-/* $Id: term.c,v 1.123 2009/11/06 10:31:32 kristaps Exp $ */
+/* $Id: term.c,v 1.253 2015/10/12 00:08:16 schwarze Exp $ */
/*
- * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
+ * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
+ * Copyright (c) 2010-2015 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.
*/
+#include "config.h"
+
+#include <sys/types.h>
+
#include <assert.h>
#include <ctype.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
-#include <time.h>
-#include "chars.h"
+#include "mandoc.h"
+#include "mandoc_aux.h"
#include "out.h"
#include "term.h"
-#include "man.h"
-#include "mdoc.h"
#include "main.h"
-/* FIXME: accomodate non-breaking, non-collapsing white-space. */
-/* FIXME: accomodate non-breaking, collapsing white-space. */
-
-static struct termp *term_alloc(enum termenc);
-static void term_free(struct termp *);
-
-static void do_escaped(struct termp *, const char **);
-static void do_special(struct termp *,
- const char *, size_t);
-static void do_reserved(struct termp *,
- const char *, size_t);
-static void buffer(struct termp *, char);
-static void encode(struct termp *, char);
+static size_t cond_width(const struct termp *, int, int *);
+static void adjbuf(struct termp *p, size_t);
+static void bufferc(struct termp *, char);
+static void encode(struct termp *, const char *, size_t);
+static void encode1(struct termp *, int);
-void *
-ascii_alloc(void)
+void
+term_free(struct termp *p)
{
- return(term_alloc(TERMENC_ASCII));
+ free(p->buf);
+ free(p->fontq);
+ free(p);
}
-
void
-terminal_free(void *arg)
+term_begin(struct termp *p, term_margin head,
+ term_margin foot, const struct roff_meta *arg)
{
- term_free((struct termp *)arg);
+ p->headf = head;
+ p->footf = foot;
+ p->argf = arg;
+ (*p->begin)(p);
}
-
-static void
-term_free(struct termp *p)
-{
-
- if (p->buf)
- free(p->buf);
- if (p->symtab)
- chars_free(p->symtab);
-
- free(p);
-}
-
-
-static struct termp *
-term_alloc(enum termenc enc)
+void
+term_end(struct termp *p)
{
- struct termp *p;
- p = calloc(1, sizeof(struct termp));
- if (NULL == p) {
- perror(NULL);
- exit(EXIT_FAILURE);
- }
- p->maxrmargin = 78;
- p->enc = enc;
- return(p);
+ (*p->end)(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.
- *
- * Specifically, a line is whatever's in p->buf of length p->col, which
- * is zeroed after this function returns.
- *
- * The usage of termp:flags is as follows:
- *
- * - 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.
+ * 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_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_DANGLE: 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_HANG: Like TERMP_DANGLE, and also suppress padding before
+ * the next chunk if this column is not full.
*/
void
term_flushln(struct termp *p)
{
- int i; /* current input position in p->buf */
+ size_t i; /* current input position in p->buf */
+ int ntab; /* number of tabs to prepend */
size_t vis; /* current visual position on output */
size_t vbl; /* number of blanks to prepend to output */
- size_t vsz; /* visual characters to write to output */
+ size_t vend; /* end of word visual position on output */
size_t bp; /* visual right border position */
- int j; /* temporary loop index */
- size_t maxvis, mmax;
- static int overstep = 0;
+ size_t dv; /* temporary for visual pos calculations */
+ size_t j; /* temporary loop index for p->buf */
+ size_t jhy; /* last hyph before overflow w/r/t j */
+ size_t maxvis; /* output position of visible boundary */
/*
* 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.
+ * small set of values.
+ *
+ * The following unsigned-signed subtractions look strange,
+ * but they are actually correct. If the int p->overstep
+ * is negative, it gets sign extended. Subtracting that
+ * very large size_t effectively adds a small number to dv.
*/
+ dv = p->rmargin > p->offset ? p->rmargin - p->offset : 0;
+ maxvis = (int)dv > p->overstep ? dv - (size_t)p->overstep : 0;
- assert(p->offset < p->rmargin);
-
- maxvis = (int)(p->rmargin - p->offset) - overstep < 0 ?
- /* LINTED */
- 0 : p->rmargin - p->offset - overstep;
- mmax = (int)(p->maxrmargin - p->offset) - overstep < 0 ?
