version 1.15, 2008/12/08 12:46:28 |
version 1.276, 2022/06/24 11:15:53 |
|
|
/* $Id$ */ |
/* $Id$ */ |
/* |
/* |
* Copyright (c) 2008 Kristaps Dzonsons <kristaps@kth.se> |
* Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv> |
|
* Copyright (c) 2011-2015, 2017-2021 Ingo Schwarze <schwarze@openbsd.org> |
* |
* |
* Permission to use, copy, modify, and distribute this software for any |
* Permission to use, copy, modify, and distribute this software for any |
* purpose with or without fee is hereby granted, provided that the |
* purpose with or without fee is hereby granted, provided that the above |
* above copyright notice and this permission notice appear in all |
* copyright notice and this permission notice appear in all copies. |
* copies. |
|
* |
* |
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL |
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHORS DISCLAIM ALL WARRANTIES |
* WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED |
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF |
* WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE |
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR |
* AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL |
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES |
* DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR |
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN |
* PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER |
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF |
* TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR |
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. |
* PERFORMANCE OF THIS SOFTWARE. |
* |
|
* Common functions for mandoc(1) HTML formatters. |
|
* For use by individual formatters and by the main program. |
*/ |
*/ |
#include <sys/param.h> |
#include "config.h" |
|
|
|
#include <sys/types.h> |
#include <sys/stat.h> |
#include <sys/stat.h> |
|
|
#include <assert.h> |
#include <assert.h> |
#include <err.h> |
#include <ctype.h> |
#include <fcntl.h> |
#include <stdarg.h> |
#include <stdlib.h> |
#include <stddef.h> |
#include <stdio.h> |
#include <stdio.h> |
|
#include <stdint.h> |
|
#include <stdlib.h> |
#include <string.h> |
#include <string.h> |
#include <unistd.h> |
#include <unistd.h> |
|
|
#include "libmdocml.h" |
#include "mandoc_aux.h" |
#include "private.h" |
#include "mandoc_ohash.h" |
#include "ml.h" |
#include "mandoc.h" |
|
#include "roff.h" |
|
#include "out.h" |
|
#include "html.h" |
|
#include "manconf.h" |
|
#include "main.h" |
|
|
|
struct htmldata { |
|
const char *name; |
|
int flags; |
|
#define HTML_INPHRASE (1 << 0) /* Can appear in phrasing context. */ |
|
#define HTML_TOPHRASE (1 << 1) /* Establishes phrasing context. */ |
|
#define HTML_NOSTACK (1 << 2) /* Does not have an end tag. */ |
|
#define HTML_NLBEFORE (1 << 3) /* Output line break before opening. */ |
|
#define HTML_NLBEGIN (1 << 4) /* Output line break after opening. */ |
|
#define HTML_NLEND (1 << 5) /* Output line break before closing. */ |
|
#define HTML_NLAFTER (1 << 6) /* Output line break after closing. */ |
|
#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 << 7) /* Indent content by two spaces. */ |
|
#define HTML_NOINDENT (1 << 8) /* Exception: never indent content. */ |
|
}; |
|
|
/* TODO: allow head/tail-less invocations (just "div" start). */ |
static const struct htmldata htmltags[TAG_MAX] = { |
|
{"html", HTML_NLALL}, |
struct htmlnode { |
{"head", HTML_NLALL | HTML_INDENT}, |
int tok; |
{"meta", HTML_NOSTACK | HTML_NLALL}, |
enum md_ns ns; |
{"link", HTML_NOSTACK | HTML_NLALL}, |
