=================================================================== RCS file: /cvs/mandoc/term.c,v retrieving revision 1.142 retrieving revision 1.191 diff -u -p -r1.142 -r1.191 --- mandoc/term.c 2010/06/07 20:57:09 1.142 +++ mandoc/term.c 2011/05/15 22:29:50 1.191 @@ -1,6 +1,7 @@ -/* $Id: term.c,v 1.142 2010/06/07 20:57:09 kristaps Exp $ */ +/* $Id: term.c,v 1.191 2011/05/15 22:29:50 kristaps Exp $ */ /* - * Copyright (c) 2008, 2009 Kristaps Dzonsons + * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons + * Copyright (c) 2010, 2011 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 @@ -26,129 +27,29 @@ #include #include #include -#include #include "mandoc.h" -#include "chars.h" #include "out.h" #include "term.h" -#include "man.h" -#include "mdoc.h" #include "main.h" -#define PS_CHAR_WIDTH 6 -#define PS_CHAR_HEIGHT 12 -#define PS_CHAR_TOPMARG (792 - 24) -#define PS_CHAR_TOP (PS_CHAR_TOPMARG - 36) -#define PS_CHAR_LEFT 36 -#define PS_CHAR_BOTMARG 24 -#define PS_CHAR_BOT (PS_CHAR_BOTMARG + 36) +static void adjbuf(struct termp *p, int); +static void bufferc(struct termp *, char); +static void encode(struct termp *, const char *, size_t); -static struct termp *alloc(char *, enum termenc, enum termtype); -static void term_free(struct termp *); -static void spec(struct termp *, const char *, size_t); -static void res(struct termp *, const char *, size_t); -static void buffera(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 void advance(struct termp *, size_t); -static void endline(struct termp *); -static void letter(struct termp *, char); -static void pageopen(struct termp *); - - -void * -ascii_alloc(char *outopts) -{ - - return(alloc(outopts, TERMENC_ASCII, TERMTYPE_CHAR)); -} - - -void * -ps_alloc(void) -{ - - return(alloc(NULL, TERMENC_ASCII, TERMTYPE_PS)); -} - - void -terminal_free(void *arg) -{ - - term_free((struct termp *)arg); -} - - -static void term_free(struct termp *p) { if (p->buf) free(p->buf); if (p->symtab) - chars_free(p->symtab); + mchars_free(p->symtab); + free(p); } -/* - * Push a single letter into our output engine. - */ -static void -letter(struct termp *p, char c) -{ - - if (TERMTYPE_CHAR == p->type) { - /* - * If using the terminal device, just push the letter - * out into the screen. - */ - putchar(c); - return; - } - - if ( ! (PS_INLINE & p->psstate)) { - /* - * If we're not in a PostScript "word" context, then - * open one now at the current cursor. - */ - printf("%zu %zu moveto\n", p->pscol, p->psrow); - putchar('('); - p->psstate |= PS_INLINE; - } - - /* - * We need to escape these characters as per the PostScript - * specification. We would also escape non-graphable characters - * (like tabs), but none of them would get to this point and - * it's superfluous to abort() on them. - */ - - switch (c) { - case ('('): - /* FALLTHROUGH */ - case (')'): - /* FALLTHROUGH */ - case ('\\'): - putchar('\\'); - break; - default: - break; - } - - /* Write the character and adjust where we are on the page. */ - putchar(c); - p->pscol += PS_CHAR_WIDTH; -} - - -/* - * Begin a "terminal" context. Since terminal encompasses PostScript, - * the actual terminal, etc., there are a few things we can do here. - */ void term_begin(struct termp *p, term_margin head, term_margin foot, const void *arg) @@ -157,174 +58,25 @@ term_begin(struct termp *p, term_margin head, p->headf = head; p->footf = foot; p->argf = arg; - - if (TERMTYPE_CHAR == p->type) { - /* Emit the header and be done. */ - (*p->headf)(p, p->argf); - return; - } - - /* - * Emit the standard PostScript prologue, set our initial page - * position, then run pageopen() on the initial page. - */ - - printf("%s\n", "%!PS"); - printf("%s\n", "/Courier"); - printf("%s\n", "10 selectfont"); - - p->pspage = 1; - p->psstate = 0; - pageopen(p); + (*p->begin)(p); } -/* - * Open a page. This is only used for -Tps at the moment. It opens a - * page context, printing the header and the footer. THE OUTPUT BUFFER - * MUST BE EMPTY. If it is not, output will ghost on the next line and - * we'll be all gross and out of state. - */ -static void -pageopen(struct termp *p) -{ - - assert(TERMTYPE_PS == p->type); - assert(0 == p->psstate); - - p->pscol = PS_CHAR_LEFT; - p->psrow = PS_CHAR_TOPMARG; - p->psstate |= PS_MARGINS; - - (*p->headf)(p, p->argf); - endline(p); - - p->psstate &= ~PS_MARGINS; - assert(0 == p->psstate); - - p->pscol = PS_CHAR_LEFT; - p->psrow = PS_CHAR_BOTMARG; - p->psstate |= PS_MARGINS; - - (*p->footf)(p, p->argf); - endline(p); - - p->psstate &= ~PS_MARGINS; - assert(0 == p->psstate); - - p->pscol = PS_CHAR_LEFT; - p->psrow = PS_CHAR_TOP; - -} - - void term_end(struct termp *p) { - if (TERMTYPE_CHAR == p->type) { - (*p->footf)(p, p->argf); - return; - } - - printf("%s\n", "%%END"); + (*p->end)(p); } -static void -endline(struct termp *p) +struct termp * +term_alloc(enum termenc enc) { - - if (TERMTYPE_CHAR == p->type) { - putchar('\n'); - return; - } - - if (PS_INLINE & p->psstate) { - printf(") show\n"); - p->psstate &= ~PS_INLINE; - } - - if (PS_MARGINS & p->psstate) - return; - - p->pscol = PS_CHAR_LEFT; - if (p->psrow >= PS_CHAR_HEIGHT + PS_CHAR_BOT) { - p->psrow -= PS_CHAR_HEIGHT; - return; - } - - /* - * XXX: can't run pageopen() until we're certain a flushln() has - * occured, else the buf will reopen in an awkward state on the - * next line. - */ - printf("showpage\n"); - p->psrow = PS_CHAR_TOP; -} - - -/* - * Advance the output engine by a certain amount of whitespace. - */ -static void -advance(struct termp *p, size_t len) -{ - size_t i; - - if (TERMTYPE_CHAR == p->type) { - /* Just print whitespace on the terminal. */ - for (i = 0; i < len; i++) - putchar(' '); - return; - } - - if (PS_INLINE & p->psstate) { - /* Dump out any existing line scope. */ - printf(") show\n"); - p->psstate &= ~PS_INLINE; - } - - p->pscol += len ? len * PS_CHAR_WIDTH : 0; -} - - -static struct termp * -alloc(char *outopts, enum termenc enc, enum termtype type) -{ struct termp *p; - const char *toks[2]; - char *v; - size_t width; - toks[0] = "width"; - toks[1] = NULL; - - p = calloc(1, sizeof(struct termp)); - if (NULL == p) { - perror(NULL); - exit(EXIT_FAILURE); - } - - p->type = type; - p->tabwidth = 5; + p = mandoc_calloc(1, sizeof(struct termp)); p->enc = enc; - - width = 80; - - while (outopts && *outopts) - switch (getsubopt(&outopts, UNCONST(toks), &v)) { - case (0): - width = atoi(v); - break; - default: - break; - } - - /* Enforce some lower boundary. */ - if (width < 60) - width = 60; - p->defrmargin = width - 2; return(p); } @@ -372,9 +124,11 @@ term_flushln(struct termp *p) 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 */ - int j; /* temporary loop index */ - int jhy; /* last hyphen before line overflow */ - size_t maxvis, mmax; + 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. @@ -382,40 +136,32 @@ term_flushln(struct termp *p) * 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; - assert(p->offset < p->rmargin); - - maxvis = (int)(p->rmargin - p->offset) - p->overstep < 0 ? - /* LINTED */ - 0 : p->rmargin - p->offset - p->overstep; - mmax = (int)(p->maxrmargin - p->offset) - p->overstep < 0 ? - /* LINTED */ - 0 : p->maxrmargin - p->offset - p->overstep; - bp = TERMP_NOBREAK & p->flags ? mmax : maxvis; /* * Indent the first line of a paragraph. */ - vbl = p->flags & TERMP_NOLPAD ? 