- /* LINTED */
- 0 : p->maxrmargin - p->offset - overstep;
-
- bp = TERMP_NOBREAK & p->flags ? mmax : maxvis;
-
- /*
- * FIXME: if bp is zero, we still output the first word before
- * breaking the line.
- */
-
- vis = 0;
+ if (p->flags & TERMP_NOBREAK) {
+ dv = p->maxrmargin > p->offset ?
+ p->maxrmargin - p->offset : 0;
+ bp = (int)dv > p->overstep ?
+ dv - (size_t)p->overstep : 0;
+ } else
+ bp = maxvis;
/*
- * If in the standard case (left-justified), then begin with our
- * indentation, otherwise (columns, etc.) just start spitting
- * out text.
+ * Calculate the required amount of padding.
*/
+ vbl = p->offset + p->overstep > p->viscol ?
+ p->offset + p->overstep - p->viscol : 0;
+
+ vis = vend = 0;
+ i = 0;
- if ( ! (p->flags & TERMP_NOLPAD))
- /* LINTED */
- for (j = 0; j < (int)p->offset; j++)
- putchar(' ');
+ while (i < p->col) {
+ /*
+ * Handle literal tab characters: collapse all
+ * subsequent tabs into a single huge set of spaces.
+ */
+ ntab = 0;
+ while (i < p->col && '\t' == p->buf[i]) {
+ vend = (vis / p->tabwidth + 1) * p->tabwidth;
+ vbl += vend - vis;
+ vis = vend;
+ ntab++;
+ i++;
+ }
- for (i = 0; i < (int)p->col; i++) {
/*
* Count up visible word characters. Control sequences
* (starting with the CSI) aren't counted. A space
* space is printed according to regular spacing rules).
*/
- /* LINTED */
- for (j = i, vsz = 0; j < (int)p->col; j++) {
- if (j && ' ' == p->buf[j])
+ for (j = i, jhy = 0; j < p->col; j++) {
+ if (' ' == p->buf[j] || '\t' == p->buf[j])
break;
- else if (8 == p->buf[j])
- vsz--;
- else
- vsz++;
- }
- /*
- * Choose the number of blanks to prepend: no blank at the
- * beginning of a line, one between words -- but do not
- * actually write them yet.
- */
- vbl = (size_t)(0 == vis ? 0 : 1);
+ /* Back over the the last printed character. */
+ if (8 == p->buf[j]) {
+ assert(j);
+ vend -= (*p->width)(p, p->buf[j - 1]);
+ continue;
+ }
+
+ /* Regular word. */
+ /* Break at the hyphen point if we overrun. */
+ if (vend > vis && vend < bp &&
+ (ASCII_HYPH == p->buf[j] ||
+ ASCII_BREAK == p->buf[j]))
+ jhy = j;
+
+ /*
+ * Hyphenation now decided, put back a real
+ * hyphen such that we get the correct width.
+ */
+ if (ASCII_HYPH == p->buf[j])
+ p->buf[j] = '-';
+
+ vend += (*p->width)(p, p->buf[j]);
+ }
/*
* Find out whether we would exceed the right margin.
- * If so, break to the next line. (TODO: hyphenate)
- * Otherwise, write the chosen number of blanks now.
+ * If so, break to the next line.
*/
- if (vis && vis + vbl + vsz > bp) {
- putchar('\n');
- if (TERMP_NOBREAK & p->flags) {
- for (j = 0; j < (int)p->rmargin; j++)
- putchar(' ');
- vis = p->rmargin - p->offset;
- } else {
- for (j = 0; j < (int)p->offset; j++)
- putchar(' ');
- vis = 0;
- }
- /* Remove the overstep width. */
- bp += (int)/* LINTED */
- overstep;
- overstep = 0;
- } else {
- for (j = 0; j < (int)vbl; j++)
- putchar(' ');
- vis += vbl;
+ if (vend > bp && 0 == jhy && vis > 0) {
+ vend -= vis;
+ (*p->endline)(p);
+ p->viscol = 0;
+ if (TERMP_BRIND & p->flags) {
+ vbl = p->rmargin;
+ vend += p->rmargin;
+ vend -= p->offset;
+ } else
+ vbl = p->offset;
+
+ /* use pending tabs on the new line */
+
+ if (0 < ntab)
+ vbl += ntab * p->tabwidth;
+
+ /*
+ * Remove the p->overstep width.