int argc[ROFF_MAXLINEARG]; |
{"style", HTML_NLALL | HTML_INDENT}, |
char *argv[ROFF_MAXLINEARG]; |
{"title", HTML_NLAROUND}, |
struct htmlnode *parent; |
{"body", HTML_NLALL}, |
|
{"div", HTML_NLAROUND}, |
|
{"section", HTML_NLALL}, |
|
{"nav", HTML_NLALL}, |
|
{"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}, |
|
{"h1", HTML_TOPHRASE | HTML_NLAROUND}, |
|
{"h2", HTML_TOPHRASE | HTML_NLAROUND}, |
|
{"p", HTML_TOPHRASE | HTML_NLAROUND | HTML_INDENT}, |
|
{"pre", HTML_TOPHRASE | HTML_NLAROUND | HTML_NOINDENT}, |
|
{"a", HTML_INPHRASE | HTML_TOPHRASE}, |
|
{"b", HTML_INPHRASE | HTML_TOPHRASE}, |
|
{"cite", HTML_INPHRASE | HTML_TOPHRASE}, |
|
{"code", HTML_INPHRASE | HTML_TOPHRASE}, |
|
{"i", HTML_INPHRASE | HTML_TOPHRASE}, |
|
{"small", HTML_INPHRASE | HTML_TOPHRASE}, |
|
{"span", HTML_INPHRASE | HTML_TOPHRASE}, |
|
{"var", HTML_INPHRASE | HTML_TOPHRASE}, |
|
{"br", HTML_INPHRASE | HTML_NOSTACK | HTML_NLALL}, |
|
{"hr", HTML_INPHRASE | HTML_NOSTACK}, |
|
{"mark", HTML_INPHRASE }, |
|
{"math", HTML_INPHRASE | 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}, |
}; |
}; |
|
|
|
/* Avoid duplicate HTML id= attributes. */ |
|
|
struct htmlq { |
struct id_entry { |
struct htmlnode *last; |
int ord; /* Ordinal number of the latest occurrence. */ |
|
char id[]; /* The id= attribute without any ordinal suffix. */ |
}; |
}; |
|
static struct ohash id_unique; |
|
|
|
static void html_reset_internal(struct html *); |
|
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 *); |
|
|
static int html_loadcss(struct md_mbuf *, const char *); |
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 *); |
|
|
static int html_alloc(void **); |
|
static void html_free(void *); |
|
static ssize_t html_endtag(struct md_mbuf *, void *, |
|
const struct md_args *, |
|
enum md_ns, int); |
|
static ssize_t html_beginstring(struct md_mbuf *, |
|
const struct md_args *, |
|
const char *, size_t); |
|
static ssize_t html_beginhttp(struct md_mbuf *, |
|
const struct md_args *, |
|
const char *, size_t); |
|
static ssize_t html_endstring(struct md_mbuf *, |
|
const struct md_args *, |
|
const char *, size_t); |
|
static ssize_t html_endhttp(struct md_mbuf *, |
|
const struct md_args *, |
|
const char *, size_t); |
|
static ssize_t html_begintag(struct md_mbuf *, void *, |
|
const struct md_args *, |
|
enum md_ns, int, |
|
const int *, const char **); |
|
static int html_begin(struct md_mbuf *, |
|
const struct md_args *, |
|
const struct tm *, |
|
const char *, const char *, |
|
enum roffmsec, const char *); |
|
static int html_printargs(struct md_mbuf *, int, |
|
const char *, const int *, |
|
const char **, size_t *); |
|
static int html_end(struct md_mbuf *, |
|
const struct md_args *); |
|
static int html_blocktagname(struct md_mbuf *, |
|
const struct md_args *, int, |
|
struct htmlq *, const int *, |
|
const char **, size_t *); |
|
static int html_blocktagargs(struct md_mbuf *, |
|
const struct md_args *, int, |
|
const int *, const char **, size_t *); |
|
static int html_headtagname(struct md_mbuf *, |
|
const struct md_args *, int, |
|
struct htmlq *, const int *, |
|
const char **, size_t *); |
|
static int html_headtagargs(struct md_mbuf *, |
|
const struct md_args *, int, |
|
const int *, const char **, size_t *); |
|
static int html_bodytagname(struct md_mbuf *, |
|
const struct md_args *, |
|
int, struct htmlq *, const int *, |
|
const char **, size_t *); |
|
static int html_bodytagargs(struct md_mbuf *, |
|
const struct md_args *, int, |
|
const int *, const char **, size_t *); |
|
static int html_inlinetagname(struct md_mbuf *, |
