCALE_PC:
r = su->scale;
break;
- case (SCALE_PT):
- r = su->scale / 8;
+ case SCALE_PT:
+ r = su->scale / 12.0;
break;
- case (SCALE_MM):
- r = su->scale / 1000;
+ case SCALE_EN:
+ case SCALE_EM:
+ r = su->scale * 0.6;
break;
- case (SCALE_VS):
+ case SCALE_VS:
r = su->scale;
break;
default:
- r = su->scale - 1;
- break;
+ abort();
}
-
- if (r < 0.0)
- r = 0.0;
- return(/* LINTED */(size_t)
- r);
+ ri = r > 0.0 ? r + 0.4995 : r - 0.4995;
+ return ri < 66 ? ri : 1;
}
-
-size_t
+/*
+ * Convert a scaling width to basic units, rounding towards 0.
+ */
+int
term_hspan(const struct termp *p, const struct roffsu *su)
{
- double v;
- v = ((*p->hspan)(p, su));
- if (v < 0.0)
- v = 0.0;
- return((size_t) /* LINTED */
- v);
+ return (*p->hspan)(p, su);
+}
+
+/*
+ * Convert a scaling width to basic units, rounding to closest.
+ */
+int
+term_hen(const struct termp *p, const struct roffsu *su)
+{
+ int bu;
+
+ if ((bu = (*p->hspan)(p, su)) >= 0)
+ return (bu + 11) / 24;
+ else
+ return -((-bu + 11) / 24);
}
git.ckatri.com / mandoc.git / blobdiff