+ * Again, if p->overstep is negative,
+ * sign extension does the right thing.
+ */
+
+ bp += (size_t)p->overstep;
+ p->overstep = 0;
}
- /*
- * Finally, write out the word.
- */
- for ( ; i < (int)p->col; i++) {
- if (' ' == p->buf[i])
+ /* Write out the [remaining] word. */
+ for ( ; i < p->col; i++) {
+ if (vend > bp && jhy > 0 && i > jhy)
break;
+ if ('\t' == p->buf[i])
+ break;
+ if (' ' == p->buf[i]) {
+ j = i;
+ while (i < p->col && ' ' == p->buf[i])
+ i++;
+ dv = (i - j) * (*p->width)(p, ' ');
+ vbl += dv;
+ vend += dv;
+ break;
+ }
+ if (ASCII_NBRSP == p->buf[i]) {
+ vbl += (*p->width)(p, ' ');
+ continue;
+ }
+ if (ASCII_BREAK == p->buf[i])
+ continue;
+
+ /*
+ * Now we definitely know there will be
+ * printable characters to output,
+ * so write preceding white space now.
+ */
+ if (vbl) {
+ (*p->advance)(p, vbl);
+ p->viscol += vbl;
+ vbl = 0;
+ }
- /* The unit sep. is a non-breaking space. */
- if (31 == p->buf[i])
- putchar(' ');
+ (*p->letter)(p, p->buf[i]);
+ if (8 == p->buf[i])
+ p->viscol -= (*p->width)(p, p->buf[i-1]);
else
- putchar(p->buf[i]);
+ p->viscol += (*p->width)(p, p->buf[i]);
}
- vis += vsz;
+ vis = vend;
}
+ /*
+ * If there was trailing white space, it was not printed;
+ * so reset the cursor position accordingly.
+ */
+ if (vis > vbl)
+ vis -= vbl;
+ else
+ vis = 0;
+
p->col = 0;
- overstep = 0;
+ p->overstep = 0;
+ p->flags &= ~(TERMP_BACKAFTER | TERMP_BACKBEFORE);
if ( ! (TERMP_NOBREAK & p->flags)) {
- putchar('\n');
+ p->viscol = 0;
+ (*p->endline)(p);
return;
}
if (TERMP_HANG & p->flags) {
- /* We need one blank after the tag. */
- overstep = /* LINTED */
- vis - maxvis + 1;
+ p->overstep += (int)(p->offset + vis - p->rmargin +
+ p->trailspace * (*p->width)(p, ' '));
/*
- * 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 there is a request to keep the columns together,
+ * allow negative overstep when the column is not full.
*/
- if (overstep >= -1) {
- assert((int)maxvis + overstep >= 0);
- /* LINTED */
- maxvis += overstep;
- } else
- overstep = 0;
+ if (p->trailspace && p->overstep < 0)
+ p->overstep = 0;
+ return;
} else if (TERMP_DANGLE & p->flags)
return;
- /* Right-pad. */
- if (maxvis > vis + /* LINTED */
- ((TERMP_TWOSPACE & p->flags) ? 1 : 0))
- for ( ; vis < maxvis; vis++)
- putchar(' ');
- else { /* ...or newline break. */
- putchar('\n');
- for (i = 0; i < (int)p->rmargin; i++)
- putchar(' ');
+ /* Trailing whitespace is significant in some columns. */
+ if (vis && vbl && (TERMP_BRTRSP & p->flags))
+ vis += vbl;
+
+ /* If the column was overrun, break the line. */
+ if (maxvis < vis + p->trailspace * (*p->width)(p, ' ')) {
+ (*p->endline)(p);
+ p->viscol = 0;
}
}
-
-/*
+/*
* 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) {
- p->flags &= ~TERMP_NOLPAD;
- return;
- }
- term_flushln(p);
- p->flags &= ~TERMP_NOLPAD;
+ if (p->col || 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);
- putchar('\n');
+ p->viscol = 0;
+ if (0 < p->skipvsp)
+ p->skipvsp--;
+ else
+ (*p->endline)(p);
}
-
-static void
-do_special(struct termp *p, const char *word, size_t len)
+/* Swap current and previous font; for \fP and .ft P */
+void
