=================================================================== RCS file: /cvs/mandoc/term.c,v retrieving revision 1.129 retrieving revision 1.138 diff -u -p -r1.129 -r1.138 --- mandoc/term.c 2010/03/23 12:42:22 1.129 +++ mandoc/term.c 2010/05/24 21:34:16 1.138 @@ -1,4 +1,4 @@ -/* $Id: term.c,v 1.129 2010/03/23 12:42:22 kristaps Exp $ */ +/* $Id: term.c,v 1.138 2010/05/24 21:34:16 schwarze Exp $ */ /* * Copyright (c) 2008, 2009 Kristaps Dzonsons * @@ -27,6 +27,7 @@ #include #include +#include "mandoc.h" #include "chars.h" #include "out.h" #include "term.h" @@ -34,7 +35,7 @@ #include "mdoc.h" #include "main.h" -static struct termp *term_alloc(enum termenc); +static struct termp *term_alloc(enum termenc, size_t); static void term_free(struct termp *); static void spec(struct termp *, const char *, size_t); static void res(struct termp *, const char *, size_t); @@ -45,10 +46,10 @@ static void encode(struct termp *, const char *, si void * -ascii_alloc(void) +ascii_alloc(size_t width) { - return(term_alloc(TERMENC_ASCII)); + return(term_alloc(TERMENC_ASCII, width)); } @@ -74,7 +75,7 @@ term_free(struct termp *p) static struct termp * -term_alloc(enum termenc enc) +term_alloc(enum termenc enc, size_t width) { struct termp *p; @@ -83,7 +84,12 @@ term_alloc(enum termenc enc) perror(NULL); exit(EXIT_FAILURE); } + p->tabwidth = 5; p->enc = enc; + /* Enforce some lower boundary. */ + if (width < 60) + width = 60; + p->defrmargin = width - 2; return(p); } @@ -92,13 +98,11 @@ term_alloc(enum termenc enc) * 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. @@ -131,7 +135,7 @@ term_flushln(struct termp *p) int i; /* current input position in p->buf */ size_t vis; /* current visual position on output */ size_t vbl; /* number of blanks to prepend to output */ - size_t 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; @@ -154,25 +158,30 @@ term_flushln(struct termp *p) bp = TERMP_NOBREAK & p->flags ? mmax : maxvis; + /* + * Indent the first line of a paragraph. + */ + vbl = p->flags & TERMP_NOLPAD ? 0 : p->offset; + /* * FIXME: if bp is zero, we still output the first word before * breaking the line. */ - vis = 0; + vis = vend = i = 0; + while (i < (int)p->col) { - /* - * If in the standard case (left-justified), then begin with our - * indentation, otherwise (columns, etc.) just start spitting - * out text. - */ + /* + * Handle literal tab characters. + */ + 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; + } - if ( ! (p->flags & TERMP_NOLPAD)) - /* LINTED */ - for (j = 0; j < (int)p->offset; j++) - putchar(' '); - - for (i = 0; i < (int)p->col; i++) { /* * Count up visible word characters. Control sequences * (starting with the CSI) aren't counted. A space @@ -181,62 +190,73 @@ term_flushln(struct termp *p) */ /* LINTED */ - for (j = i, vsz = 0; j < (int)p->col; j++) { - if (j && ' ' == p->buf[j]) + for ( ; j < (int)p->col; j++) { + if ((j && ' ' == p->buf[j]) || '\t' == p->buf[j]) break; - else if (8 == p->buf[j]) - vsz--; + if (8 == p->buf[j]) + vend--; else - vsz++; + vend++; } /* - * 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) { + if (vend > bp && vis > 0) { + vend -= vis; putchar('\n'); if (TERMP_NOBREAK & p->flags) { for (j = 0; j < (int)p->rmargin; j++) putchar(' '); - vis = p->rmargin - p->offset; + vend += p->rmargin - p->offset; } else { - for (j = 0; j < (int)p->offset; j++) - putchar(' '); - vis = 0; + vbl = p->offset; } + /* Remove the p->overstep width. */ + bp += (int)/* LINTED */ p->overstep; p->overstep = 0; - } else { - for (j = 0; j < (int)vbl; j++) - putchar(' '); - vis += vbl; } /* - * 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 ('\t' == p->buf[i]) break; + if (' ' == p->buf[i]) { + while (' ' == p->buf[i]) { + vbl++; + i++; + } + break; + } + if (ASCII_NBRSP == p->buf[i]) { + vbl++; + continue; + } - /* The unit sep. is a non-breaking space. */ - if (31 == p->buf[i]) - putchar(' '); - else - putchar(p->buf[i]); + /* + * Now we definitely know there will be + * printable characters to output, + * so write preceding white space now. + */ + if (vbl) { + for (j = 0; j < (int)vbl; j++) + putchar(' '); + vbl = 0; + } + putchar(p->buf[i]); } - vis += vsz; + vend += vbl; + vis = vend; } p->col = 0; @@ -440,8 +460,6 @@ term_word(struct termp *p, const char *word) case(')'): /* FALLTHROUGH */ case(']'): - /* FALLTHROUGH */ - case('}'): if ( ! (TERMP_IGNDELIM & p->flags)) p->flags |= TERMP_NOSPACE; break; @@ -449,12 +467,17 @@ term_word(struct termp *p, const char *word) break; } - if ( ! (TERMP_NOSPACE & p->flags)) + if ( ! (TERMP_NOSPACE & p->flags)) { bufferc(p, ' '); + if (TERMP_SENTENCE & p->flags) + bufferc(p, ' '); + } if ( ! (p->flags & TERMP_NONOSPACE)) p->flags &= ~TERMP_NOSPACE; + p->flags &= ~TERMP_SENTENCE; + /* FIXME: use strcspn. */ while (*word) { @@ -495,13 +518,15 @@ term_word(struct termp *p, const char *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('['): - /* FALLTHROUGH */ - case('{'): p->flags |= TERMP_NOSPACE; break; default: