=================================================================== RCS file: /cvs/mandoc/Attic/mdocterm.c,v retrieving revision 1.25 retrieving revision 1.36 diff -u -p -r1.25 -r1.36 --- mandoc/Attic/mdocterm.c 2009/03/02 17:14:46 1.25 +++ mandoc/Attic/mdocterm.c 2009/03/08 18:02:36 1.36 @@ -1,4 +1,4 @@ -/* $Id: mdocterm.c,v 1.25 2009/03/02 17:14:46 kristaps Exp $ */ +/* $Id: mdocterm.c,v 1.36 2009/03/08 18:02:36 kristaps Exp $ */ /* * Copyright (c) 2008 Kristaps Dzonsons * @@ -16,8 +16,6 @@ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR * PERFORMANCE OF THIS SOFTWARE. */ -#include - #include #include #include @@ -32,6 +30,11 @@ #include "mmain.h" #include "term.h" +struct termenc { + const char *enc; + int sym; +}; + static void body(struct termp *, struct termpair *, const struct mdoc_meta *, @@ -44,19 +47,82 @@ static void footer(struct termp *, static void pword(struct termp *, const char *, size_t); static void pescape(struct termp *, const char *, size_t *, size_t); -static void style(struct termp *, enum tstyle); static void nescape(struct termp *, const char *, size_t); static void chara(struct termp *, char); static void stringa(struct termp *, const char *, size_t); static void symbola(struct termp *, enum tsym); +static void stylea(struct termp *, enum tstyle); #ifdef __linux__ extern size_t strlcat(char *, const char *, size_t); extern size_t strlcpy(char *, const char *, size_t); #endif +static struct termenc termenc1[] = { + { "\\", TERMSYM_SLASH }, + { "\'", TERMSYM_RSQUOTE }, + { "`", TERMSYM_LSQUOTE }, + { "-", TERMSYM_HYPHEN }, + { " ", TERMSYM_SPACE }, + { ".", TERMSYM_PERIOD }, + { "&", TERMSYM_BREAK }, + { "e", TERMSYM_SLASH }, + { "q", TERMSYM_DQUOTE }, + { NULL, 0 } +}; + +static struct termenc termenc2[] = { + { "rB", TERMSYM_RBRACK }, + { "lB", TERMSYM_LBRACK }, + { "ra", TERMSYM_RANGLE }, + { "la", TERMSYM_LANGLE }, + { "Lq", TERMSYM_LDQUOTE }, + { "lq", TERMSYM_LDQUOTE }, + { "Rq", TERMSYM_RDQUOTE }, + { "rq", TERMSYM_RDQUOTE }, + { "oq", TERMSYM_LSQUOTE }, + { "aq", TERMSYM_RSQUOTE }, + + { "<-", TERMSYM_LARROW }, + { "->", TERMSYM_RARROW }, + { "ua", TERMSYM_UARROW }, + { "da", TERMSYM_DARROW }, + + { "bu", TERMSYM_BULLET }, + { "Ba", TERMSYM_BAR }, + { "ba", TERMSYM_BAR }, + { "co", TERMSYM_COPY }, + { "Am", TERMSYM_AMP }, + + { "Le", TERMSYM_LE }, + { "<=", TERMSYM_LE }, + { "Ge", TERMSYM_GE }, + { ">=", TERMSYM_GE }, + { "==", TERMSYM_EQ }, + { "Ne", TERMSYM_NEQ }, + { "!=", TERMSYM_NEQ }, + { "Pm", TERMSYM_PLUSMINUS }, + { "+-", TERMSYM_PLUSMINUS }, + { "If", TERMSYM_INF2 }, + { "if", TERMSYM_INF }, + { "Na", TERMSYM_NAN }, + { "na", TERMSYM_NAN }, + { "**", TERMSYM_ASTERISK }, + { "Gt", TERMSYM_GT }, + { "Lt", TERMSYM_LT }, + + { "aa", TERMSYM_ACUTE }, + { "ga", TERMSYM_GRAVE }, + + { "en", TERMSYM_EN }, + { "em", TERMSYM_EM }, + + { "Pi", TERMSYM_PI }, + { NULL, 0 } +}; + static struct termsym termsym_ansi[] = { { "]", 1 }, /* TERMSYM_RBRACK */ { "[", 1 }, /* TERMSYM_LBRACK */ @@ -85,8 +151,18 @@ static struct