+term_fontlast(struct termp *p)
{
- const char *rhs;
- size_t sz;
- int i;
-
- rhs = chars_a2ascii(p->symtab, word, len, &sz);
-
- if (NULL == rhs) {
-#if 0
- fputs("Unknown special character: ", stderr);
- for (i = 0; i < (int)len; i++)
- fputc(word[i], stderr);
- fputc('\n', stderr);
-#endif
- return;
- }
- for (i = 0; i < (int)sz; i++)
- encode(p, rhs[i]);
+ enum termfont f;
+
+ f = p->fontl;
+ p->fontl = p->fontq[p->fonti];
+ 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->fontl = p->fontq[p->fonti];
+ p->fontq[p->fonti] = f;
+}
-static void
-do_reserved(struct termp *p, const char *word, size_t len)
+/* Set font, save previous. */
+void
+term_fontpush(struct termp *p, enum termfont f)
{
- const char *rhs;
- size_t sz;
- int i;
-
- rhs = chars_a2res(p->symtab, word, len, &sz);
-
- if (NULL == rhs) {
-#if 0
- fputs("Unknown reserved word: ", stderr);
- for (i = 0; i < (int)len; i++)
- fputc(word[i], stderr);
- fputc('\n', stderr);
-#endif
- return;
+
+ p->fontl = p->fontq[p->fonti];
+ if (++p->fonti == p->fontsz) {
+ p->fontsz += 8;
+ p->fontq = mandoc_reallocarray(p->fontq,
+ p->fontsz, sizeof(enum termfont *));
}
- for (i = 0; i < (int)sz; i++)
- encode(p, rhs[i]);
+ p->fontq[p->fonti] = f;
}
-
-/*
- * Handle an escape sequence: determine its length and pass it to the
- * escape-symbol look table. Note that we assume mdoc(3) has validated
- * the escape sequence (we assert upon badly-formed escape sequences).
- */
-static void
-do_escaped(struct termp *p, const char **word)
+/* Flush to make the saved pointer current again. */
+void
+term_fontpopq(struct termp *p, int i)
{
- int j, type, sv, t, lim;
- const char *wp;
- wp = *word;
- type = 1;
+ assert(i >= 0);
+ if (p->fonti > i)
+ p->fonti = i;
+}
- if ('\0' == *(++wp)) {
- *word = wp;
- return;
- }
+/* Pop one font off the stack. */
+void
+term_fontpop(struct termp *p)
+{
- if ('(' == *wp) {
- wp++;
- if ('\0' == *wp || '\0' == *(wp + 1)) {
- *word = '\0' == *wp ? wp : wp + 1;
- return;
- }
+ assert(p->fonti);
+ p->fonti--;
+}
- do_special(p, wp, 2);
- *word = ++wp;
- return;
+/*
+ * 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
+ * handles word styling.
+ */
+void
+term_word(struct termp *p, const char *word)
+{
+ const char nbrsp[2] = { ASCII_NBRSP, 0 };
+ const char *seq, *cp;
+ int sz, uc;
+ size_t ssz;
+ enum mandoc_esc esc;
+
+ if ( ! (TERMP_NOSPACE & p->flags)) {
+ if ( ! (TERMP_KEEP & p->flags)) {
+ bufferc(p, ' ');
+ if (TERMP_SENTENCE & p->flags)
+ bufferc(p, ' ');
+ } else
+ bufferc(p, ASCII_NBRSP);
+ }
+ if (TERMP_PREKEEP & p->flags)
+ p->flags |= TERMP_KEEP;
- } else if ('*' == *wp) {
- if ('\0' == *(++wp)) {
- *word = wp;
- return;
+ if ( ! (p->flags & TERMP_NONOSPACE))
+ p->flags &= ~TERMP_NOSPACE;
+ else
+ p->flags |= TERMP_NOSPACE;
+
+ p->flags &= ~(TERMP_SENTENCE | TERMP_NONEWLINE);
+ p->skipvsp = 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;
+ continue;
}
- switch (*wp) {
- case ('('):
- wp++;
- if ('\0' == *wp || '\0' == *(wp + 1)) {
- *word = '\0' == *wp ? wp : wp + 1;
- return;
- }
+ word++;
+ esc = mandoc_escape(&word, &seq, &sz);
+ if (ESCAPE_ERROR == esc)
+ continue;
- do_reserved(p, wp, 2);