|
const struct md_args *, int, size_t *); |
|
static int html_inlinetagargs(struct md_mbuf *, |
|
const struct md_args *, int, |
|
const int *, const char **, size_t *); |
|
static int html_Bl_bodytagname(struct md_mbuf *, |
|
struct htmlq *, const int *, |
|
const char **, size_t *); |
|
static int html_It_blocktagname(struct md_mbuf *, |
|
struct htmlq *, const int *, |
|
const char **, size_t *); |
|
static int html_It_headtagname(struct md_mbuf *, |
|
struct htmlq *, const int *, |
|
const char **, size_t *); |
|
static int html_It_bodytagname(struct md_mbuf *, |
|
struct htmlq *, const int *, |
|
const char **, size_t *); |
|
|
|
|
void * |
/* ARGSUSED */ |
html_alloc(const struct manoutput *outopts) |
static int |
|
html_It_headtagname(struct md_mbuf *mbuf, struct htmlq *q, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
{ |
struct htmlnode *n; |
struct html *h; |
int i; |
|
|
|
for (n = q->last; n; n = n->parent) |
h = mandoc_calloc(1, sizeof(struct html)); |
if (n->tok == ROFF_Bl) |
|
break; |
|
|
|
assert(n); |
h->tag = NULL; |
|
h->metac = h->metal = ESCAPE_FONTROMAN; |
|
h->style = outopts->style; |
|
if ((h->base_man1 = outopts->man) == NULL) |
|
h->base_man2 = NULL; |
|
else if ((h->base_man2 = strchr(h->base_man1, ';')) != NULL) |
|
*h->base_man2++ = '\0'; |
|
h->base_includes = outopts->includes; |
|
if (outopts->fragment) |
|
h->oflags |= HTML_FRAGMENT; |
|
if (outopts->toc) |
|
h->oflags |= HTML_TOC; |
|
|
/* LINTED */ |
mandoc_ohash_init(&id_unique, 4, offsetof(struct id_entry, id)); |
for (i = 0; ROFF_ARGMAX != n->argc[i] && |
|
i < ROFF_MAXLINEARG; i++) { |
|
switch (n->argc[i]) { |
|
case (ROFF_Ohang): |
|
return(ml_nputs(mbuf, "div", 3, res)); |
|
|
|
case (ROFF_Tag): |
return h; |
/* FALLTHROUGH */ |
|
case (ROFF_Column): |
|
return(ml_nputs(mbuf, "td", 2, res)); |
|
default: |
|
break; |
|
} |
|
} |
|
|
|
abort(); |
|
/* NOTREACHED */ |
|
} |
} |
|
|
|
static void |
/* ARGSUSED */ |
html_reset_internal(struct html *h) |
static int |
|
html_It_bodytagname(struct md_mbuf *mbuf, struct htmlq *q, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
{ |
struct htmlnode *n; |
struct tag *tag; |
int i; |
struct id_entry *entry; |
|
unsigned int slot; |
|
|
for (n = q->last; n; n = n->parent) |
while ((tag = h->tag) != NULL) { |
if (n->tok == ROFF_Bl) |
h->tag = tag->next; |
break; |
free(tag); |
|
|
assert(n); |
|
|
|
/* LINTED */ |
|
for (i = 0; ROFF_ARGMAX != n->argc[i] && |
|
i < ROFF_MAXLINEARG; i++) { |
|
switch (n->argc[i]) { |
|
case (ROFF_Enum): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Bullet): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Dash): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Hyphen): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Item): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Diag): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Hang): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Ohang): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Inset): |
|
return(ml_nputs(mbuf, "div", 3, res)); |
|
case (ROFF_Tag): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Column): |
|
return(ml_nputs(mbuf, "td", 2, res)); |
|
default: |
|
break; |
|
} |
|
} |
} |
|
entry = ohash_first(&id_unique, &slot); |
|
while (entry != NULL) { |
|
free(entry); |
|
entry = ohash_next(&id_unique, &slot); |
|
} |
|
ohash_delete(&id_unique); |
|
} |
|
|
assert(i != ROFF_MAXLINEARG); |
void |
abort(); |
html_reset(void *p) |
/* NOTREACHED */ |
{ |
|
html_reset_internal(p); |
|
mandoc_ohash_init(&id_unique, 4, offsetof(struct id_entry, id)); |
