=================================================================== RCS file: /cvs/mandoc/html.c,v retrieving revision 1.30 retrieving revision 1.236 diff -u -p -r1.30 -r1.236 --- mandoc/html.c 2009/09/16 22:17:27 1.30 +++ mandoc/html.c 2018/06/25 14:00:28 1.236 @@ -1,787 +1,1015 @@ -/* $Id: html.c,v 1.30 2009/09/16 22:17:27 kristaps Exp $ */ +/* $Id: html.c,v 1.236 2018/06/25 14:00:28 schwarze Exp $ */ /* - * Copyright (c) 2008, 2009 Kristaps Dzonsons + * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons + * Copyright (c) 2011-2015, 2017, 2018 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 * copyright notice and this permission notice appear in all copies. * - * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES + * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR + * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. */ -#include +#include "config.h" +#include + #include -#include +#include +#include +#include #include +#include #include +#include +#include -#include "mdoc.h" -#include "man.h" +#include "mandoc_aux.h" +#include "mandoc_ohash.h" +#include "mandoc.h" +#include "roff.h" +#include "out.h" +#include "html.h" +#include "manconf.h" +#include "main.h" -#define DOCTYPE "-//W3C//DTD HTML 4.01//EN" -#define DTD "http://www.w3.org/TR/html4/strict.dtd" - -enum htmltag { - TAG_HTML, - TAG_HEAD, - TAG_BODY, - TAG_META, - TAG_TITLE, - TAG_DIV, - TAG_H1, - TAG_H2, - TAG_P, - TAG_SPAN, - TAG_LINK, - TAG_BR, - TAG_A, - TAG_MAX -}; - -enum htmlattr { - ATTR_HTTPEQUIV, - ATTR_CONTENT, - ATTR_NAME, - ATTR_REL, - ATTR_HREF, - ATTR_TYPE, - ATTR_MEDIA, - ATTR_CLASS, - ATTR_MAX -}; - struct htmldata { - char *name; + const char *name; int flags; -#define HTML_CLRLINE (1 << 0) -#define HTML_NOSTACK (1 << 1) +#define HTML_NOSTACK (1 << 0) +#define HTML_AUTOCLOSE (1 << 1) +#define HTML_NLBEFORE (1 << 2) +#define HTML_NLBEGIN (1 << 3) +#define HTML_NLEND (1 << 4) +#define HTML_NLAFTER (1 << 5) +#define HTML_NLAROUND (HTML_NLBEFORE | HTML_NLAFTER) +#define HTML_NLINSIDE (HTML_NLBEGIN | HTML_NLEND) +#define HTML_NLALL (HTML_NLAROUND | HTML_NLINSIDE) +#define HTML_INDENT (1 << 6) +#define HTML_NOINDENT (1 << 7) }; static const struct htmldata htmltags[TAG_MAX] = { - {"html", HTML_CLRLINE}, /* TAG_HTML */ - {"head", HTML_CLRLINE}, /* TAG_HEAD */ - {"body", HTML_CLRLINE}, /* TAG_BODY */ - {"meta", HTML_CLRLINE | HTML_NOSTACK}, /* TAG_META */ - {"title", HTML_CLRLINE | HTML_NOSTACK}, /* TAG_TITLE */ - {"div", HTML_CLRLINE}, /* TAG_DIV */ - {"h1", 0}, /* TAG_H1 */ - {"h2", 0}, /* TAG_H2 */ - {"p", HTML_CLRLINE}, /* TAG_P */ - {"span", 0}, /* TAG_SPAN */ - {"link", HTML_CLRLINE | HTML_NOSTACK}, /* TAG_LINK */ - {"br", HTML_CLRLINE | HTML_NOSTACK}, /* TAG_LINK */ - {"a", 0}, /* TAG_A */ + {"html", HTML_NLALL}, + {"head", HTML_NLALL | HTML_INDENT}, + {"body", HTML_NLALL}, + {"meta", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL}, + {"title", HTML_NLAROUND}, + {"div", HTML_NLAROUND}, + {"div", 0}, + {"h1", HTML_NLAROUND}, + {"h2", HTML_NLAROUND}, + {"span", 0}, + {"link", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL}, + {"br", HTML_NOSTACK | HTML_AUTOCLOSE | HTML_NLALL}, + {"a", 0}, + {"table", HTML_NLALL | HTML_INDENT}, + {"tr", HTML_NLALL | HTML_INDENT}, + {"td", HTML_NLAROUND}, + {"li", HTML_NLAROUND | HTML_INDENT}, + {"ul", HTML_NLALL | HTML_INDENT}, + {"ol", HTML_NLALL | HTML_INDENT}, + {"dl", HTML_NLALL | HTML_INDENT}, + {"dt", HTML_NLAROUND}, + {"dd", HTML_NLAROUND | HTML_INDENT}, + {"pre", HTML_NLALL | HTML_NOINDENT}, + {"var", 0}, + {"cite", 0}, + {"b", 0}, + {"i", 0}, + {"code", 0}, + {"small", 0}, + {"style", HTML_NLALL | HTML_INDENT}, + {"math", HTML_NLALL | HTML_INDENT}, + {"mrow", 0}, + {"mi", 0}, + {"mn", 0}, + {"mo", 0}, + {"msup", 0}, + {"msub", 0}, + {"msubsup", 0}, + {"mfrac", 0}, + {"msqrt", 0}, + {"mfenced", 0}, + {"mtable", 0}, + {"mtr", 0}, + {"mtd", 0}, + {"munderover", 0}, + {"munder", 0}, + {"mover", 0}, }; -static const char *const htmlattrs[ATTR_MAX] = { - "http-equiv", - "content", - "name", - "rel", - "href", - "type", - "media", - "class" +static const char *const roffscales[SCALE_MAX] = { + "cm", /* SCALE_CM */ + "in", /* SCALE_IN */ + "pc", /* SCALE_PC */ + "pt", /* SCALE_PT */ + "em", /* SCALE_EM */ + "em", /* SCALE_MM */ + "ex", /* SCALE_EN */ + "ex", /* SCALE_BU */ + "em", /* SCALE_VS */ + "ex", /* SCALE_FS */ }; -struct htmlpair { - enum htmlattr key; - char *val; -}; +/* Avoid duplicate HTML id= attributes. */ +static struct ohash id_unique; -struct tag { - enum htmltag tag; - SLIST_ENTRY(tag) entry; -}; +static void print_byte(struct html *, char); +static void print_endword(struct html *); +static void print_indent(struct html *); +static void print_word(struct html *, const char *); -SLIST_HEAD(tagq, tag); +static void print_ctag(struct html *, struct tag *); +static int print_escape(struct html *, char); +static int print_encode(struct html *, const char *, const char *, int); +static void print_href(struct html *, const char *, const char *, int); +static void print_metaf(struct html *, enum mandoc_esc); -struct html { - int flags; -#define HTML_NOSPACE (1 << 0) -#define HTML_NEWLINE (1 << 1) - struct tagq stack; -}; -#define MDOC_ARGS const struct mdoc_meta *m, \ - const struct mdoc_node *n, \ - struct html *h -#define MAN_ARGS const struct man_meta *m, \ - const struct man_node *n, \ - struct html *h -struct htmlmdoc { - int (*pre)(MDOC_ARGS); - void (*post)(MDOC_ARGS); -}; - -static void print_gen_doctype(struct html *); -static void print_gen_head(struct html *); -static void print_mdoc(MDOC_ARGS); -static void print_mdoc_head(MDOC_ARGS); -static void print_mdoc_title(MDOC_ARGS); -static void print_mdoc_node(MDOC_ARGS); -static void print_man(MAN_ARGS); -static void print_man_head(MAN_ARGS); -static void print_man_body(MAN_ARGS); -static struct tag *print_otag(struct html *, enum htmltag, - int, const struct htmlpair *); -static void print_tagq(struct html *, const struct tag *); -static void print_stagq(struct html *, const struct tag *); -static void print_ctag(struct html *, enum htmltag); -static void print_encode(const char *); -static void print_text(struct html *, const char *); -static int mdoc_root_pre(MDOC_ARGS); - -static int mdoc_fl_pre(MDOC_ARGS); -static int mdoc_nd_pre(MDOC_ARGS); -static int mdoc_nm_pre(MDOC_ARGS); -static int mdoc_op_pre(MDOC_ARGS); -static void mdoc_op_post(MDOC_ARGS); -static int mdoc_pp_pre(MDOC_ARGS); -static int mdoc_sh_pre(MDOC_ARGS); -static int mdoc_ss_pre(MDOC_ARGS); -static int mdoc_xr_pre(MDOC_ARGS); - -static const struct htmlmdoc mdocs[MDOC_MAX] = { - {NULL, NULL}, /* Ap */ - {NULL, NULL}, /* Dd */ - {NULL, NULL}, /* Dt */ - {NULL, NULL}, /* Os */ - {mdoc_sh_pre, NULL }, /* Sh */ - {mdoc_ss_pre, NULL }, /* Ss */ - {mdoc_pp_pre, NULL}, /* Pp */ - {NULL, NULL}, /* D1 */ - {NULL, NULL}, /* Dl */ - {NULL, NULL}, /* Bd */ - {NULL, NULL}, /* Ed */ - {NULL, NULL}, /* Bl */ - {NULL, NULL}, /* El */ - {NULL, NULL}, /* It */ - {NULL, NULL}, /* Ad */ - {NULL, NULL}, /* An */ - {NULL, NULL}, /* Ar */ - {NULL, NULL}, /* Cd */ - {NULL, NULL}, /* Cm */ - {NULL, NULL}, /* Dv */ - {NULL, NULL}, /* Er */ - {NULL, NULL}, /* Ev */ - {NULL, NULL}, /* Ex */ - {NULL, NULL}, /* Fa */ - {NULL, NULL}, /* Fd */ - {mdoc_fl_pre, NULL}, /* Fl */ - {NULL, NULL}, /* Fn */ - {NULL, NULL}, /* Ft */ - {NULL, NULL}, /* Ic */ - {NULL, NULL}, /* In */ - {NULL, NULL}, /* Li */ - {mdoc_nd_pre, NULL}, /* Nd */ - {mdoc_nm_pre, NULL}, /* Nm */ - {mdoc_op_pre, mdoc_op_post}, /* Op */ - {NULL, NULL}, /* Ot */ - {NULL, NULL}, /* Pa */ - {NULL, NULL}, /* Rv */ - {NULL, NULL}, /* St */ - {NULL, NULL}, /* Va */ - {NULL, NULL}, /* Vt */ - {mdoc_xr_pre, NULL}, /* Xr */ - {NULL, NULL}, /* %A */ - {NULL, NULL}, /* %B */ - {NULL, NULL}, /* %D */ - {NULL, NULL}, /* %I */ - {NULL, NULL}, /* %J */ - {NULL, NULL}, /* %N */ - {NULL, NULL}, /* %O */ - {NULL, NULL}, /* %P */ - {NULL, NULL}, /* %R */ - {NULL, NULL}, /* %T */ - {NULL, NULL}, /* %V */ - {NULL, NULL}, /* Ac */ - {NULL, NULL}, /* Ao */ - {NULL, NULL}, /* Aq */ - {NULL, NULL}, /* At */ - {NULL, NULL}, /* Bc */ - {NULL, NULL}, /* Bf */ - {NULL, NULL}, /* Bo */ - {NULL, NULL}, /* Bq */ - {NULL, NULL}, /* Bsx */ - {NULL, NULL}, /* Bx */ - {NULL, NULL}, /* Db */ - {NULL, NULL}, /* Dc */ - {NULL, NULL}, /* Do */ - {NULL, NULL}, /* Dq */ - {NULL, NULL}, /* Ec */ - {NULL, NULL}, /* Ef */ - {NULL, NULL}, /* Em */ - {NULL, NULL}, /* Eo */ - {NULL, NULL}, /* Fx */ - {NULL, NULL}, /* Ms */ - {NULL, NULL}, /* No */ - {NULL, NULL}, /* Ns */ - {NULL, NULL}, /* Nx */ - {NULL, NULL}, /* Ox */ - {NULL, NULL}, /* Pc */ - {NULL, NULL}, /* Pf */ - {NULL, NULL}, /* Po */ - {NULL, NULL}, /* Pq */ - {NULL, NULL}, /* Qc */ - {NULL, NULL}, /* Ql */ - {NULL, NULL}, /* Qo */ - {NULL, NULL}, /* Qq */ - {NULL, NULL}, /* Re */ - {NULL, NULL}, /* Rs */ - {NULL, NULL}, /* Sc */ - {NULL, NULL}, /* So */ - {NULL, NULL}, /* Sq */ - {NULL, NULL}, /* Sm */ - {NULL, NULL}, /* Sx */ - {NULL, NULL}, /* Sy */ - {NULL, NULL}, /* Tn */ - {NULL, NULL}, /* Ux */ - {NULL, NULL}, /* Xc */ - {NULL, NULL}, /* Xo */ - {NULL, NULL}, /* Fo */ - {NULL, NULL}, /* Fc */ - {NULL, NULL}, /* Oo */ - {NULL, NULL}, /* Oc */ - {NULL, NULL}, /* Bk */ - {NULL, NULL}, /* Ek */ - {NULL, NULL}, /* Bt */ - {NULL, NULL}, /* Hf */ - {NULL, NULL}, /* Fr */ - {NULL, NULL}, /* Ud */ - {NULL, NULL}, /* Lb */ - {NULL, NULL}, /* Lp */ - {NULL, NULL}, /* Lk */ - {NULL, NULL}, /* Mt */ - {NULL, NULL}, /* Brq */ - {NULL, NULL}, /* Bro */ - {NULL, NULL}, /* Brc */ - {NULL, NULL}, /* %C */ - {NULL, NULL}, /* Es */ - {NULL, NULL}, /* En */ - {NULL, NULL}, /* Dx */ - {NULL, NULL}, /* %Q */ - {NULL, NULL}, /* br */ - {NULL, NULL}, /* sp */ -}; - - -void -html_mdoc(void *arg, const struct mdoc *m) +void * +html_alloc(const struct manoutput *outopts) { - struct html *h; - struct tag *t; - - h = (struct html *)arg; - - print_gen_doctype(h); - t = print_otag(h, TAG_HTML, 0, NULL); - print_mdoc(mdoc_meta(m), mdoc_node(m), h); - print_tagq(h, t); - - printf("\n"); -} - - -void -html_man(void *arg, const struct man *m) -{ struct html *h; - struct tag *t; - h = (struct html *)arg; + h = mandoc_calloc(1, sizeof(struct html)); - print_gen_doctype(h); - t = print_otag(h, TAG_HTML, 0, NULL); - print_man(man_meta(m), man_node(m), h); - print_tagq(h, t); + h->tag = NULL; + h->style = outopts->style; + h->base_man = outopts->man; + h->base_includes = outopts->includes; + if (outopts->fragment) + h->oflags |= HTML_FRAGMENT; - printf("\n"); -} + mandoc_ohash_init(&id_unique, 4, 0); - -void * -html_alloc(void) -{ - struct html *h; - - if (NULL == (h = calloc(1, sizeof(struct html)))) - return(NULL); - - SLIST_INIT(&h->stack); - return(h); + return h; } - void html_free(void *p) { struct tag *tag; struct html *h; + char *cp; + unsigned int slot; h = (struct html *)p; - - while ( ! SLIST_EMPTY(&h->stack)) { - tag = SLIST_FIRST(&h->stack); - SLIST_REMOVE_HEAD(&h->stack, entry); + while ((tag = h->tag) != NULL) { + h->tag = tag->next; free(tag); } free(h); + + cp = ohash_first(&id_unique, &slot); + while (cp != NULL) { + free(cp); + cp = ohash_next(&id_unique, &slot); + } + ohash_delete(&id_unique); } - -static void -print_mdoc(MDOC_ARGS) +void +print_gen_head(struct html *h) { struct tag *t; - t = print_otag(h, TAG_HEAD, 0, NULL); - print_mdoc_head(m, n, h); - print_tagq(h, t); + print_otag(h, TAG_META, "?", "charset", "utf-8"); + if (h->style != NULL) { + print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet", + h->style, "type", "text/css", "media", "all"); + return; + } - t = print_otag(h, TAG_BODY, 0, NULL); - print_mdoc_title(m, n, h); - print_mdoc_node(m, n, h); + /* + * Print a minimal embedded style sheet. + */ + + t = print_otag(h, TAG_STYLE, ""); + print_text(h, "table.head, table.foot { width: 100%; }"); + print_endline(h); + print_text(h, "td.head-rtitle, td.foot-os { text-align: right; }"); + print_endline(h); + print_text(h, "td.head-vol { text-align: center; }"); + print_endline(h); + print_text(h, "div.Pp { margin: 1ex 0ex; }"); + print_endline(h); + print_text(h, "div.Nd, div.Bf, div.Op { display: inline; }"); + print_endline(h); + print_text(h, "span.Pa, span.Ad { font-style: italic; }"); + print_endline(h); + print_text(h, "span.Ms { font-weight: bold; }"); + print_endline(h); + print_text(h, "dl.Bl-diag "); + print_byte(h, '>'); + print_text(h, " dt { font-weight: bold; }"); + print_endline(h); + print_text(h, "code.Nm, code.Fl, code.Cm, code.Ic, " + "code.In, code.Fd, code.Fn,"); + print_endline(h); + print_text(h, "code.Cd { font-weight: bold; " + "font-family: inherit; }"); print_tagq(h, t); } - static void -print_gen_head(struct html *h) +print_metaf(struct html *h, enum mandoc_esc deco) { - struct htmlpair meta0[2]; - struct htmlpair meta1[2]; - struct htmlpair link[4]; + enum htmlfont font; - meta0[0].key = ATTR_HTTPEQUIV; - meta0[0].val = "Content-Type"; - meta0[1].key = ATTR_CONTENT; - meta0[1].val = "text/html; charest-utf-8"; + switch (deco) { + case ESCAPE_FONTPREV: + font = h->metal; + break; + case ESCAPE_FONTITALIC: + font = HTMLFONT_ITALIC; + break; + case ESCAPE_FONTBOLD: + font = HTMLFONT_BOLD; + break; + case ESCAPE_FONTBI: + font = HTMLFONT_BI; + break; + case ESCAPE_FONT: + case ESCAPE_FONTROMAN: + font = HTMLFONT_NONE; + break; + default: + abort(); + } - meta1[0].key = ATTR_NAME; - meta1[0].val = "resource-type"; - meta1[1].key = ATTR_CONTENT; - meta1[1].val = "document"; + if (h->metaf) { + print_tagq(h, h->metaf); + h->metaf = NULL; + } - link[0].key = ATTR_REL; - link[0].val = "stylesheet"; - link[1].key = ATTR_HREF; - link[1].val = "style.css"; /* XXX */ - link[2].key = ATTR_TYPE; - link[2].val = "text/css"; - link[3].key = ATTR_MEDIA; - link[3].val = "all"; + h->metal = h->metac; + h->metac = font; - print_otag(h, TAG_META, 2, meta0); - print_otag(h, TAG_META, 2, meta1); - print_otag(h, TAG_LINK, 4, link); + switch (font) { + case HTMLFONT_ITALIC: + h->metaf = print_otag(h, TAG_I, ""); + break; + case HTMLFONT_BOLD: + h->metaf = print_otag(h, TAG_B, ""); + break; + case HTMLFONT_BI: + h->metaf = print_otag(h, TAG_B, ""); + print_otag(h, TAG_I, ""); + break; + default: + break; + } } - -/* ARGSUSED */ -static void -print_mdoc_head(MDOC_ARGS) +char * +html_make_id(const struct roff_node *n, int unique) { + const struct roff_node *nch; + char *buf, *bufs, *cp; + unsigned int slot; + int suffix; - print_gen_head(h); - print_otag(h, TAG_TITLE, 0, NULL); - print_encode(m->title); -} + for (nch = n->child; nch != NULL; nch = nch->next) + if (nch->type != ROFFT_TEXT) + return NULL; + buf = NULL; + deroff(&buf, n); + if (buf == NULL) + return NULL; -/* ARGSUSED */ -static void -print_mdoc_title(MDOC_ARGS) -{ + /* + * In ID attributes, only use ASCII characters that are + * permitted in URL-fragment strings according to the + * explicit list at: + * https://url.spec.whatwg.org/#url-fragment-string + */ - /* TODO */ -} + for (cp = buf; *cp != '\0'; cp++) + if (isalnum((unsigned char)*cp) == 0 && + strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL) + *cp = '_'; + if (unique == 0) + return buf; -static void -print_mdoc_node(MDOC_ARGS) + /* Avoid duplicate HTML id= attributes. */ + + bufs = NULL; + suffix = 1; + slot = ohash_qlookup(&id_unique, buf); + cp = ohash_find(&id_unique, slot); + if (cp != NULL) { + while (cp != NULL) { + free(bufs); + if (++suffix > 127) { + free(buf); + return NULL; + } + mandoc_asprintf(&bufs, "%s_%d", buf, suffix); + slot = ohash_qlookup(&id_unique, bufs); + cp = ohash_find(&id_unique, slot); + } + free(buf); + buf = bufs; + } + ohash_insert(&id_unique, slot, buf); + return buf; +} + +int +html_strlen(const char *cp) { - int child; - struct tag *t; + size_t rsz; + int skip, sz; - child = 1; - t = SLIST_FIRST(&h->stack); + /* + * Account for escaped sequences within string length + * calculations. This follows the logic in term_strlen() as we + * must calculate the width of produced strings. + * Assume that characters are always width of "1". This is + * hacky, but it gets the job done for approximation of widths. + */ - switch (n->type) { - case (MDOC_ROOT): - child = mdoc_root_pre(m, n, h); - break; - case (MDOC_TEXT): - print_text(h, n->string); - break; - default: - if (mdocs[n->tok].pre) - child = (*mdocs[n->tok].pre)(m, n, h); - break; + sz = 0; + skip = 0; + while (1) { + rsz = strcspn(cp, "\\"); + if (rsz) { + cp += rsz; + if (skip) { + skip = 0; + rsz--; + } + sz += rsz; + } + if ('\0' == *cp) + break; + cp++; + switch (mandoc_escape(&cp, NULL, NULL)) { + case ESCAPE_ERROR: + return sz; + case ESCAPE_UNICODE: + case ESCAPE_NUMBERED: + case ESCAPE_SPECIAL: + case ESCAPE_OVERSTRIKE: + if (skip) + skip = 0; + else + sz++; + break; + case ESCAPE_SKIPCHAR: + skip = 1; + break; + default: + break; + } } + return sz; +} - if (child && n->child) - print_mdoc_node(m, n->child, h); +static int +print_escape(struct html *h, char c) +{ - print_stagq(h, t); - - switch (n->type) { - case (MDOC_ROOT): + switch (c) { + case '<': + print_word(h, "<"); break; - case (MDOC_TEXT): + case '>': + print_word(h, ">"); break; - default: - if (mdocs[n->tok].post) - (*mdocs[n->tok].post)(m, n, h); + case '&': + print_word(h, "&"); break; + case '"': + print_word(h, """); + break; + case ASCII_NBRSP: + print_word(h, " "); + break; + case ASCII_HYPH: + print_byte(h, '-'); + break; + case ASCII_BREAK: + break; + default: + return 0; } - - if (n->next) - print_mdoc_node(m, n->next, h); + return 1; } - -static void -print_man(MAN_ARGS) +static int +print_encode(struct html *h, const char *p, const char *pend, int norecurse) { + char numbuf[16]; struct tag *t; + const char *seq; + size_t sz; + int c, len, breakline, nospace; + enum mandoc_esc esc; + static const char rejs[10] = { ' ', '\\', '<', '>', '&', '"', + ASCII_NBRSP, ASCII_HYPH, ASCII_BREAK, '\0' }; - t = print_otag(h, TAG_HEAD, 0, NULL); - print_man_head(m, n, h); - print_tagq(h, t); + if (pend == NULL) + pend = strchr(p, '\0'); - t = print_otag(h, TAG_BODY, 0, NULL); - print_man_body(m, n, h); - print_tagq(h, t); -} + breakline = 0; + nospace = 0; + while (p < pend) { + if (HTML_SKIPCHAR & h->flags && '\\' != *p) { + h->flags &= ~HTML_SKIPCHAR; + p++; + continue; + } -/* ARGSUSED */ -static void -print_man_head(MAN_ARGS) -{ + for (sz = strcspn(p, rejs); sz-- && p < pend; p++) + print_byte(h, *p); - print_gen_head(h); - print_otag(h, TAG_TITLE, 0, NULL); - print_encode(m->title); -} + if (breakline && + (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) { + t = print_otag(h, TAG_DIV, ""); + print_text(h, "\\~"); + print_tagq(h, t); + breakline = 0; + while (p < pend && (*p == ' ' || *p == ASCII_NBRSP)) + p++; + continue; + } + if (p >= pend) + break; -/* ARGSUSED */ -static void -print_man_body(MAN_ARGS) -{ + if (*p == ' ') { + print_endword(h); + p++; + continue; + } - /* TODO */ -} + if (print_escape(h, *p++)) + continue; + esc = mandoc_escape(&p, &seq, &len); + if (ESCAPE_ERROR == esc) + break; + switch (esc) { + case ESCAPE_FONT: + case ESCAPE_FONTPREV: + case ESCAPE_FONTBOLD: + case ESCAPE_FONTITALIC: + case ESCAPE_FONTBI: + case ESCAPE_FONTROMAN: + if (0 == norecurse) + print_metaf(h, esc); + continue; + case ESCAPE_SKIPCHAR: + h->flags |= HTML_SKIPCHAR; + continue; + default: + break; + } + + if (h->flags & HTML_SKIPCHAR) { + h->flags &= ~HTML_SKIPCHAR; + continue; + } + + switch (esc) { + case ESCAPE_UNICODE: + /* Skip past "u" header. */ + c = mchars_num2uc(seq + 1, len - 1); + break; + case ESCAPE_NUMBERED: + c = mchars_num2char(seq, len); + if (c < 0) + continue; + break; + case ESCAPE_SPECIAL: + c = mchars_spec2cp(seq, len); + if (c <= 0) + continue; + break; + case ESCAPE_BREAK: + breakline = 1; + continue; + case ESCAPE_NOSPACE: + if ('\0' == *p) + nospace = 1; + continue; + case ESCAPE_OVERSTRIKE: + if (len == 0) + continue; + c = seq[len - 1]; + break; + default: + continue; + } + if ((c < 0x20 && c != 0x09) || + (c > 0x7E && c < 0xA0)) + c = 0xFFFD; + if (c > 0x7E) { + (void)snprintf(numbuf, sizeof(numbuf), "&#x%.4X;", c); + print_word(h, numbuf); + } else if (print_escape(h, c) == 0) + print_byte(h, c); + } + + return nospace; +} + static void -print_encode(const char *p) +print_href(struct html *h, const char *name, const char *sec, int man) { + const char *p, *pp; - printf("%s", p); /* XXX */ + pp = man ? h->base_man : h->base_includes; + while ((p = strchr(pp, '%')) != NULL) { + print_encode(h, pp, p, 1); + if (man && p[1] == 'S') { + if (sec == NULL) + print_byte(h, '1'); + else + print_encode(h, sec, NULL, 1); + } else if ((man && p[1] == 'N') || + (man == 0 && p[1] == 'I')) + print_encode(h, name, NULL, 1); + else + print_encode(h, p, p + 2, 1); + pp = p + 2; + } + if (*pp != '\0') + print_encode(h, pp, NULL, 1); } - -static struct tag * -print_otag(struct html *h, enum htmltag tag, - int sz, const struct htmlpair *p) +struct tag * +print_otag(struct html *h, enum htmltag tag, const char *fmt, ...) { - int i; + va_list