- *word = ++wp;
- return;
- case ('['):
- type = 0;
+ switch (esc) {
+ case ESCAPE_UNICODE:
+ uc = mchars_num2uc(seq + 1, sz - 1);
+ break;
+ case ESCAPE_NUMBERED:
+ uc = mchars_num2char(seq, sz);
+ if (uc < 0)
+ continue;
break;
+ case ESCAPE_SPECIAL:
+ if (p->enc == TERMENC_ASCII) {
+ cp = mchars_spec2str(p->symtab,
+ seq, sz, &ssz);
+ if (cp != NULL)
+ encode(p, cp, ssz);
+ } else {
+ uc = mchars_spec2cp(p->symtab, seq, sz);
+ if (uc > 0)
+ encode1(p, uc);
+ }
+ continue;
+ case ESCAPE_FONTBOLD:
+ term_fontrepl(p, TERMFONT_BOLD);
+ continue;
+ case ESCAPE_FONTITALIC:
+ term_fontrepl(p, TERMFONT_UNDER);
+ continue;
+ case ESCAPE_FONTBI:
+ term_fontrepl(p, TERMFONT_BI);
+ continue;
+ case ESCAPE_FONT:
+ case ESCAPE_FONTROMAN:
+ term_fontrepl(p, TERMFONT_NONE);
+ continue;
+ case ESCAPE_FONTPREV:
+ term_fontlast(p);
+ 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_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->col > 2 && p->buf[p->col - 1] == ' ')
+ p->col -= 2;
+ continue;
default:
- do_reserved(p, wp, 1);
- *word = wp;
- return;
+ continue;
}
- } else if ('s' == *wp) {
- /* This closely follows mandoc_special(). */
- if ('\0' == *(++wp)) {
- *word = wp;
- return;
- }
+ /*
+ * Common handling for Unicode and numbered
+ * character escape sequences.
+ */
- t = 0;
- lim = 1;
-
- if (*wp == '\'') {
- lim = 0;
- t = 1;
- ++wp;
- } else if (*wp == '[') {
- lim = 0;
- t = 2;
- ++wp;
- } else if (*wp == '(') {
- lim = 2;
- t = 3;
- ++wp;
+ 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;
+}
- if (*wp == '+' || *wp == '-')
- ++wp;
+static void
+adjbuf(struct termp *p, size_t sz)
+{
- if (*wp == '\'') {
- if (t) {
- *word = wp;
- return;
- }
- lim = 0;
- t = 1;
- ++wp;
- } else if (*wp == '[') {
- if (t) {
- *word = wp;
- return;
- }
- lim = 0;
- t = 2;
- ++wp;
- } else if (*wp == '(') {
- if (t) {
- *word = wp;
- return;
- }
- lim = 2;
- t = 3;
- ++wp;
- }
+ if (0 == p->maxcols)
+ p->maxcols = 1024;
+ while (sz >= p->maxcols)
+ p->maxcols <<= 2;
- if ( ! isdigit((u_char)*wp)) {
- *word = --wp;
- return;
- }
+ p->buf = mandoc_reallocarray(p->buf, p->maxcols, sizeof(int));
+}
- for (j = 0; isdigit((u_char)*wp); j++) {
- if (lim && j >= lim)
- break;
- ++wp;
- }
+static void
+bufferc(struct termp *p, char c)
+{
- if (t && t < 3) {
- if (1 == t && *wp != '\'') {
- *word = --wp;
- return;
- }
- if (2 == t && *wp != ']') {
- *word = --wp;
- return;
- }
- ++wp;
- }
- *word = --wp;
- return;
+ if (p->col + 1 >= p->maxcols)
+ adjbuf(p, p->col + 1);
- } else if ('f' == *wp) {
- if ('\0' == *(++wp)) {
- *word = wp;
- return;
- }
+ p->buf[p->col++] = c;
+}
- switch (*wp) {
- case ('3'):
- /* FALLTHROUGH */
- case ('B'):
- p->metamask = p->metafont;
- p->metafont |= METAF_BOLD;
- break;
- case ('2'):
- /* FALLTHROUGH */
- case ('I'):
- p->metamask = p->metafont;
- p->metafont |= METAF_UNDER;
- break;
- case ('P'):
- sv = p->metamask;
- p->metamask = p->metafont;
- p->metafont = sv;
- break;
- case ('1'):
- /* FALLTHROUGH */
- case ('R'):
- p->metamask = p->metafont;
- p->metafont &= ~METAF_UNDER;
- p->metafont &= ~METAF_BOLD;
- break;
- default:
- break;
- }
+/*
+ * See encode().
