=================================================================== RCS file: /cvs/mandoc/Attic/mdocterm.c,v retrieving revision 1.32 retrieving revision 1.42 diff -u -p -r1.32 -r1.42 --- mandoc/Attic/mdocterm.c 2009/03/04 14:41:40 1.32 +++ mandoc/Attic/mdocterm.c 2009/03/14 12:35:02 1.42 @@ -1,4 +1,4 @@ -/* $Id: mdocterm.c,v 1.32 2009/03/04 14:41:40 kristaps Exp $ */ +/* $Id: mdocterm.c,v 1.42 2009/03/14 12:35:02 kristaps Exp $ */ /* * Copyright (c) 2008 Kristaps Dzonsons * @@ -16,7 +16,7 @@ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR * PERFORMANCE OF THIS SOFTWARE. */ -#include +#include #include #include @@ -25,18 +25,16 @@ #include #include #include -#ifndef __OpenBSD__ -#include -#endif #include "mmain.h" #include "term.h" -struct termenc { +struct termseq { const char *enc; int sym; }; +static int option(void *, int, const char *); static void body(struct termp *, struct termpair *, const struct mdoc_meta *, @@ -55,14 +53,14 @@ 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); +static void sanity(const struct mdoc_node *); #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[] = { +static struct termseq termenc1[] = { { "\\", TERMSYM_SLASH }, { "\'", TERMSYM_RSQUOTE }, { "`", TERMSYM_LSQUOTE }, @@ -72,10 +70,13 @@ static struct termenc termenc1[] = { { "&", TERMSYM_BREAK }, { "e", TERMSYM_SLASH }, { "q", TERMSYM_DQUOTE }, + { "|", TERMSYM_BREAK }, { NULL, 0 } }; -static struct termenc termenc2[] = { +static struct termseq termenc2[] = { + { "rC", TERMSYM_RBRACE }, + { "lC", TERMSYM_LBRACE }, { "rB", TERMSYM_RBRACK }, { "lB", TERMSYM_LBRACK }, { "ra", TERMSYM_RANGLE }, @@ -125,7 +126,7 @@ static struct termenc termenc2[] = { { NULL, 0 } }; -static struct termsym termsym_ansi[] = { +static struct termsym termsym_ascii[TERMSYM_MAX] = { { "]", 1 }, /* TERMSYM_RBRACK */ { "[", 1 }, /* TERMSYM_LBRACK */ { "<-", 2 }, /* TERMSYM_LARROW */ @@ -165,41 +166,38 @@ static struct termsym termsym_ansi[] = { { "", 0 }, /* TERMSYM_BREAK */ { "<", 1 }, /* TERMSYM_LANGLE */ { ">", 1 }, /* TERMSYM_RANGLE */ + { "{", 1 }, /* TERMSYM_LBRACE */ + { "}", 1 }, /* TERMSYM_RBRACE */ }; -static const char ansi_clear[] = { 27, '[', '0', 'm' }; -static const char ansi_bold[] = { 27, '[', '1', 'm' }; -static const char ansi_under[] = { 27, '[', '4', 'm' }; -static struct termsym termstyle_ansi[] = { - { ansi_clear, 4 }, - { ansi_bold, 4 }, - { ansi_under, 4 } -}; - - 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(); + (void)memset(&termp, 0, sizeof(struct termp)); - if ( ! mmain_getopt(p, argc, argv, NULL, NULL, NULL, NULL)) - mmain_exit(p, 1); - - if (NULL == (mdoc = mmain_mdoc(p))) - mmain_exit(p, 1); - termp.maxrmargin = termp.rmargin = 78; /* XXX */ termp.maxcols = 1024; termp.offset = termp.col = 0; termp.flags = TERMP_NOSPACE; - termp.symtab = termsym_ansi; - termp.styletab = termstyle_ansi; + termp.symtab = termsym_ascii; + termp.enc = TERMENC_ANSI; + p = mmain_alloc(); + c = mmain_getopt(p, argc, argv, "[-Ooption...]", + "O:", &termp, option); + + if (1 != c) + mmain_exit(p, -1 == c ? 