|
} |
|
|
return(1); |
void |
|
html_free(void *p) |
|
{ |
|
html_reset_internal(p); |
|
free(p); |
} |
} |
|
|
|
void |
/* ARGSUSED */ |
print_gen_head(struct html *h) |
static int |
|
html_Bl_bodytagname(struct md_mbuf *mbuf, struct htmlq *q, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
{ |
int i; |
struct tag *t; |
|
|
for (i = 0; ROFF_ARGMAX != argc[i] |
print_otag(h, TAG_META, "?", "charset", "utf-8"); |
&& i < ROFF_MAXLINEARG; i++) { |
print_otag(h, TAG_META, "??", "name", "viewport", |
switch (argc[i]) { |
"content", "width=device-width, initial-scale=1.0"); |
case (ROFF_Enum): |
if (h->style != NULL) { |
return(ml_nputs(mbuf, "ol", 2, res)); |
print_otag(h, TAG_LINK, "?h??", "rel", "stylesheet", |
case (ROFF_Bullet): |
h->style, "type", "text/css", "media", "all"); |
/* FALLTHROUGH */ |
return; |
case (ROFF_Dash): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Hyphen): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Item): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Diag): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Hang): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Ohang): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Inset): |
|
return(ml_nputs(mbuf, "ul", 2, res)); |
|
case (ROFF_Tag): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Column): |
|
return(ml_nputs(mbuf, "table", 5, res)); |
|
default: |
|
break; |
|
} |
|
} |
} |
|
|
assert(i != ROFF_MAXLINEARG); |
/* |
abort(); |
* Print a minimal embedded style sheet. |
/* NOTREACHED */ |
*/ |
|
|
|
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, ".Nd, .Bf, .Op { display: inline; }"); |
|
print_endline(h); |
|
print_text(h, ".Pa, .Ad { font-style: italic; }"); |
|
print_endline(h); |
|
print_text(h, ".Ms { font-weight: bold; }"); |
|
print_endline(h); |
|
print_text(h, ".Bl-diag "); |
|
print_byte(h, '>'); |
|
print_text(h, " dt { font-weight: bold; }"); |
|
print_endline(h); |
|
print_text(h, "code.Nm, .Fl, .Cm, .Ic, code.In, .Fd, .Fn, .Cd " |
|
"{ font-weight: bold; font-family: inherit; }"); |
|
print_tagq(h, t); |
} |
} |
|
|
|
int |
/* ARGSUSED */ |
html_setfont(struct html *h, enum mandoc_esc font) |
static int |
|
html_It_blocktagname(struct md_mbuf *mbuf, struct htmlq *q, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
{ |
struct htmlnode *n; |
switch (font) { |
int i; |
case ESCAPE_FONTPREV: |
|
font = h->metal; |
for (n = q->last; n; n = n->parent) |
break; |
if (n->tok == ROFF_Bl) |
case ESCAPE_FONTITALIC: |
break; |
case ESCAPE_FONTBOLD: |
|
case ESCAPE_FONTBI: |
assert(n); |
case ESCAPE_FONTROMAN: |
|
case ESCAPE_FONTCR: |
/* LINTED */ |
case ESCAPE_FONTCB: |
for (i = 0; ROFF_ARGMAX != n->argc[i] && |
case ESCAPE_FONTCI: |
i < ROFF_MAXLINEARG; i++) { |
break; |
switch (n->argc[i]) { |
case ESCAPE_FONT: |
case (ROFF_Enum): |
font = ESCAPE_FONTROMAN; |
/* FALLTHROUGH */ |
break; |
case (ROFF_Bullet): |
default: |
/* FALLTHROUGH */ |
return 0; |
case (ROFF_Dash): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Hyphen): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Item): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Diag): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Hang): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Ohang): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Inset): |
|
return(ml_nputs(mbuf, "li", 2, res)); |
|
case (ROFF_Tag): |
|
/* FALLTHROUGH */ |
|
case (ROFF_Column): |
|
return(ml_nputs(mbuf, "tr", 2, res)); |
|
default: |
|
break; |
|
} |
|
} |
} |
|
h->metal = h->metac; |
assert(i != ROFF_MAXLINEARG); |