+ * Do this for a single (probably unicode) value.
+ * Does not check for non-decorated glyphs.
+ */
+static void
+encode1(struct termp *p, int c)
+{
+ enum termfont f;
- *word = wp;
- return;
+ if (p->col + 7 >= p->maxcols)
+ adjbuf(p, p->col + 7);
- } else if ('[' != *wp) {
- do_special(p, wp, 1);
- *word = wp;
- return;
+ f = (c == ASCII_HYPH || isgraph(c)) ?
+ p->fontq[p->fonti] : TERMFONT_NONE;
+
+ if (p->flags & TERMP_BACKBEFORE) {
+ if (p->buf[p->col - 1] == ' ')
+ p->col--;
+ else
+ p->buf[p->col++] = 8;
+ p->flags &= ~TERMP_BACKBEFORE;
+ }
+ if (TERMFONT_UNDER == f || TERMFONT_BI == f) {
+ p->buf[p->col++] = '_';
+ p->buf[p->col++] = 8;
+ }
+ if (TERMFONT_BOLD == f || TERMFONT_BI == f) {
+ if (ASCII_HYPH == c)
+ p->buf[p->col++] = '-';
+ else
+ p->buf[p->col++] = c;
+ p->buf[p->col++] = 8;
}
+ p->buf[p->col++] = c;
+ if (p->flags & TERMP_BACKAFTER) {
+ p->flags |= TERMP_BACKBEFORE;
+ p->flags &= ~TERMP_BACKAFTER;
+ }
+}
- wp++;
- for (j = 0; *wp && ']' != *wp; wp++, j++)
- /* Loop... */ ;
+static void
+encode(struct termp *p, const char *word, size_t sz)
+{
+ size_t i;
- if ('\0' == *wp) {
- *word = wp;
- return;
- }
+ if (p->col + 2 + (sz * 5) >= p->maxcols)
+ adjbuf(p, p->col + 2 + (sz * 5));
- if (type)
- do_special(p, wp - j, (size_t)j);
- else
- do_reserved(p, wp - j, (size_t)j);
- *word = wp;
+ for (i = 0; i < sz; i++) {
+ if (ASCII_HYPH == word[i] ||
+ isgraph((unsigned char)word[i]))
+ encode1(p, word[i]);
+ else
+ p->buf[p->col++] = word[i];
+ }
}
-
-/*
- * 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
- * handles word styling.
- */
void
-term_word(struct termp *p, const char *word)
+term_setwidth(struct termp *p, const char *wstr)
{
- const char *sv;
-
- 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(']'):
- /* FALLTHROUGH */
- case('}'):
- if ( ! (TERMP_IGNDELIM & p->flags))
- p->flags |= TERMP_NOSPACE;
+ struct roffsu su;
+ int iop, width;
+
+ iop = 0;
+ width = 0;
+ if (NULL != wstr) {
+ switch (*wstr) {
+ case '+':
+ iop = 1;
+ wstr++;
break;
- default:
- break;
- }
-
- if ( ! (TERMP_NOSPACE & p->flags))
- buffer(p, ' ');
-
- if ( ! (p->flags & TERMP_NONOSPACE))
- p->flags &= ~TERMP_NOSPACE;
-
- for ( ; *word; word++)
- if ('\\' != *word)
- encode(p, *word);
- else
- do_escaped(p, &word);
-
- if (sv[0] && 0 == sv[1])
- switch (sv[0]) {
- case('('):
- /* FALLTHROUGH */
- case('['):
- /* FALLTHROUGH */
- case('{'):
- p->flags |= TERMP_NOSPACE;
+ case '-':
+ iop = -1;
+ wstr++;
break;
default:
break;
}
+ if (a2roffsu(wstr, &su, SCALE_MAX))
+ width = term_hspan(p, &su);
+ else
+ iop = 0;
+ }
+ (*p->setwidth)(p, iop, width);
}
-
-/*
- * Insert a single character into the line-buffer. If the buffer's
- * space is exceeded, then allocate more space by doubling the buffer
- * size.