termsym termsym_ansi[] = { { "NaN", 3 }, /* TERMSYM_NAN */ { "|", 1 }, /* TERMSYM_BAR */ { "o", 1 }, /* TERMSYM_BULLET */ - { "&", 1 }, /* TERMSYM_AND */ - { "|", 1 }, /* TERMSYM_OR */ + { "&", 1 }, /* TERMSYM_AMP */ + { "--", 2 }, /* TERMSYM_EM */ + { "-", 1 }, /* TERMSYM_EN */ + { "(C)", 3 }, /* TERMSYM_COPY */ + { "*", 1 }, /* TERMSYM_ASTERISK */ + { "\\", 1 }, /* TERMSYM_SLASH */ + { "-", 1 }, /* TERMSYM_HYPHEN */ + { " ", 1 }, /* TERMSYM_SPACE */ + { ".", 1 }, /* TERMSYM_PERIOD */ + { "", 0 }, /* TERMSYM_BREAK */ + { "<", 1 }, /* TERMSYM_LANGLE */ + { ">", 1 }, /* TERMSYM_RANGLE */ }; static const char ansi_clear[] = { 27, '[', '0', 'm' }; @@ -103,20 +179,21 @@ static struct termsym termstyle_ansi[] = { int main(int argc, char *argv[]) { - struct mmain *p; + struct mmain *p; + int c; const struct mdoc *mdoc; - struct termp termp; + struct termp termp; p = mmain_alloc(); - if ( ! mmain_getopt(p, argc, argv, NULL, NULL, NULL, NULL)) - mmain_exit(p, 1); + c = mmain_getopt(p, argc, argv, NULL, NULL, NULL, NULL); + if (1 != c) + mmain_exit(p, -1 == c ? 1 : 0); if (NULL == (mdoc = mmain_mdoc(p))) mmain_exit(p, 1); - termp.maxrmargin = 78; /* XXX */ - termp.rmargin = termp.maxrmargin; + termp.maxrmargin = termp.rmargin = 78; /* XXX */ termp.maxcols = 1024; termp.offset = termp.col = 0; termp.flags = TERMP_NOSPACE; @@ -153,9 +230,6 @@ main(int argc, char *argv[]) * offset value. This is useful when doing columnar lists where the * prior column has right-padded. * - * - TERMP_LITERAL: don't break apart words. Note that a long literal - * word will violate the right margin. - * * - 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. @@ -213,7 +287,7 @@ flushln(struct termp *p) /* LINTED */ for (j = i, vsz = 0; j < p->col; j++) { - if (isspace((int)p->buf[j])) + if (isspace((u_char)p->buf[j])) break; else if (27 == p->buf[j]) { assert(j + 4 <= p->col); @@ -246,7 +320,7 @@ flushln(struct termp *p) putchar('\n'); for (j = 0; j < p->rmargin; j++) putchar(' '); - vis = p->rmargin; + vis = p->rmargin - p->offset; } else if (vis + vsz > bp) warnx("word breaks right margin"); @@ -260,7 +334,7 @@ flushln(struct termp *p) */ for ( ; i < p->col; i++) { - if (isspace((int)p->buf[i])) + if (isspace((u_char)p->buf[i])) break; putchar(p->buf[i]); } @@ -277,9 +351,11 @@ flushln(struct termp *p) */ if ((TERMP_NOBREAK & p->flags) && vis >= maxvis) { - putchar('\n'); - for (i = 0; i < p->rmargin; i++) - putchar(' '); + if ( ! (TERMP_NONOBREAK & p->flags)) { + putchar('\n'); + for (i = 0; i < p->rmargin; i++) + putchar(' '); + } p->col = 0; return; } @@ -290,8 +366,9 @@ flushln(struct termp *p) */ if (p->flags & TERMP_NOBREAK) { - for ( ; vis < maxvis; vis++) - putchar(' '); + if ( ! (TERMP_NONOBREAK & p->flags)) + for ( ; vis < maxvis; vis++) + putchar(' '); } else putchar('\n'); @@ -360,13 +437,13 @@ word(struct termp *p, const char *word) /* LINTED */ for (j = i = 0; i < len; i++) { - if ( ! isspace((int)word[i])) { + if ( ! isspace((u_char)word[i])) { j++; continue; } /* Escaped