=================================================================== RCS file: /cvs/mandoc/term.c,v retrieving revision 1.87 retrieving revision 1.207 diff -u -p -r1.87 -r1.207 --- mandoc/term.c 2009/07/15 15:53:57 1.87 +++ mandoc/term.c 2013/05/29 15:17:52 1.207 @@ -1,6 +1,7 @@ -/* $Id: term.c,v 1.87 2009/07/15 15:53:57 kristaps Exp $ */ +/* $Id: term.c,v 1.207 2013/05/29 15:17:52 schwarze Exp $ */ /* - * Copyright (c) 2008, 2009 Kristaps Dzonsons + * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons + * Copyright (c) 2010, 2011, 2012 Ingo Schwarze * * Permission to use, copy, modify, and distribute this software for any * purpose with or without fee is hereby granted, provided that the above @@ -14,181 +15,83 @@ * 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 + #include -#include +#include +#include #include #include #include +#include "mandoc.h" +#include "out.h" #include "term.h" -#include "man.h" -#include "mdoc.h" +#include "main.h" -extern int man_run(struct termp *, - const struct man *); -extern int mdoc_run(struct termp *, - const struct mdoc *); +static size_t cond_width(const struct termp *, int, int *); +static void adjbuf(struct termp *p, int); +static void bufferc(struct termp *, char); +static void encode(struct termp *, const char *, size_t); +static void encode1(struct termp *, int); -static struct termp *term_alloc(enum termenc); -static void term_free(struct termp *); -static void term_pescape(struct termp *, const char **); -static void term_nescape(struct termp *, - const char *, size_t); -static void term_chara(struct termp *, char); -static void term_encodea(struct termp *, char); -static int term_isopendelim(const char *); -static int term_isclosedelim(const char *); - - -void * -ascii_alloc(void) -{ - - return(term_alloc(TERMENC_ASCII)); -} - - -int -terminal_man(void *arg, const struct man *man) -{ - struct termp *p; - - p = (struct termp *)arg; - if (NULL == p->symtab) - p->symtab = term_ascii2htab(); - - return(man_run(p, man)); -} - - -int -terminal_mdoc(void *arg, const struct mdoc *mdoc) -{ - struct termp *p; - - p = (struct termp *)arg; - if (NULL == p->symtab) - p->symtab = term_ascii2htab(); - - return(mdoc_run(p, mdoc)); -} - - void -terminal_free(void *arg) -{ - - term_free((struct termp *)arg); -} - - -static void term_free(struct termp *p) { if (p->buf) free(p->buf); - if (TERMENC_ASCII == p->enc && p->symtab) - term_asciifree(p->symtab); + if (p->symtab) + mchars_free(p->symtab); free(p); } -static struct termp * -term_alloc(enum termenc enc) +void +term_begin(struct termp *p, term_margin head, + term_margin foot, const void *arg) { - struct termp *p; - if (NULL == (p = malloc(sizeof(struct termp)))) - err(1, "malloc"); - bzero(p, sizeof(struct termp)); - p->maxrmargin = 78; - p->enc = enc; - return(p); + p->headf = head; + p->footf = foot; + p->argf = arg; + (*p->begin)(p); } -static int -term_isclosedelim(const char *p) +void +term_end(struct termp *p) { - if ( ! (*p && 0 == *(p + 1))) - return(0); - - switch (*p) { - case('.'): - /* FALLTHROUGH */ - case(','): - /* FALLTHROUGH */ - case(';'): - /* FALLTHROUGH */ - case(':'): - /* FALLTHROUGH */ - case('?'): - /* FALLTHROUGH */ - case('!'): - /* FALLTHROUGH */ - case(')'): - /* FALLTHROUGH */ - case(']'): - /* FALLTHROUGH */ - case('}'): - return(1); - default: - break; - } - - return(0); + (*p->end)(p); } - -static int -term_isopendelim(const char *p) -{ - - if ( ! (*p && 0 == *(p + 1))) - return(0); - - switch (*p) { - case('('): - /* FALLTHROUGH */ - case('['): - /* FALLTHROUGH */ - case('{'): - return(1); - default: - break; - } - - return(0); -} - - /* * 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. + * fragment of a columnar list (`Bl -tag' or `Bl -column'), which does + * not have a trailing newline. * - * Specifically, a line is whatever's in p->buf of length p->col, which - * is zeroed after this function returns. + * The following flags may be specified: * - * 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. - * * - 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. + * columns. In short: don't print a newline and instead expect the + * next call to do the padding up to the start of the next column. * + * - TERMP_TWOSPACE: make sure there is room for 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 + * overrunning, instead save the position and continue at that point * when the next invocation. * * In-line line breaking: @@ -196,141 +99,197 @@ term_isopendelim(const char *p) * 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. Extremely long - * lines will cause the system to emit a warning (TODO: hyphenate, if - * possible). - * - * FIXME: newline breaks occur (in groff) also occur when a single - * space follows a NOBREAK! + * writing from the right-margin, which will lead to the above scenario + * upon exit. Otherwise, the line will break at the right margin. */ void term_flushln(struct termp *p) { - int i, j; - size_t vbl, vsz, vis, maxvis, mmax, bp; - static int sv = -1; + int 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 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 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. + * 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; - assert(p->offset < p->rmargin); - maxvis = p->rmargin - p->offset; - mmax = p->maxrmargin - p->offset; bp = TERMP_NOBREAK & p->flags ? mmax : maxvis; - vis = 0; - if (sv >= 0) { - vis = (size_t)sv; - sv = -1; - } - /* - * 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; - if ( ! (p->flags & TERMP_NOLPAD)) - /* LINTED */ - for (j = 0; j < (int)p->offset; j++) - putchar(' '); + vis = vend = 0; + i = 0; - for (i = 0; i < (int)p->col; i++) { + 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++; + } + + /* * Count up visible word characters. Control sequences * (starting with the CSI) aren't counted. A space * generates a non-printing word, which is valid (the * space is printed according to regular spacing rules). */ - /* LINTED */ - for (j = i, vsz = 0; j < (int)p->col; j++) { - if (' ' == p->buf[j]) + for (j = i, jhy = 0; j < p->col; j++) { + if ((j && ' ' == p->buf[j]) || '\t' == p->buf[j]) break; - else if (8 == p->buf[j]) - j += 1; - else - vsz++; + + /* 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]) + jhy = j; + + vend += (*p->width)(p, p->buf[j]); } /* - * 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); - - /* * 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 (vend > bp && 0 == jhy && vis > 0) { + vend -= vis; + (*p->endline)(p); + p->viscol = 0; 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; - } - } else { - for (j = 0; j < (int)vbl; j++) - putchar(' '); - vis += vbl; + vbl = p->rmargin; + vend += p->rmargin - 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. */ + + 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; - putchar(p->buf[i]); + if ('\t' == p->buf[i]) + break; + if (' ' == p->buf[i]) { + j = i; + while (' ' == p->buf[i]) + i++; + dv = (size_t)(i - j) * (*p->width)(p, ' '); + vbl += dv; + vend += dv; + break; + } + if (ASCII_NBRSP == p->buf[i]) { + vbl += (*p->width)(p, ' '); + 