1 : 0); + + if (NULL == (mdoc = mmain_mdoc(p))) + mmain_exit(p, 1); + if (NULL == (termp.buf = malloc(termp.maxcols))) err(1, "malloc"); @@ -214,6 +212,26 @@ main(int argc, char *argv[]) } +int +option(void *ptr, int c, const char *arg) +{ + struct termp *p; + + p = (struct termp *)ptr; + + if (0 == strcmp(arg, "nroff")) { + p->enc = TERMENC_NROFF; + return(1); + } else if (0 == strcmp(arg, "ansi")) { + p->enc = TERMENC_ANSI; + return(1); + } + + warnx("unknown option: -O%s", arg); + 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 @@ -283,15 +301,18 @@ flushln(struct termp *p) * space is printed according to regular spacing rules). */ - /* FIXME: make non-ANSI friendly. */ - /* 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); - j += 3; + } else if (27 == p->buf[j]) { + assert(TERMENC_ANSI == p->enc); + assert(j + 5 <= p->col); + j += 4; + } else if (8 == p->buf[j]) { + assert(TERMENC_NROFF == p->enc); + assert(j + 2 <= p->col); + j += 1; } else vsz++; } @@ -334,7 +355,7 @@ 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]); } @@ -426,8 +447,8 @@ word(struct termp *p, const char *word) return; } - len = strlen(word); - assert(len > 0); + if (0 == (len = strlen(word))) + errx(1, "blank line not in literal context"); if (mdoc_isdelim(word)) { if ( ! (p->flags & TERMP_IGNDELIM)) @@ -437,13 +458,13 @@ 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; @@ -476,6 +497,10 @@ body(struct termp *p, struct termpair *ppair, int dochild; struct termpair pair; + /* Some quick sanity-checking. */ + + sanity(node); + /* Pre-processing. */ dochild = 1; @@ -490,7 +515,7 @@ 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. */ @@ -570,9 +595,7 @@ 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; @@ -581,49 +604,7 @@ 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: @@ -637,19 +618,20 @@ 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; @@ -678,7 +660,6 @@ header(struct termp *p, const struct mdoc_meta *meta) p->flags &= ~TERMP_NOSPACE; free(title); - free(vbuf); free(buf); } @@ -691,7 +672,7 @@ header(struct termp *p, const struct mdoc_meta *meta) static void nescape(struct termp *p, const char *word, size_t len) { - struct termenc *enc; + struct termseq *enc; switch (len) { case (1): @@ -725,24 +706,34 @@ 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; @@ -762,7 +753,10 @@ 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); } @@ -785,26 +779,70 @@ pword(struct termp *p, const char *word, size_t len) p->flags &= ~TERMP_NOSPACE; /* - * XXX - if literal and underlining, this will underline the - * spaces between literal words. + * If ANSI (word-length styling), then apply our style now, + * before the word. */ - if (p->flags & TERMP_BOLD) - stylea(p, TERMSTYLE_BOLD); - if (p->flags & TERMP_UNDERLINE) - stylea(p, TERMSTYLE_UNDER); + if (TERMENC_ANSI == p->enc && TERMP_STYLE & p->flags) { + if (TERMP_BOLD & p->flags) { + chara(p, 27); + stringa(p, "[01m", 4); + } + if (TERMP_UNDER & p->flags) { + chara(p, 27); + stringa(p, "[04m", 4); + } + if (TERMP_RED & p->flags) { + chara(p, 27); + stringa(p, "[31m", 4); + } + if (TERMP_GREEN & p->flags) { + chara(p, 27); + stringa(p, "[32m", 4); + } + if (TERMP_YELLOW & p->flags) { + chara(p, 27); + stringa(p, "[33m", 4); + } + if (TERMP_BLUE & p->flags) { + chara(p, 27); + stringa(p, "[34m", 4); + } + if (TERMP_MAGENTA & p->flags) { + chara(p, 27); + stringa(p, "[35m", 4); + } + if (TERMP_CYAN & p->flags) { + chara(p, 27); + stringa(p, "[36m", 4); + } + } for (i = 0; i < len; i++) { if ('\\' == word[i]) { pescape(p, word, &i, len); continue; } + + if (TERMENC_NROFF == p->enc && + TERMP_STYLE & p->flags) { + if (TERMP_BOLD & p->flags) { + chara(p, word[i]); + chara(p, 8); + } + if (TERMP_UNDER & p->flags) { + chara(p, '_'); + chara(p, 8); + } + } + chara(p, word[i]); } - if (p->flags & TERMP_BOLD || - p->flags & TERMP_UNDERLINE) - stylea(p, TERMSTYLE_CLEAR); + if (TERMENC_ANSI == p->enc && TERMP_STYLE & p->flags) { + chara(p, 27); + stringa(p, "[00m", 4); + } } @@ -821,18 +859,6 @@ 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. @@ -877,3 +903,114 @@ chara(struct termp *p, char c) } p->buf[(p->col)++] = c; } + + +static void +sanity(const struct mdoc_node *n) +{ + + switch (n->type) { + case (MDOC_TEXT): + if (n->child) + errx(1, "regular form violated (1)"); + if (NULL == n->parent) + errx(1, "regular form violated (2)"); + if (NULL == n->string) + errx(1, "regular form violated (3)"); + switch (n->parent->type) { + case (MDOC_TEXT): + /* FALLTHROUGH */ + case (MDOC_ROOT): + errx(1, "regular form violated (4)"); + /* NOTREACHED */ + default: + break; + } + break; + case (MDOC_ELEM): + if (NULL == n->parent) + errx(1, "regular form violated (5)"); + switch (n->parent->type) { + case (MDOC_TAIL): + /* FALLTHROUGH */ + case (MDOC_BODY): + /* FALLTHROUGH */ + case (MDOC_HEAD): + break; + default: + errx(1, "regular form violated (6)"); + /* NOTREACHED */ + } + if (n->child) switch (n->child->type) { + case (MDOC_TEXT): + break; + default: + errx(1, "regular form violated (7("); + /* NOTREACHED */ + } + break; + case (MDOC_HEAD): + /* FALLTHROUGH */ + case (MDOC_BODY): + /* FALLTHROUGH */ + case (MDOC_TAIL): + if (NULL == n->parent) + errx(1, "regular form violated (8)"); + if (MDOC_BLOCK != n->parent->type) + errx(1, "regular form violated (9)"); + if (n->child) switch (n->child->type) { + case (MDOC_BLOCK): + /* FALLTHROUGH */ + case (MDOC_ELEM): + /* FALLTHROUGH */ + case (MDOC_TEXT): + break; + default: + errx(1, "regular form violated (a)"); + /* NOTREACHED */ + } + break; + case (MDOC_BLOCK): + if (NULL == n->parent) + errx(1, "regular form violated (b)"); + if (NULL == n->child) + errx(1, "regular form violated (c)"); + switch (n->parent->type) { + case (MDOC_ROOT): + /* FALLTHROUGH */ + case (MDOC_HEAD): + /* FALLTHROUGH */ + case (MDOC_BODY): + /* FALLTHROUGH */ + case (MDOC_TAIL): + break; + default: + errx(1, "regular form violated (d)"); + /* NOTREACHED */ + } + switch (n->child->type) { + case (MDOC_ROOT): + /* FALLTHROUGH */ + case (MDOC_ELEM): + errx(1, "regular form violated (e)"); + /* NOTREACHED */ + default: + break; + } + break; + case (MDOC_ROOT): + if (n->parent) + errx(1, "regular form violated (f)"); + if (NULL == n->child) + errx(1, "regular form violated (10)"); + switch (n->child->type) { + case (MDOC_BLOCK): + break; + default: + errx(1, "regular form violated (11)"); + /* NOTREACHED */ + } + break; + } +} +