0 : p->offset; + vbl = p->flags & TERMP_NOLPAD ? (size_t)0 : p->offset; - /* - * FIXME: if bp is zero, we still output the first word before - * breaking the line. - */ + vis = vend = 0; + i = 0; - vis = vend = i = 0; - while (i < (int)p->col) { - + while (i < p->col) { /* - * Handle literal tab characters. + * Handle literal tab characters: collapse all + * subsequent tabs into a single huge set of spaces. */ - for (j = i; j < (int)p->col; j++) { - if ('\t' != p->buf[j]) - break; - vend = (vis/p->tabwidth+1)*p->tabwidth; + while (i < p->col && '\t' == p->buf[i]) { + vend = (vis / p->tabwidth + 1) * p->tabwidth; vbl += vend - vis; vis = vend; + i++; } /* @@ -425,17 +171,24 @@ term_flushln(struct termp *p) * space is printed according to regular spacing rules). */ - /* LINTED */ - for (jhy = 0; j < (int)p->col; j++) { + for (j = i, jhy = 0; j < p->col; j++) { if ((j && ' ' == p->buf[j]) || '\t' == p->buf[j]) break; - if (8 != p->buf[j]) { - if (vend > vis && vend < bp && - ASCII_HYPH == p->buf[j]) - jhy = j; - vend++; - } else - vend--; + + /* 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]); } /* @@ -444,10 +197,10 @@ term_flushln(struct termp *p) */ if (vend > bp && 0 == jhy && vis > 0) { vend -= vis; - endline(p); + (*p->endline)(p); if (TERMP_NOBREAK & p->flags) { p->viscol = p->rmargin; - advance(p, p->rmargin); + (*p->advance)(p, p->rmargin); vend += p->rmargin - p->offset; } else { p->viscol = 0; @@ -456,32 +209,27 @@ term_flushln(struct termp *p) /* Remove the p->overstep width. */ - bp += (int)/* LINTED */ - p->overstep; + bp += (size_t)p->overstep; p->overstep = 0; } - /* - * Skip leading tabs, they were handled above. - */ - while (i < (int)p->col && '\t' == p->buf[i]) - i++; - /* Write out the [remaining] word. */ - for ( ; i < (int)p->col; i++) { + for ( ; i < p->col; i++) { if (vend > bp && jhy > 0 && i > jhy) break; if ('\t' == p->buf[i]) break; if (' ' == p->buf[i]) { - while (' ' == p->buf[i]) { - vbl++; + 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++; + vbl += (*p->width)(p, ' '); continue; } @@ -491,35 +239,40 @@ term_flushln(struct termp *p) * so write preceding white space now. */ if (vbl) { - advance(p, vbl); + (*p->advance)(p, vbl); p->viscol += vbl; vbl = 0; } - if (ASCII_HYPH == p->buf[i]) - letter(p, '-'); - else - letter(p, p->buf[i]); - - p->viscol += 1; + 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]); + } } - vend += vbl; vis = vend; } + /* + * 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 ( ! (TERMP_NOBREAK & p->flags)) { p->viscol = 0; - endline(p); + (*p->endline)(p); return; } if (TERMP_HANG & p->flags) { /* We need one blank after the tag. */ - p->overstep = /* LINTED */ - vis - maxvis + 1; + p->overstep = (int)(vis - maxvis + (*p->width)(p, ' ')); /* * Behave exactly the same way as groff: @@ -533,8 +286,7 @@ term_flushln(struct termp *p) */ if (p->overstep >= -1) { assert((int)maxvis + p->overstep >= 0); - /* LINTED */ - maxvis += p->overstep; + maxvis += (size_t)p->overstep; } else p->overstep = 0; @@ -542,15 +294,15 @@ term_flushln(struct termp *p) return; /* Right-pad. */ - if (maxvis > vis + /* LINTED */ - ((TERMP_TWOSPACE & p->flags) ? 