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; + } + + if (ASCII_HYPH == p->buf[i]) { + (*p->letter)(p, '-'); + p->viscol += (*p->width)(p, '-'); + continue; + } + + (*p->letter)(p, p->buf[i]); + if (8 == p->buf[i]) + p->viscol -= (*p->width)(p, p->buf[i-1]); + else + p->viscol += (*p->width)(p, p->buf[i]); } - vis += vsz; + vis = vend; } /* - * If we've overstepped our maximum visible no-break space, then - * cause a newline and offset at the right margin. + * If there was trailing white space, it was not printed; + * so reset the cursor position accordingly. */ + if (vis) + vis -= vbl; - if ((TERMP_NOBREAK & p->flags) && vis >= maxvis) { - if ( ! (TERMP_DANGLE & p->flags) && - ! (TERMP_HANG & p->flags)) { - putchar('\n'); - for (i = 0; i < (int)p->rmargin; i++) - putchar(' '); - } - if (TERMP_HANG & p->flags) - sv = (int)(vis - maxvis); - p->col = 0; + p->col = 0; + p->overstep = 0; + + if ( ! (TERMP_NOBREAK & p->flags)) { + p->viscol = 0; + (*p->endline)(p); return; } - /* - * If we're not to right-marginalise it (newline), then instead - * pad to the right margin and stay off. - */ + if (TERMP_HANG & p->flags) { + /* We need one blank after the tag. */ + p->overstep = (int)(vis - maxvis + (*p->width)(p, ' ')); - if (p->flags & TERMP_NOBREAK) { - if ( ! (TERMP_DANGLE & p->flags)) - for ( ; vis < maxvis; vis++) - putchar(' '); - } else - putchar('\n'); + /* + * If we have overstepped the margin, temporarily move + * it to the right and flag the rest of the line to be + * shorter. + */ + if (p->overstep < 0) + p->overstep = 0; + return; - p->col = 0; + } else if (TERMP_DANGLE & p->flags) + return; + + /* If the column was overrun, break the line. */ + if (maxvis <= vis + + ((TERMP_TWOSPACE & p->flags) ? (*p->width)(p, ' ') : 0)) { + (*p->endline)(p); + p->viscol = 0; + } } @@ -344,12 +303,8 @@ term_newln(struct termp *p) { 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); } @@ -364,200 +319,464 @@ term_vspace(struct termp *p) { term_newln(p); - putchar('\n'); + p->viscol = 0; + if (0 < p->skipvsp) + p->skipvsp--; + else + (*p->endline)(p); } +void +term_fontlast(struct termp *p) +{ + enum termfont f; -/* - * Determine the symbol indicated by an escape sequences, that is, one - * starting with a backslash. Once done, we pass this value into the - * output buffer by way of the symbol table. - */ -static void -term_nescape(struct termp *p, const char *word, size_t len) + f = p->fontl; + p->fontl = p->fontq[p->fonti]; + p->fontq[p->fonti] = f; +} + + +void +term_fontrepl(struct termp *p, enum termfont f) { - const char *rhs; - size_t sz; - int i; - rhs = term_a2ascii(p->symtab, word, len, &sz); + p->fontl = p->fontq[p->fonti]; + p->fontq[p->fonti] = f; +} - if (rhs) - for (i = 0; i < (int)sz; i++) - term_encodea(p, rhs[i]); + +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]); +} + + +void +term_fontpopq(struct termp *p, const void *key) +{ + + while (p->fonti >= 0 && key < (void *)(p->fontq + p->fonti)) + p->fonti--; + assert(p->fonti >= 0); +} + + +void +term_fontpop(struct termp *p) +{ + + assert(p->fonti); + p->fonti--; +} + /* - * 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). + * 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. */ -static void -term_pescape(struct termp *p, const char **word) +void +term_word(struct termp *p, const char *word) { - int j; - const char *wp; + const char *seq, *cp; + char c; + int sz, uc; + size_t ssz; + enum mandoc_esc esc; - wp = *word; - - if (0 == *(++wp)) { - *word = wp; - return; + 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; - if ('(' == *wp) { - wp++; - if (0 == *wp || 0 == *(wp + 1)) { - *word = 0 == *wp ? wp : wp + 1; - return; - } + if ( ! (p->flags & TERMP_NONOSPACE)) + p->flags &= ~TERMP_NOSPACE; + else + p->flags |= TERMP_NOSPACE; - term_nescape(p, wp, 2); - *word = ++wp; - return; + p->flags &= ~(TERMP_SENTENCE | TERMP_IGNDELIM); - } else if ('*' == *wp) { - if (0 == *(++wp)) { - *word = wp; - return; + while ('\0' != *word) { + if ('\\' != *word) { + if (TERMP_SKIPCHAR & p->flags) { + p->flags &= ~TERMP_SKIPCHAR; + word++; + continue; + } + 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) + break; + + if (TERMENC_ASCII != p->enc) + switch (esc) { + case (ESCAPE_UNICODE): + uc = mchars_num2uc(seq + 1, sz - 1); + if ('\0' == uc) + break; + encode1(p, uc); + continue; + case (ESCAPE_SPECIAL): + uc = mchars_spec2cp(p->symtab, seq, sz); + if (uc <= 0) + break; + encode1(p, uc); + continue; + default: + break; } - term_nescape(p, wp, 2); - *word = ++wp; - return; - case ('['): + switch (esc) { + case (ESCAPE_UNICODE): + encode1(p, '?'); break; - default: - term_nescape(p, wp, 1); - *word = wp; - return; - } - - } else if ('f' == *wp) { - if (0 == *(++wp)) { - *word = wp; - return; - } - - switch (*wp) { - case ('B'): - p->flags |= TERMP_BOLD; + case (ESCAPE_NUMBERED): + c = mchars_num2char(seq, sz); + if ('\0' != c) + encode(p, &c, 1); break; - case ('I'): - p->flags |= TERMP_UNDER; + case (ESCAPE_SPECIAL): + cp = mchars_spec2str(p->symtab, seq, sz, &ssz); + if (NULL != cp) + encode(p, cp, ssz); + else if (1 == ssz) + encode(p, seq, sz); break; - case ('P'): + case (ESCAPE_FONTBOLD): + term_fontrepl(p, TERMFONT_BOLD); + break; + case (ESCAPE_FONTITALIC): + term_fontrepl(p, TERMFONT_UNDER); + break; + case (ESCAPE_FONT): /* FALLTHROUGH */ - case ('R'): - p->flags &= ~TERMP_STYLE; + case (ESCAPE_FONTROMAN): + term_fontrepl(p, TERMFONT_NONE); break; + case (ESCAPE_FONTPREV): + term_fontlast(p); + break; + case (ESCAPE_NOSPACE): + if (TERMP_SKIPCHAR & p->flags) + p->flags &= ~TERMP_SKIPCHAR; + else if ('\0' == *word) + p->flags |= TERMP_NOSPACE; + break; + case (ESCAPE_SKIPCHAR): + p->flags |= TERMP_SKIPCHAR; + break; default: break; } - - *word = wp; - return; - - } else if ('[' != *wp) { - term_nescape(p, wp, 1); - *word = wp; - return; } +} - wp++; - for (j = 0; *wp && ']' != *wp; wp++, j++) - /* Loop... */ ; +static void +adjbuf(struct termp *p, int sz) +{ - if (0 == *wp) { - *word = wp; - return; - } + if (0 == p->maxcols) + p->maxcols = 1024; + while (sz >= p->maxcols) + p->maxcols <<= 2; - term_nescape(p, wp - j, (size_t)j); - *word = wp; + p->buf = mandoc_realloc + (p->buf, sizeof(int) * (size_t)p->maxcols); } +static void +bufferc(struct termp *p, char c) +{ + if (p->col + 1 >= p->maxcols) + adjbuf(p, p->col + 1); + + p->buf[p->col++] = c; +} + /* - * 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. + * See encode(). + * Do this for a single (probably unicode) value. + * Does not check for non-decorated glyphs. */ -void -term_word(struct termp *p, const char *word) +static void +encode1(struct termp *p, int c) { + enum termfont f; - if (term_isclosedelim(word)) - if ( ! (TERMP_IGNDELIM & p->flags)) - p->flags |= TERMP_NOSPACE; + if (TERMP_SKIPCHAR & p->flags) { + p->flags &= ~TERMP_SKIPCHAR; + return; + } - if ( ! (TERMP_NOSPACE & p->flags)) - term_chara(p, ' '); + if (p->col + 4 >= p->maxcols) + adjbuf(p, p->col + 4); - if ( ! (p->flags & TERMP_NONOSPACE)) - p->flags &= ~TERMP_NOSPACE; + f = term_fonttop(p); - /* - * If ANSI (word-length styling), then apply our style now, - * before the word. + if (TERMFONT_NONE == f) { + p->buf[p->col++] = c; + return; + } else if (TERMFONT_UNDER == f) { + p->buf[p->col++] = '_'; + } else + p->buf[p->col++] = c; + + p->buf[p->col++] = 8; + p->buf[p->col++] = c; +} + +static void +encode(struct termp *p, const char *word, size_t sz) +{ + enum termfont f; + int i, len; + + if (TERMP_SKIPCHAR & p->flags) { + p->flags &= ~TERMP_SKIPCHAR; + return; + } + + /* LINTED */ + len = sz; + + /* + * Encode and buffer a string of characters. If the current + * font mode is unset, buffer directly, else encode then buffer + * character by character. */ - for ( ; *word; word++) - if ('\\' != *word) - term_encodea(p, *word); + if (TERMFONT_NONE == (f = term_fonttop(p))) { + if (p->col + len >= p->maxcols) + adjbuf(p, p->col + len); + for (i = 0; i < len; i++) + p->buf[p->col++] = word[i]; + return; + } + + /* Pre-buffer, assuming worst-case. */ + + if (p->col + 1 + (len * 3) >= p->maxcols) + adjbuf(p, p->col + 1 + (len * 3)); + + for (i = 0; i < len; i++) { + if (ASCII_HYPH != word[i] && + ! isgraph((unsigned char)word[i])) { + p->buf[p->col++] = word[i]; + continue; + } + + if (TERMFONT_UNDER == f) + p->buf[p->col++] = '_'; + else if (ASCII_HYPH == word[i]) + p->buf[p->col++] = '-'; else - term_pescape(p, &word); + p->buf[p->col++] = word[i]; - if (term_isopendelim(word)) - p->flags |= TERMP_NOSPACE; + p->buf[p->col++] = 8; + p->buf[p->col++] = word[i]; + } } - -/* - * 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 -term_chara(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) - err(1, "realloc"); - p->maxcols = s; - } - p->buf[(int)(p->col)++] = c; + return((*p->width)(p, ' ') * sz); } +static size_t +cond_width(const struct termp *p, int c, int *skip) +{ -static void -term_encodea(struct termp *p, char c) + 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, rsz, i; + int ssz, skip, c; + const char *seq, *rhs; + enum mandoc_esc esc; + static const char rej[] = { '\\', ASCII_HYPH, ASCII_NBRSP, '\0' }; - if (TERMP_STYLE & p->flags) { - if (TERMP_BOLD & p->flags) { - term_chara(p, c); - term_chara(p, 8); + /* + * 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); + + c = 0; + switch (*cp) { + case ('\\'): + cp++; + esc = mandoc_escape(&cp, &seq, &ssz); + if (ESCAPE_ERROR == esc) + return(sz); + + if (TERMENC_ASCII != p->enc) + switch (esc) { + case (ESCAPE_UNICODE): + c = mchars_num2uc + (seq + 1, ssz - 1); + if ('\0' == c) + break; + sz += cond_width(p, c, &skip); + continue; + case (ESCAPE_SPECIAL): + c = mchars_spec2cp + (p->symtab, seq, ssz); + if (c <= 0) + break; + sz += cond_width(p, c, &skip); + continue; + default: + break; + } + + rhs = NULL; + + switch (esc) { + case (ESCAPE_UNICODE): + sz += cond_width(p, '?', &skip); + break; + case (ESCAPE_NUMBERED): + c = mchars_num2char(seq, ssz); + if ('\0' != c) + sz += cond_width(p, c, &skip); + break; + case (ESCAPE_SPECIAL): + rhs = mchars_spec2str + (p->symtab, seq, ssz, &rsz); + + if (ssz != 1 || rhs) + break; + + rhs = seq; + rsz = ssz; + break; + case (ESCAPE_SKIPCHAR): + skip = 1; + break; + default: + break; + } + + if (NULL == rhs) + break; + + if (skip) { + skip = 0; + break; + } + + 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 (TERMP_UNDER & p->flags) { - term_chara(p, '_'); - term_chara(p, 8); - } } - term_chara(p, c); + + return(sz); +} + +/* ARGSUSED */ +size_t +term_vspan(const struct termp *p, const struct roffsu *su) +{ + double r; + + 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; + } + + if (r < 0.0) + r = 0.0; + return(/* LINTED */(size_t) + r); +} + +size_t +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); }