- */
-static void
-buffer(struct termp *p, char c)
+size_t
+term_len(const struct termp *p, size_t sz)
{
- size_t s;
-
- if (p->col + 1 >= p->maxcols) {
- if (0 == p->maxcols)
- p->maxcols = 256;
- s = p->maxcols * 2;
- p->buf = realloc(p->buf, s);
- if (NULL == p->buf) {
- perror(NULL);
- exit(EXIT_FAILURE);
- }
- p->maxcols = s;
- }
- p->buf[(int)(p->col)++] = c;
-}
+ return (*p->width)(p, ' ') * sz;
+}
-static void
-encode(struct termp *p, char c)
+static size_t
+cond_width(const struct termp *p, int c, int *skip)
{
-
- if (isgraph((u_char)c)) {
- if (p->under || METAF_UNDER & p->metafont) {
- buffer(p, '_');
- buffer(p, 8);
- }
- if (p->bold || METAF_BOLD & p->metafont) {
- buffer(p, c);
- buffer(p, 8);
- }
- }
- buffer(p, c);
-}
+ if (*skip) {
+ (*skip) = 0;
+ return 0;
+ } else
+ return (*p->width)(p, c);
+}
size_t
-term_vspan(const struct roffsu *su)
+term_strlen(const struct termp *p, const char *cp)
{
- double r;
+ 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' };
- switch (su->unit) {
- case (SCALE_CM):
- r = su->scale * 2;
- break;
- case (SCALE_IN):
- r = su->scale * 6;
- break;
- case (SCALE_PC):
- r = su->scale;
- break;
- case (SCALE_PT):
- r = su->scale / 8;
- break;
- case (SCALE_MM):
- r = su->scale / 1000;
- break;
- case (SCALE_VS):
- r = su->scale;
- break;
- default:
- r = su->scale - 1;
- 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 = 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(p->symtab,
+ seq, ssz, &rsz);
+ if (rhs != NULL)
+ break;
+ } else {
+ uc = mchars_spec2cp(p->symtab,
+ 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:
+ 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;
+ }
+
+ /*
+ * 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++;
+ break;
+ case ASCII_HYPH:
+ sz += cond_width(p, '-', &skip);
+ cp++;
+ break;
+ default:
+ break;
+ }
}
- if (r < 0.0)
- r = 0.0;
- return(/* LINTED */(size_t)
- r);
+ return sz;
}
-
-size_t
-term_hspan(const struct roffsu *su)
+int
+term_vspan(const struct termp *p, const struct roffsu *su)
{
double r;
-
- /* XXX: CM, IN, and PT are approximations. */
+ int ri;
switch (su->unit) {
- case (SCALE_CM):
- r = 4 * su->scale;
+ case SCALE_BU:
+ r = su->scale / 40.0;
+ break;
+ case SCALE_CM:
+ r = su->scale * 6.0 / 2.54;
break;
- case (SCALE_IN):
- /* XXX: this is an approximation. */
- r = 10 * su->scale;
+ case SCALE_FS:
+ r = su->scale * 65536.0 / 40.0;
break;
- case (SCALE_PC):
- r = (10 * su->scale) / 6;
+ case SCALE_IN:
+ r = su->scale * 6.0;
break;
- case (SCALE_PT):
- r = (10 * su->scale) / 72;
+ case SCALE_MM:
+ r = su->scale * 0.006;
break;
- case (SCALE_MM):
- r = su->scale / 1000; /* FIXME: double-check. */
+ case SCALE_PC:
+ r = su->scale;
break;
- case (SCALE_VS):
- r = su->scale * 2 - 1; /* FIXME: double-check. */
+ case SCALE_PT:
+ r = su->scale / 12.0;
break;
- default:
+ case SCALE_EN:
+ case SCALE_EM:
+ r = su->scale * 0.6;
+ break;
+ case SCALE_VS:
r = su->scale;
break;
+ default:
+ abort();
}
-
- if (r < 0.0)
- r = 0.0;
- return((size_t)/* LINTED */
- r);
+ ri = r > 0.0 ? r + 0.4995 : r - 0.4995;
+ return ri < 66 ? ri : 1;
}
+/*
+ * Convert a scaling width to basic units, rounding down.
+ */
+int
+term_hspan(const struct termp *p, const struct roffsu *su)
+{
+ return (*p->hspan)(p, su);
+}
git.ckatri.com / mandoc.git / blobdiff