1 : 0)) { + if (maxvis > vis + + ((TERMP_TWOSPACE & p->flags) ? (*p->width)(p, ' ') : 0)) { p->viscol += maxvis - vis; - advance(p, maxvis - vis); + (*p->advance)(p, maxvis - vis); vis += (maxvis - vis); } else { /* ...or newline break. */ - endline(p); + (*p->endline)(p); p->viscol = p->rmargin; - advance(p, p->rmargin); + (*p->advance)(p, p->rmargin); } } @@ -586,34 +338,9 @@ term_vspace(struct termp *p) term_newln(p); p->viscol = 0; - endline(p); + (*p->endline)(p); } - -static void -spec(struct termp *p, const char *word, size_t len) -{ - const char *rhs; - size_t sz; - - rhs = chars_a2ascii(p->symtab, word, len, &sz); - if (rhs) - encode(p, rhs, sz); -} - - -static void -res(struct termp *p, const char *word, size_t len) -{ - const char *rhs; - size_t sz; - - rhs = chars_a2res(p->symtab, word, len, &sz); - if (rhs) - encode(p, rhs, sz); -} - - void term_fontlast(struct termp *p) { @@ -678,7 +405,6 @@ 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 @@ -687,107 +413,84 @@ term_fontpop(struct termp *p) void term_word(struct termp *p, const char *word) { - const char *sv, *seq; + const char *seq, *cp; + char c; int sz; size_t ssz; - enum roffdeco deco; + enum mandoc_esc esc; - 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)) { - bufferc(p, ' '); - if (TERMP_SENTENCE & p->flags) + if ( ! (TERMP_KEEP & p->flags)) { + if (TERMP_PREKEEP & p->flags) + p->flags |= TERMP_KEEP; bufferc(p, ' '); + if (TERMP_SENTENCE & p->flags) + bufferc(p, ' '); + } else + bufferc(p, ASCII_NBRSP); } if ( ! (p->flags & TERMP_NONOSPACE)) p->flags &= ~TERMP_NOSPACE; + else + p->flags |= TERMP_NOSPACE; - p->flags &= ~TERMP_SENTENCE; + p->flags &= ~(TERMP_SENTENCE | TERMP_IGNDELIM); - /* FIXME: use strcspn. */ + while ('\0' != *word) { + if ((ssz = strcspn(word, "\\")) > 0) + encode(p, word, ssz); - while (*word) { - if ('\\' != *word) { - encode(p, word, 1); - word++; + word += (int)ssz; + if ('\\' != *word) continue; - } - seq = ++word; - sz = a2roffdeco(&deco, &seq, &ssz); + word++; + esc = mandoc_escape(&word, &seq, &sz); + if (ESCAPE_ERROR == esc) + break; - switch (deco) { - case (DECO_RESERVED): - res(p, seq, ssz); + switch (esc) { + case (ESCAPE_NUMBERED): + if ('\0' != (c = mchars_num2char(seq, sz))) + encode(p, &c, 1); break; - case (DECO_SPECIAL): - spec(p, seq, ssz); + case (ESCAPE_PREDEF): + cp = mchars_res2str(p->symtab, seq, sz, &ssz); + if (NULL != cp) + encode(p, cp, ssz); break; - case (DECO_BOLD): + 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 (ESCAPE_FONTBOLD): term_fontrepl(p, TERMFONT_BOLD); break; - case (DECO_ITALIC): + case (ESCAPE_FONTITALIC): term_fontrepl(p, TERMFONT_UNDER); break; - case (DECO_ROMAN): + case (ESCAPE_FONTROMAN): term_fontrepl(p, TERMFONT_NONE); break; - case (DECO_PREVIOUS): + case (ESCAPE_FONTPREV): term_fontlast(p); break; - default: + case (ESCAPE_NOSPACE): + if ('\0' == *word) + p->flags |= TERMP_NOSPACE; break; - } - - word += sz; - if (DECO_NOSPACE == deco && '\0' == *word) - p->flags |= TERMP_NOSPACE; - } - - /* - * 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; } + } } - static void -adjbuf(struct termp *p, size_t sz) +adjbuf(struct termp *p, int sz) { if (0 == p->maxcols) @@ -795,77 +498,148 @@ adjbuf(struct termp *p, size_t sz) while (sz >= p->maxcols) p->maxcols <<= 2; - p->buf = realloc(p->buf, p->maxcols); - if (NULL == p->buf) { - perror(NULL); - exit(EXIT_FAILURE); - } + p->buf = mandoc_realloc + (p->buf, sizeof(int) * (size_t)p->maxcols); } - static