=================================================================== RCS file: /cvs/mandoc/term.c,v retrieving revision 1.95 retrieving revision 1.158 diff -u -p -r1.95 -r1.158 --- mandoc/term.c 2009/07/27 12:35:54 1.95 +++ mandoc/term.c 2010/07/03 16:02:12 1.158 @@ -1,6 +1,6 @@ -/* $Id: term.c,v 1.95 2009/07/27 12:35:54 kristaps Exp $ */ +/* $Id: term.c,v 1.158 2010/07/03 16:02:12 schwarze Exp $ */ /* - * Copyright (c) 2008, 2009 Kristaps Dzonsons + * Copyright (c) 2008, 2009 Kristaps Dzonsons * * Permission to use, copy, modify, and distribute this software for any * purpose with or without fee is hereby granted, provided that the above @@ -14,157 +14,80 @@ * 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 "chars.h" +#include "out.h" +#include "regs.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 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 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 int isopendelim(const char *); -static int 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) + chars_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 -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 -isopendelim(const char *p) +struct termp * +term_alloc(enum termenc enc) { + struct termp *p; - if ( ! (*p && 0 == *(p + 1))) - return(0); - - switch (*p) { - case('('): - /* FALLTHROUGH */ - case('['): - /* FALLTHROUGH */ - case('{'): - return(1); - default: - break; + p = calloc(1, sizeof(struct termp)); + if (NULL == p) { + perror(NULL); + exit(EXIT_FAILURE); } - return(0); + p->enc = enc; + return(p); } @@ -172,13 +95,11 @@ isopendelim(const char *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. + * 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. @@ -202,52 +123,61 @@ 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). + * 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 overstep = 0; + 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 */ + 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->offset < p->rmargin); - assert((int)(p->rmargin - p->offset) - overstep > 0); - maxvis = /* LINTED */ - p->rmargin - p->offset - overstep; - mmax = /* LINTED */ - p->maxrmargin - p->offset - overstep; + 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; - vis = 0; - overstep = 0; /* - * If in the standard case (left-justified), then begin with our - * indentation, otherwise (columns, etc.) just start spitting - * out text. + * Indent the first line of a paragraph. */ + vbl = p->flags & TERMP_NOLPAD ? 0 : p->offset; - if ( ! (p->flags & TERMP_NOLPAD)) - /* LINTED */ - for (j = 0; j < (int)p->offset; j++) - putchar(' '); + vis = vend = i = 0; - for (i = 0; i < (int)p->col; i++) { + while (i < (int)p->col) { /* + * 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; + vbl += vend - vis; + vis = vend; + } + + /* * 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 @@ -255,65 +185,109 @@ term_flushln(struct termp *p) */ /* LINTED */ - for (j = i, vsz = 0; j < (int)p->col; j++) { - if (j && ' ' == p->buf[j]) + for (jhy = 0; j < (int)p->col; j++) { + if ((j && ' ' == p->buf[j]) || '\t' == p->buf[j]) break; - else if (8 == p->buf[j]) - vsz--; - 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); if (TERMP_NOBREAK & p->flags) { - for (j = 0; j < (int)p->rmargin; j++) - putchar(' '); - vis = p->rmargin - p->offset; + p->viscol = p->rmargin; + (*p->advance)(p, p->rmargin); + vend += p->rmargin - p->offset; } else { - for (j = 0; j < (int)p->offset; j++) - putchar(' '); - vis = 0; + p->viscol = 0; + vbl = p->offset; } - } else { - for (j = 0; j < (int)vbl; j++) - putchar(' '); - vis += vbl; + + /* Remove the p->overstep width. */ + + bp += (int)/* LINTED */ + p->overstep; + p->overstep = 0; } /* - * Finally, write out the word. + * 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++) { - if (' ' == p->buf[i]) + if (vend > bp && jhy > 0 && i > jhy) break; - putchar(p->buf[i]); + if ('\t' == p->buf[i]) + break; + if (' ' == p->buf[i]) { + while (' ' == p->buf[i]) { + vbl += (*p->width)(p, p->buf[i]); + i++; + } + 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, '-'); + } else { + (*p->letter)(p, p->buf[i]); + p->viscol += (*p->width)(p, p->buf[i]); + } } - vis += vsz; + vend += vbl; + vis = vend; } + p->col = 0; + p->overstep = 0; 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 = /* LINTED */ + vis - maxvis + (*p->width)(p, ' '); /* * Behave exactly the same way as groff: @@ -325,25 +299,27 @@ term_flushln(struct termp *p) * move it one step LEFT and flag the rest of the line * to be longer. */ - if (overstep >= -1) { - assert((int)maxvis + overstep >= 0); + if (p->overstep >= -1) { + assert((int)maxvis + p->overstep >= 0); /* LINTED */ - maxvis += overstep; + maxvis += p->overstep; } else - overstep = 0; + p->overstep = 0; } 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(' '); + ((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); } } @@ -358,7 +334,7 @@ term_newln(struct termp *p) { p->flags |= TERMP_NOSPACE; - if (0 == p->col) { + if (0 == p->col && 0 == p->viscol) { p->flags &= ~TERMP_NOLPAD; return; } @@ -378,208 +354,356 @@ term_vspace(struct termp *p) { term_newln(p); - putchar('\n'); + p->viscol = 0; + (*p->endline)(p); } static void -do_special(struct termp *p, const char *word, size_t len) +spec(struct termp *p, const char *word, size_t len) { const char *rhs; size_t sz; - int i; - rhs = term_a2ascii(p->symtab, word, len, &sz); - - if (NULL == rhs) - return; - for (i = 0; i < (int)sz; i++) - encode(p, rhs[i]); + rhs = chars_a2ascii(p->symtab, word, len, &sz); + if (rhs) + encode(p, rhs, sz); } static void -do_reserved(struct termp *p, const char *word, size_t len) +res(struct termp *p, const char *word, size_t len) { const char *rhs; size_t sz; - int i; - rhs = term_a2res(p->symtab, word, len, &sz); + rhs = chars_a2res(p->symtab, word, len, &sz); + if (NULL == rhs) { + rhs = roff_getstrn(word, len); + if (rhs) + sz = strlen(rhs); + } + if (rhs) + encode(p, rhs, sz); +} - if (NULL == rhs) - return; - for (i = 0; i < (int)sz; i++) - encode(p, rhs[i]); + +void +term_fontlast(struct termp *p) +{ + enum termfont f; + + f = p->fontl; + p->fontl = p->fontq[p->fonti]; + p->fontq[p->fonti] = f; } +void +term_fontrepl(struct termp *p, enum termfont f) +{ + + p->fontl = p->fontq[p->fonti]; + p->fontq[p->fonti] = f; +} + + +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 != &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 -do_escaped(struct termp *p, const char **word) +void +term_word(struct termp *p, const char *word) { - int j; - const char *wp; + const char *sv, *seq; + int sz; + size_t ssz; + enum roffdeco deco; - wp = *word; + sv = word; - if (0 == *(++wp)) { - *word = wp; - return; + 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) + bufferc(p, ' '); + } else + bufferc(p, ASCII_NBRSP); } - 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; - do_special(p, wp, 2); - *word = ++wp; - return; + p->flags &= ~TERMP_SENTENCE; - } else if ('*' == *wp) { - if (0 == *(++wp)) { - *word = wp; - return; + /* FIXME: use strcspn. */ + + while (*word) { + if ('\\' != *word) { + encode(p, word, 1); + word++; + continue; } - switch (*wp) { - case ('('): - wp++; - if (0 == *wp || 0 == *(wp + 1)) { - *word = 0 == *wp ? wp : wp + 1; - return; - } + seq = ++word; + sz = a2roffdeco(&deco, &seq, &ssz); - do_reserved(p, wp, 2); - *word = ++wp; - return; - case ('['): + switch (deco) { + case (DECO_RESERVED): + res(p, seq, ssz); break; + case (DECO_SPECIAL): + spec(p, seq, ssz); + break; + case (DECO_BOLD): + term_fontrepl(p, TERMFONT_BOLD); + break; + case (DECO_ITALIC): + term_fontrepl(p, TERMFONT_UNDER); + break; + case (DECO_ROMAN): + term_fontrepl(p, TERMFONT_NONE); + break; + case (DECO_PREVIOUS): + term_fontlast(p); + break; default: - do_reserved(p, wp, 1); - *word = wp; - return; + break; } - - } else if ('f' == *wp) { - if (0 == *(++wp)) { - *word = wp; - return; - } - switch (*wp) { - case ('B'): - p->flags |= TERMP_BOLD; - break; - case ('I'): - p->flags |= TERMP_UNDER; - break; - case ('P'): + 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 ('R'): - p->flags &= ~TERMP_STYLE; + case('['): + p->flags |= TERMP_NOSPACE; break; default: break; } +} - *word = wp; - return; - } else if ('[' != *wp) { - do_special(p, wp, 1); - *word = wp; - return; +static void +adjbuf(struct termp *p, size_t sz) +{ + + 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(EXIT_FAILURE); } +} - wp++; - for (j = 0; *wp && ']' != *wp; wp++, j++) - /* Loop... */ ; - if (0 == *wp) { - *word = wp; +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; +} + + +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. + */ + + if (TERMFONT_NONE == (f = term_fonttop(p))) { + buffera(p, word, sz); return; } - do_special(p, wp - j, (size_t)j); - *word = wp; + for (i = 0; i < (int)sz; i++) { + if ( ! isgraph((u_char)word[i])) { + bufferc(p, word[i]); + continue; + } + + if (TERMFONT_UNDER == f) + bufferc(p, '_'); + else + bufferc(p, word[i]); + + bufferc(p, 8); + bufferc(p, 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) +size_t +term_len(const struct termp *p, size_t sz) { - const char *sv; - if (isclosedelim(word)) - if ( ! (TERMP_IGNDELIM & p->flags)) - p->flags |= TERMP_NOSPACE; + return((*p->width)(p, ' ') * sz); +} - if ( ! (TERMP_NOSPACE & p->flags)) - buffer(p, ' '); - if ( ! (p->flags & TERMP_NONOSPACE)) - p->flags &= ~TERMP_NOSPACE; +size_t +term_strlen(const struct termp *p, const char *cp) +{ + size_t sz; - for (sv = word; *word; word++) - if ('\\' != *word) - encode(p, *word); - else - do_escaped(p, &word); + for (sz = 0; *cp; cp++) + sz += (*p->width)(p, *cp); - if (isopendelim(sv)) - p->flags |= TERMP_NOSPACE; + return(sz); } -/* - * 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) +/* ARGSUSED */ +size_t +term_vspan(const struct termp *p, const struct roffsu *su) { - size_t s; + double r; - 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; + 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; } - p->buf[(int)(p->col)++] = c; + + if (r < 0.0) + r = 0.0; + return(/* LINTED */(size_t) + r); } -static void -encode(struct termp *p, char c) +size_t +term_hspan(const struct termp *p, const struct roffsu *su) { - - if (' ' != c && TERMP_STYLE & p->flags) { - if (TERMP_BOLD & p->flags) { - buffer(p, c); - buffer(p, 8); - } - if (TERMP_UNDER & p->flags) { - buffer(p, '_'); - buffer(p, 8); - } - } - buffer(p, c); + double v; + + v = ((*p->hspan)(p, su)); + if (v < 0.0) + v = 0.0; + return((size_t) /* LINTED */ + v); }