spaces don't delimit... */ - if (i > 0 && isspace((int)word[i]) && + if (i > 0 && isspace((u_char)word[i]) && '\\' == word[i - 1]) { j++; continue; @@ -413,7 +490,7 @@ body(struct termp *p, struct termpair *ppair, if ( ! (*termacts[node->tok].pre)(p, &pair, meta, node)) dochild = 0; } else /* MDOC_TEXT == node->type */ - word(p, node->data.text.string); + word(p, node->string); /* Children. */ @@ -493,9 +570,7 @@ footer(struct termp *p, const struct mdoc_meta *meta) static void header(struct termp *p, const struct mdoc_meta *meta) { - char *buf, *title, *bufp, *vbuf; - const char *pp; - struct utsname uts; + char *buf, *title, *bufp; p->rmargin = p->maxrmargin; p->offset = 0; @@ -504,49 +579,7 @@ header(struct termp *p, const struct mdoc_meta *meta) err(1, "malloc"); if (NULL == (title = malloc(p->rmargin))) err(1, "malloc"); - if (NULL == (vbuf = malloc(p->rmargin))) - err(1, "malloc"); - if (NULL == (pp = mdoc_vol2a(meta->vol))) { - switch (meta->msec) { - case (MSEC_1): - /* FALLTHROUGH */ - case (MSEC_6): - /* FALLTHROUGH */ - case (MSEC_7): - pp = mdoc_vol2a(VOL_URM); - break; - case (MSEC_8): - pp = mdoc_vol2a(VOL_SMM); - break; - case (MSEC_2): - /* FALLTHROUGH */ - case (MSEC_3): - /* FALLTHROUGH */ - case (MSEC_4): - /* FALLTHROUGH */ - case (MSEC_5): - pp = mdoc_vol2a(VOL_PRM); - break; - case (MSEC_9): - pp = mdoc_vol2a(VOL_KM); - break; - default: - break; - } - } - vbuf[0] = 0; - - if (pp) { - if (-1 == uname(&uts)) - err(1, "uname"); - (void)strlcat(vbuf, uts.sysname, p->rmargin); - (void)strlcat(vbuf, " ", p->rmargin); - } else if (NULL == (pp = mdoc_msec2a(meta->msec))) - pp = mdoc_msec2a(MSEC_local); - - (void)strlcat(vbuf, pp, p->rmargin); - /* * The header is strange. It has three components, which are * really two with the first duplicated. It goes like this: @@ -560,19 +593,20 @@ header(struct termp *p, const struct mdoc_meta *meta) * switches on the manual section. */ - if (mdoc_arch2a(meta->arch)) - (void)snprintf(buf, p->rmargin, "%s (%s)", - vbuf, mdoc_arch2a(meta->arch)); - else - (void)strlcpy(buf, vbuf, p->rmargin); + assert(meta->vol); + (void)strlcpy(buf, meta->vol, p->rmargin); - pp = mdoc_msec2a(meta->msec); + if (meta->arch) { + (void)strlcat(buf, " (", p->rmargin); + (void)strlcat(buf, meta->arch, p->rmargin); + (void)strlcat(buf, ")", p->rmargin); + } - (void)snprintf(title, p->rmargin, "%s(%s)", - meta->title, pp ? pp : ""); + (void)snprintf(title, p->rmargin, "%s(%d)", + meta->title, meta->msec); for (bufp = title; *bufp; bufp++) - *bufp = toupper(*bufp); + *bufp = toupper((u_char)*bufp); p->offset = 0; p->rmargin = (p->maxrmargin - strlen(buf)) / 2; @@ -601,7 +635,6 @@ header(struct termp *p, const struct mdoc_meta *meta) p->flags &= ~TERMP_NOSPACE; free(title); - free(vbuf); free(buf); } @@ -614,149 +647,27 @@ header(struct termp *p, const struct mdoc_meta *meta) static void nescape(struct termp *p, const char *word, size_t len) { + struct termenc *enc; switch (len) { case (1): - switch (word[0]) { - case ('\\'): - /* FALLTHROUGH */ - case ('\''): - /* FALLTHROUGH */ - case ('`'): - /* FALLTHROUGH */ - case ('-'): - /* FALLTHROUGH */ - case (' '): - /* FALLTHROUGH */ - case ('.'): - chara(p, word[0]); /* FIXME */ - break; - case ('&'): - break; - case ('e'): - chara(p, '\\'); /* FIXME */ - break; - case ('q'): - symbola(p, TERMSYM_DQUOTE); - break; - default: - warnx("escape sequence not supported: %c", - word[0]); - break; - } + enc = termenc1; break; - case (2): - if ('r' == word[0] && 'B' == word[1]) - symbola(p, TERMSYM_RBRACK); - else if ('l' == word[0] && 'B' == word[1]) - symbola(p, TERMSYM_LBRACK); - else if ('l' == word[0] && 'q' == word[1]) - symbola(p, TERMSYM_LDQUOTE); - else if ('r' == word[0] && 'q' == word[1]) - symbola(p, TERMSYM_RDQUOTE); - else if ('o' == word[0] && 'q' == word[1]) - symbola(p, TERMSYM_LSQUOTE); - else if ('a' == word[0] && 'q' == word[1]) - symbola(p, TERMSYM_RSQUOTE); - else if ('<' == word[0] && '-' == word[1]) - symbola(p, TERMSYM_LARROW); - else if ('-' == word[0] && '>' == word[1]) - symbola(p, TERMSYM_RARROW); - else if ('b' == word[0] && 'u' == word[1]) - symbola(p, TERMSYM_BULLET); - else if ('<' == word[0] && '=' == word[1]) - symbola(p, TERMSYM_LE); - else if ('>' == word[0] && '=' == word[1]) - symbola(p, TERMSYM_GE); - else if ('=' == word[0] && '=' == word[1]) - symbola(p, TERMSYM_EQ); - else if ('+' == word[0] && '-' == word[1]) - symbola(p, TERMSYM_PLUSMINUS); - else if ('u' == word[0] && 'a' == word[1]) - symbola(p, TERMSYM_UARROW); - else if ('d' == word[0] && 'a' == word[1]) - symbola(p, TERMSYM_DARROW); - else if ('a' == word[0] && 'a' == word[1]) - symbola(p, TERMSYM_ACUTE); - else if ('g' == word[0] && 'a' == word[1]) - symbola(p, TERMSYM_GRAVE); - else if ('!' == word[0] && '=' == word[1]) - symbola(p, TERMSYM_NEQ); - else if ('i' == word[0] && 'f' == word[1]) - symbola(p, TERMSYM_INF); - else if ('n' == word[0] && 'a' == word[1]) - symbola(p, TERMSYM_NAN); - else if ('b' == word[0] && 'a' == word[1]) - symbola(p, TERMSYM_BAR); - - /* Deprecated forms. */ - else if ('A' == word[0] && 'm' == word[1]) - symbola(p, TERMSYM_AMP); - else if ('B' == word[0] && 'a' == word[1]) - symbola(p, TERMSYM_BAR); - else if ('I' == word[0] && 'f' == word[1]) - symbola(p, TERMSYM_INF2); - else if ('G' == word[0] && 'e' == word[1]) - symbola(p, TERMSYM_GE); - else if ('G' == word[0] && 't' == word[1]) - symbola(p, TERMSYM_GT); - else if ('L' == word[0] && 'e' == word[1]) - symbola(p, TERMSYM_LE); - else if ('L' == word[0] && 'q' == word[1]) - symbola(p, TERMSYM_LDQUOTE); - else if ('L' == word[0] && 't' == word[1]) - symbola(p, TERMSYM_LT); - else if ('N' == word[0] && 'a' == word[1]) - symbola(p, TERMSYM_NAN); - else if ('N' == word[0] && 'e' == word[1]) - symbola(p, TERMSYM_NEQ); - else if ('P' == word[0] && 'i' == word[1]) - symbola(p, TERMSYM_PI); - else if ('P' == word[0] && 'm' == word[1]) - symbola(p, TERMSYM_PLUSMINUS); - else if ('R' == word[0] && 'q' == word[1]) - symbola(p, TERMSYM_RDQUOTE); - else - warnx("escape sequence not supported: %c%c", - word[0], word[1]); + enc = termenc2; break; - default: - warnx("escape sequence not supported"); - break; + warnx("unsupported %zu-byte escape sequence", len); + return; } -} + for ( ; enc->enc; enc++) + if (0 == memcmp(enc->enc, word, len)) { + symbola(p, enc->sym); + return; + } -/* - * Apply a style to the output buffer. This is looked up by means of - * the styletab. - */ -static void -style(struct termp *p, enum tstyle esc) -{ - - if (p->col + 4 >= p->maxcols) - errx(1, "line overrun"); - - p->buf[(p->col)++] = 27; - p->buf[(p->col)++] = '['; - switch (esc) { - case (TERMSTYLE_CLEAR): - p->buf[(p->col)++] = '0'; - break; - case (TERMSTYLE_BOLD): - p->buf[(p->col)++] = '1'; - break; - case (TERMSTYLE_UNDER): - p->buf[(p->col)++] = '4'; - break; - default: - abort(); - /* NOTREACHED */ - } - p->buf[(p->col)++] = 'm'; + warnx("unsupported %zu-byte escape sequence", len); } @@ -770,24 +681,34 @@ pescape(struct termp *p, const char *word, size_t *i, { size_t j; - (*i)++; - assert(*i < len); + if (++(*i) >= len) { + warnx("ignoring bad escape sequence"); + return; + } if ('(' == word[*i]) { (*i)++; - assert(*i + 1 < len); + if (*i + 1 >= len) { + warnx("ignoring bad escape sequence"); + return; + } nescape(p, &word[*i], 2); (*i)++; return; } else if ('*' == word[*i]) { - /* XXX - deprecated! */ (*i)++; - assert(*i < len); + if (*i >= len) { + warnx("ignoring bad escape sequence"); + return; + } switch (word[*i]) { case ('('): (*i)++; - assert(*i + 1 < len); + if (*i + 1 >= len) { + warnx("ignoring bad escape sequence"); + return; + } nescape(p, &word[*i], 2); (*i)++; return; @@ -807,7 +728,10 @@ pescape(struct termp *p, const char *word, size_t *i, for (j = 0; word[*i] && ']' != word[*i]; (*i)++, j++) /* Loop... */ ; - assert(word[*i]); + if (0 == word[*i]) { + warnx("ignoring bad escape sequence"); + return; + } nescape(p, &word[*i - j], j); } @@ -835,9 +759,9 @@ pword(struct termp *p, const char *word, size_t len) */ if (p->flags & TERMP_BOLD) - style(p, TERMSTYLE_BOLD); + stylea(p, TERMSTYLE_BOLD); if (p->flags & TERMP_UNDERLINE) - style(p, TERMSTYLE_UNDER); + stylea(p, TERMSTYLE_UNDER); for (i = 0; i < len; i++) { if ('\\' == word[i]) { @@ -849,7 +773,7 @@ pword(struct termp *p, const char *word, size_t len) if (p->flags & TERMP_BOLD || p->flags & TERMP_UNDERLINE) - style(p, TERMSTYLE_CLEAR); + stylea(p, TERMSTYLE_CLEAR); } @@ -866,6 +790,18 @@ symbola(struct termp *p, enum tsym sym) /* + * Add a style to the output line buffer. + */ +static void +stylea(struct termp *p, enum tstyle style) +{ + + assert(p->styletab[style].sym); + stringa(p, p->styletab[style].sym, p->styletab[style].sz); +} + + +/* * Like chara() but for arbitrary-length buffers. Resize the buffer by * a factor of two (if the buffer is less than that) or the buffer's * size. @@ -874,6 +810,9 @@ static void stringa(struct termp *p, const char *c, size_t sz) { size_t s; + + if (0 == sz) + return; s = sz > p->maxcols * 2 ? sz : p->maxcols * 2;