void -buffera(struct termp *p, const char *word, size_t sz) -{ - - if (p->col + sz >= p->maxcols) - adjbuf(p, p->col + sz); - - memcpy(&p->buf[(int)p->col], word, sz); - p->col += sz; -} - - -static void bufferc(struct termp *p, char c) { if (p->col + 1 >= p->maxcols) adjbuf(p, p->col + 1); - p->buf[(int)p->col++] = c; + p->buf[p->col++] = c; } - static void encode(struct termp *p, const char *word, size_t sz) { enum termfont f; - int i; + int i, len; + /* 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. */ - if (TERMTYPE_PS == p->type) { - buffera(p, word, sz); + 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; - } else if (TERMFONT_NONE == (f = term_fonttop(p))) { - buffera(p, word, sz); - return; } - for (i = 0; i < (int)sz; i++) { - if ( ! isgraph((u_char)word[i])) { - bufferc(p, word[i]); + /* 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 ( ! isgraph((unsigned char)word[i])) { + p->buf[p->col++] = word[i]; continue; } if (TERMFONT_UNDER == f) - bufferc(p, '_'); + p->buf[p->col++] = '_'; else - bufferc(p, word[i]); + p->buf[p->col++] = word[i]; - bufferc(p, 8); - bufferc(p, word[i]); + p->buf[p->col++] = 8; + p->buf[p->col++] = word[i]; } } +size_t +term_len(const struct termp *p, size_t sz) +{ + return((*p->width)(p, ' ') * sz); +} + + size_t -term_vspan(const struct roffsu *su) +term_strlen(const struct termp *p, const char *cp) { + size_t sz, rsz, i; + int ssz, c; + const char *seq, *rhs; + static const char rej[] = { '\\', ASCII_HYPH, ASCII_NBRSP, '\0' }; + + /* + * 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; + while ('\0' != *cp) { + rsz = strcspn(cp, rej); + for (i = 0; i < rsz; i++) + sz += (*p->width)(p, *cp++); + + switch (*cp) { + case ('\\'): + cp++; + rhs = NULL; + switch (mandoc_escape(&cp, &seq, &ssz)) { + case (ESCAPE_ERROR): + return(sz); + case (ESCAPE_NUMBERED): + c = mchars_num2char(seq, ssz); + if ('\0' != c) + sz += (*p->width)(p, c); + break; + case (ESCAPE_PREDEF): + rhs = mchars_res2str + (p->symtab, seq, ssz, &rsz); + break; + case (ESCAPE_SPECIAL): + rhs = mchars_spec2str + (p->symtab, seq, ssz, &rsz); + + if (ssz != 1 || rhs) + break; + + rhs = seq; + rsz = ssz; + break; + default: + break; + } + + if (NULL == rhs) + break; + + for (i = 0; i < rsz; i++) + sz += (*p->width)(p, *rhs++); + break; + case (ASCII_NBRSP): + sz += (*p->width)(p, ' '); + cp++; + break; + case (ASCII_HYPH): + sz += (*p->width)(p, '-'); + cp++; + break; + default: + break; + } + } + + return(sz); +} + +/* ARGSUSED */ +size_t +term_vspan(const struct termp *p, const struct roffsu *su) +{ double r; switch (su->unit) { @@ -898,43 +672,14 @@ term_vspan(const struct roffsu *su) r); } - size_t -term_hspan(const struct roffsu *su) +term_hspan(const struct termp *p, const struct roffsu *su) { - double r; + double v; - /* XXX: CM, IN, and PT are approximations. */ - - switch (su->unit) { - case (SCALE_CM): - r = 4 * su->scale; - break; - case (SCALE_IN): - /* XXX: this is an approximation. */ - r = 10 * su->scale; - break; - case (SCALE_PC): - r = (10 * su->scale) / 6; - break; - case (SCALE_PT): - r = (10 * su->scale) / 72; - break; - case (SCALE_MM): - r = su->scale / 1000; /* FIXME: double-check. */ - break; - case (SCALE_VS): - r = su->scale * 2 - 1; /* FIXME: double-check. */ - break; - default: - r = su->scale; - break; - } - - if (r < 0.0) - r = 0.0; - return((size_t)/* LINTED */ - r); + v = ((*p->hspan)(p, su)); + if (v < 0.0) + v = 0.0; + return((size_t) /* LINTED */ + v); } - -