ap; + struct roffsu mysu, *su; + char numbuf[16]; struct tag *t; + const char *attr; + char *arg1, *arg2; + double v; + int i, have_style, tflags; - if ( ! (HTML_NOSTACK & htmltags[tag].flags)) { - if (NULL == (t = malloc(sizeof(struct tag)))) - err(EXIT_FAILURE, "malloc"); + tflags = htmltags[tag].flags; + + /* Push this tag onto the stack of open scopes. */ + + if ((tflags & HTML_NOSTACK) == 0) { + t = mandoc_malloc(sizeof(struct tag)); t->tag = tag; - SLIST_INSERT_HEAD(&h->stack, t, entry); + t->next = h->tag; + h->tag = t; } else t = NULL; - if ( ! (HTML_NOSPACE & h->flags)) - if ( ! (HTML_CLRLINE & htmltags[tag].flags)) - printf(" "); - - printf("<%s", htmltags[tag].name); - for (i = 0; i < sz; i++) { - printf(" %s=\"", htmlattrs[p[i].key]); - assert(p->val); - print_encode(p[i].val); - printf("\""); + if (tflags & HTML_NLBEFORE) + print_endline(h); + if (h->col == 0) + print_indent(h); + else if ((h->flags & HTML_NOSPACE) == 0) { + if (h->flags & HTML_KEEP) + print_word(h, " "); + else { + if (h->flags & HTML_PREKEEP) + h->flags |= HTML_KEEP; + print_endword(h); + } } - printf(">"); - h->flags |= HTML_NOSPACE; - if (HTML_CLRLINE & htmltags[tag].flags) - h->flags |= HTML_NEWLINE; + if ( ! (h->flags & HTML_NONOSPACE)) + h->flags &= ~HTML_NOSPACE; else - h->flags &= ~HTML_NEWLINE; + h->flags |= HTML_NOSPACE; - return(t); -} + /* Print out the tag name and attributes. */ + print_byte(h, '<'); + print_word(h, htmltags[tag].name); -/* ARGSUSED */ -static void -print_ctag(struct html *h, enum htmltag tag) -{ - - printf("", htmltags[tag].name); - if (HTML_CLRLINE & htmltags[tag].flags) - h->flags |= HTML_NOSPACE; - if (HTML_CLRLINE & htmltags[tag].flags) - h->flags |= HTML_NEWLINE; - else - h->flags &= ~HTML_NEWLINE; -} + va_start(ap, fmt); + have_style = 0; + while (*fmt != '\0') { + if (*fmt == 's') { + have_style = 1; + fmt++; + break; + } -/* ARGSUSED */ -static void -print_gen_doctype(struct html *h) -{ - - printf("\n", DOCTYPE, DTD); -} + /* Parse a non-style attribute and its arguments. */ + arg1 = va_arg(ap, char *); + switch (*fmt++) { + case 'c': + attr = "class"; + break; + case 'h': + attr = "href"; + break; + case 'i': + attr = "id"; + break; + case '?': + attr = arg1; + arg1 = va_arg(ap, char *); + break; + default: + abort(); + } + arg2 = NULL; + if (*fmt == 'M') + arg2 = va_arg(ap, char *); + if (arg1 == NULL) + continue; -static void -print_text(struct html *h, const char *p) -{ + /* Print the non-style attributes. */ - if (*p && 0 == *(p + 1)) - switch (*p) { - case('.'): - /* FALLTHROUGH */ - case(','): - /* FALLTHROUGH */ - case(';'): - /* FALLTHROUGH */ - case(':'): - /* FALLTHROUGH */ - case('?'): - /* FALLTHROUGH */ - case('!'): - /* FALLTHROUGH */ - case(')'): - /* FALLTHROUGH */ - case(']'): - /* FALLTHROUGH */ - case('}'): - h->flags |= HTML_NOSPACE; + print_byte(h, ' '); + print_word(h, attr); + print_byte(h, '='); + print_byte(h, '"'); + switch (*fmt) { + case 'I': + print_href(h, arg1, NULL, 0); + fmt++; break; + case 'M': + print_href(h, arg1, arg2, 1); + fmt++; + break; + case 'R': + print_byte(h, '#'); + print_encode(h, arg1, NULL, 1); + fmt++; + break; + case 'T': + print_encode(h, arg1, NULL, 1); + print_word(h, "\" title=\""); + print_encode(h, arg1, NULL, 1); + fmt++; + break; default: + print_encode(h, arg1, NULL, 1); break; } + print_byte(h, '"'); + } - if ( ! (h->flags & HTML_NOSPACE)) - printf(" "); + /* Print out styles. */ - h->flags &= ~HTML_NOSPACE; - h->flags &= ~HTML_NEWLINE; + while (*fmt != '\0') { + arg1 = NULL; + su = NULL; - if (p) - print_encode(p); + /* First letter: input argument type. */ - if (*p && 0 == *(p + 1)) - switch (*p) { - case('('): - /* FALLTHROUGH */ - case('['): - /* FALLTHROUGH */ - case('{'): - h->flags |= HTML_NOSPACE; + switch (*fmt++) { + case 'h': + i = va_arg(ap, int); + su = &mysu; + SCALE_HS_INIT(su, i); break; - default: + case 's': + arg1 = va_arg(ap, char *); break; + case 'u': + su = va_arg(ap, struct roffsu *); + break; + default: + abort(); } -} + /* Second letter: style name. */ -static void -print_tagq(struct html *h, const struct tag *until) -{ - struct tag *tag; + switch (*fmt++) { + case 'h': + attr = "height"; + break; + case 'l': + attr = "margin-left"; + break; + case '?': + attr = arg1; + arg1 = va_arg(ap, char *); + break; + default: + abort(); + } + if (su == NULL && arg1 == NULL) + continue; - while ( ! SLIST_EMPTY(&h->stack)) { - tag = SLIST_FIRST(&h->stack); - print_ctag(h, tag->tag); - SLIST_REMOVE_HEAD(&h->stack, entry); - free(tag); - if (until && tag == until) - return; + if (have_style == 1) + print_word(h, " style=\""); + else + print_byte(h, ' '); + print_word(h, attr); + print_byte(h, ':'); + print_byte(h, ' '); + if (su != NULL) { + v = su->scale; + if (su->unit == SCALE_MM && (v /= 100.0) == 0.0) + v = 1.0; + else if (su->unit == SCALE_BU) + v /= 24.0; + (void)snprintf(numbuf, sizeof(numbuf), "%.2f", v); + print_word(h, numbuf); + print_word(h, roffscales[su->unit]); + } else + print_word(h, arg1); + print_byte(h, ';'); + have_style = 2; } -} + if (have_style == 2) + print_byte(h, '"'); + va_end(ap); + /* Accommodate for "well-formed" singleton escaping. */ + + if (HTML_AUTOCLOSE & htmltags[tag].flags) + print_byte(h, '/'); + + print_byte(h, '>'); + + if (tflags & HTML_NLBEGIN) + print_endline(h); + else + h->flags |= HTML_NOSPACE; + + if (tflags & HTML_INDENT) + h->indent++; + if (tflags & HTML_NOINDENT) + h->noindent++; + + return t; +} + static void -print_stagq(struct html *h, const struct tag *suntil) +print_ctag(struct html *h, struct tag *tag) { - struct tag *tag; + int tflags; - while ( ! SLIST_EMPTY(&h->stack)) { - tag = SLIST_FIRST(&h->stack); - if (suntil && tag == suntil) - return; - print_ctag(h, tag->tag); - SLIST_REMOVE_HEAD(&h->stack, entry); - free(tag); - } -} + /* + * Remember to close out and nullify the current + * meta-font and table, if applicable. + */ + if (tag == h->metaf) + h->metaf = NULL; + if (tag == h->tblt) + h->tblt = NULL; + tflags = htmltags[tag->tag].flags; -/* ARGSUSED */ -static int -mdoc_root_pre(MDOC_ARGS) -{ - struct htmlpair tag; + if (tflags & HTML_INDENT) + h->indent--; + if (tflags & HTML_NOINDENT) + h->noindent--; + if (tflags & HTML_NLEND) + print_endline(h); + print_indent(h); + print_byte(h, '<'); + print_byte(h, '/'); + print_word(h, htmltags[tag->tag].name); + print_byte(h, '>'); + if (tflags & HTML_NLAFTER) + print_endline(h); - tag.key = ATTR_CLASS; - tag.val = "body"; + h->tag = tag->next; + free(tag); +} - print_otag(h, TAG_DIV, 1, &tag); - return(1); +void +print_gen_decls(struct html *h) +{ + print_word(h, ""); + print_endline(h); } - -/* ARGSUSED */ -static int -mdoc_ss_pre(MDOC_ARGS) +void +print_gen_comment(struct html *h, struct roff_node *n) { + int wantblank; - if (MDOC_BODY == n->type) - print_otag(h, TAG_P, 0, NULL); - if (MDOC_HEAD == n->type) - print_otag(h, TAG_H2, 0, NULL); - return(1); + print_word(h, "") == NULL && + (wantblank || *n->string != '\0')) { + print_endline(h); + print_indent(h); + print_word(h, n->string); + wantblank = *n->string != '\0'; + } + n = n->next; + } + if (wantblank) + print_endline(h); + print_word(h, " -->"); + print_endline(h); + h->indent = 0; } - -/* ARGSUSED */ -static int -mdoc_fl_pre(MDOC_ARGS) +void +print_text(struct html *h, const char *word) { - struct htmlpair tag; + if (h->col && (h->flags & HTML_NOSPACE) == 0) { + if ( ! (HTML_KEEP & h->flags)) { + if (HTML_PREKEEP & h->flags) + h->flags |= HTML_KEEP; + print_endword(h); + } else + print_word(h, " "); + } - tag.key = ATTR_CLASS; - tag.val = "flag"; + assert(NULL == h->metaf); + switch (h->metac) { + case HTMLFONT_ITALIC: + h->metaf = print_otag(h, TAG_I, ""); + break; + case HTMLFONT_BOLD: + h->metaf = print_otag(h, TAG_B, ""); + break; + case HTMLFONT_BI: + h->metaf = print_otag(h, TAG_B, ""); + print_otag(h, TAG_I, ""); + break; + default: + print_indent(h); + break; + } - print_otag(h, TAG_SPAN, 1, &tag); - print_text(h, "\\-"); - h->flags |= HTML_NOSPACE; - return(1); -} + assert(word); + if ( ! print_encode(h, word, NULL, 0)) { + if ( ! (h->flags & HTML_NONOSPACE)) + h->flags &= ~HTML_NOSPACE; + h->flags &= ~HTML_NONEWLINE; + } else + h->flags |= HTML_NOSPACE | HTML_NONEWLINE; + if (h->metaf) { + print_tagq(h, h->metaf); + h->metaf = NULL; + } -/* ARGSUSED */ -static int -mdoc_pp_pre(MDOC_ARGS) + h->flags &= ~HTML_IGNDELIM; +} + +void +print_tagq(struct html *h, const struct tag *until) { + struct tag *tag; - print_otag(h, TAG_BR, 0, NULL); - print_otag(h, TAG_BR, 0, NULL); - return(0); + while ((tag = h->tag) != NULL) { + print_ctag(h, tag); + if (until && tag == until) + return; + } } - -/* ARGSUSED */ -static int -mdoc_nd_pre(MDOC_ARGS) +void +print_stagq(struct html *h, const struct tag *suntil) { + struct tag *tag; - if (MDOC_BODY == n->type) - print_text(h, "--"); - return(1); + while ((tag = h->tag) != NULL) { + if (suntil && tag == suntil) + return; + print_ctag(h, tag); + } } - -/* ARGSUSED */ -static int -mdoc_op_pre(MDOC_ARGS) +void +print_paragraph(struct html *h) { + struct tag *t; - if (MDOC_BODY == n->type) { - print_text(h, "\\(lB"); - h->flags |= HTML_NOSPACE; - } - return(1); + t = print_otag(h, TAG_DIV, "c", "Pp"); + print_tagq(h, t); } -/* ARGSUSED */ +/*********************************************************************** + * Low level output functions. + * They implement line breaking using a short static buffer. + ***********************************************************************/ + +/* + * Buffer one HTML output byte. + * If the buffer is full, flush and deactivate it and start a new line. + * If the buffer is inactive, print directly. + */ static void -mdoc_op_post(MDOC_ARGS) +print_byte(struct html *h, char c) { + if ((h->flags & HTML_BUFFER) == 0) { + putchar(c); + h->col++; + return; + } - if (MDOC_BODY != n->type) + if (h->col + h->bufcol < sizeof(h->buf)) { + h->buf[h->bufcol++] = c; return; - h->flags |= HTML_NOSPACE; - print_text(h, "\\(rB"); + } + + putchar('\n'); + h->col = 0; + print_indent(h); + putchar(' '); + putchar(' '); + fwrite(h->buf, h->bufcol, 1, stdout); + putchar(c); + h->col = (h->indent + 1) * 2 + h->bufcol + 1; + h->bufcol = 0; + h->flags &= ~HTML_BUFFER; } - -static int -mdoc_nm_pre(MDOC_ARGS) +/* + * If something was printed on the current output line, end it. + * Not to be called right after print_indent(). + */ +void +print_endline(struct html *h) { - struct htmlpair class; + if (h->col == 0) + return; - if ( ! (HTML_NEWLINE & h->flags)) - if (SEC_SYNOPSIS == n->sec) - print_otag(h, TAG_BR, 0, NULL); - - class.key = ATTR_CLASS; - class.val = "name"; - - print_otag(h, TAG_SPAN, 1, &class); - if (NULL == n->child) - print_text(h, m->name); - - return(1); + if (h->bufcol) { + putchar(' '); + fwrite(h->buf, h->bufcol, 1, stdout); + h->bufcol = 0; + } + putchar('\n'); + h->col = 0; + h->flags |= HTML_NOSPACE; + h->flags &= ~HTML_BUFFER; } - -/* ARGSUSED */ -static int -mdoc_sh_pre(MDOC_ARGS) +/* + * Flush the HTML output buffer. + * If it is inactive, activate it. + */ +static void +print_endword(struct html *h) { + if (h->noindent) { + print_byte(h, ' '); + return; + } - if (MDOC_BODY == n->type) - print_otag(h, TAG_P, 0, NULL); - if (MDOC_HEAD == n->type) - print_otag(h, TAG_H1, 0, NULL); - return(1); + if ((h->flags & HTML_BUFFER) == 0) { + h->col++; + h->flags |= HTML_BUFFER; + } else if (h->bufcol) { + putchar(' '); + fwrite(h->buf, h->bufcol, 1, stdout); + h->col += h->bufcol + 1; + } + h->bufcol = 0; } - -/* ARGSUSED */ -static int -mdoc_xr_pre(MDOC_ARGS) +/* + * If at the beginning of a new output line, + * perform indentation and mark the line as containing output. + * Make sure to really produce some output right afterwards, + * but do not use print_otag() for producing it. + */ +static void +print_indent(struct html *h) { - struct htmlpair tag; + size_t i; - tag.key = ATTR_HREF; - tag.val = "#"; /* TODO */ + if (h->col) + return; - print_otag(h, TAG_A, 1, &tag); + if (h->noindent == 0) { + h->col = h->indent * 2; + for (i = 0; i < h->col; i++) + putchar(' '); + } + h->flags &= ~HTML_NOSPACE; +} - n = n->child; - print_text(h, n->string); - if (NULL == (n = n->next)) - return(0); - - h->flags |= HTML_NOSPACE; - print_text(h, "("); - h->flags |= HTML_NOSPACE; - print_text(h, n->string); - h->flags |= HTML_NOSPACE; - print_text(h, ")"); - - return(0); +/* + * Print or buffer some characters + * depending on the current HTML output buffer state. + */ +static void +print_word(struct html *h, const char *cp) +{ + while (*cp != '\0') + print_byte(h, *cp++); }