h->metac = font; |
abort(); |
return 1; |
/* NOTREACHED */ |
|
} |
} |
|
|
|
static void |
static int |
print_metaf(struct html *h) |
html_loadcss(struct md_mbuf *mbuf, const char *css) |
|
{ |
{ |
size_t res, bufsz; |
if (h->metaf) { |
char *buf; |
print_tagq(h, h->metaf); |
struct stat st; |
h->metaf = NULL; |
int fd, c; |
|
ssize_t ssz; |
|
|
|
c = 0; |
|
res = 0; |
|
buf = NULL; |
|
|
|
if (-1 == (fd = open(css, O_RDONLY, 0))) { |
|
warn("%s", css); |
|
return(0); |
|
} |
|
|
|
if (-1 == fstat(fd, &st)) { |
|
warn("%s", css); |
|
goto out; |
|
} |
} |
|
switch (h->metac) { |
bufsz = MAX(st.st_blksize, BUFSIZ); |
case ESCAPE_FONTITALIC: |
if (NULL == (buf = malloc(bufsz))) { |
h->metaf = print_otag(h, TAG_I, ""); |
warn("malloc"); |
break; |
goto out; |
case ESCAPE_FONTBOLD: |
|
h->metaf = print_otag(h, TAG_B, ""); |
|
break; |
|
case ESCAPE_FONTBI: |
|
h->metaf = print_otag(h, TAG_B, ""); |
|
print_otag(h, TAG_I, ""); |
|
break; |
|
case ESCAPE_FONTCR: |
|
h->metaf = print_otag(h, TAG_SPAN, "c", "Li"); |
|
break; |
|
case ESCAPE_FONTCB: |
|
h->metaf = print_otag(h, TAG_SPAN, "c", "Li"); |
|
print_otag(h, TAG_B, ""); |
|
break; |
|
case ESCAPE_FONTCI: |
|
h->metaf = print_otag(h, TAG_SPAN, "c", "Li"); |
|
print_otag(h, TAG_I, ""); |
|
break; |
|
default: |
|
break; |
} |
} |
|
} |
|
|
|
void |
|
html_close_paragraph(struct html *h) |
|
{ |
|
struct tag *this, *next; |
|
int flags; |
|
|
|
this = h->tag; |
for (;;) { |
for (;;) { |
if (-1 == (ssz = read(fd, buf, bufsz))) { |
next = this->next; |
warn("%s", css); |
flags = htmltags[this->tag].flags; |
goto out; |
if (flags & (HTML_INPHRASE | HTML_TOPHRASE)) |
} else if (0 == ssz) |
print_ctag(h, this); |
|
if ((flags & HTML_INPHRASE) == 0) |
break; |
break; |
if ( ! ml_nputs(mbuf, buf, (size_t)ssz, &res)) |
this = next; |
goto out; |
|
} |
} |
|
} |
|
|
c = 1; |
/* |
|
* ROFF_nf switches to no-fill mode, ROFF_fi to fill mode. |
|
* TOKEN_NONE does not switch. The old mode is returned. |
|
*/ |
|
enum roff_tok |
|
html_fillmode(struct html *h, enum roff_tok want) |
|
{ |
|
struct tag *t; |
|
enum roff_tok had; |
|
|
out: |
for (t = h->tag; t != NULL; t = t->next) |
if (-1 == close(fd)) { |
if (t->tag == TAG_PRE) |
warn("%s", css); |
break; |
c = 0; |
|
} |
|
|
|
if (buf) |
had = t == NULL ? ROFF_fi : ROFF_nf; |
free(buf); |
|
|
|
return(c); |
if (want != had) { |
|
switch (want) { |
|
case ROFF_fi: |
|
print_tagq(h, t); |
|
break; |
|
case ROFF_nf: |
|
html_close_paragraph(h); |
|
print_otag(h, TAG_PRE, ""); |
|
break; |
|
case TOKEN_NONE: |
|
break; |
|
default: |
|
abort(); |
|
} |
|
} |
|
return had; |
} |
} |
|
|
|
/* |
/* ARGSUSED */ |
* Allocate a string to be used for the "id=" attribute of an HTML |
static int |
* element and/or as a segment identifier for a URI in an <a> element. |
html_begin(struct md_mbuf *mbuf, const struct md_args *args, |
* The function may fail and return NULL if the node lacks text data |
const struct tm *tm, const char *os, |
* to create the attribute from. |
const char *title, enum roffmsec section, |
* The caller is responsible for free(3)ing the returned string. |
const char *vol) |
* |
|
* If the "unique" argument is non-zero, the "id_unique" ohash table |
|
* is used for de-duplication. If the "unique" argument is 1, |
|
* it is the first time the function is called for this tag and |
|
* location, so if an ordinal suffix is needed, it is incremented. |
|
* If the "unique" argument is 2, it is the second time the function |
|
* is called for this tag and location, so the ordinal suffix |
|
* remains unchanged. |
|
*/ |
|
char * |
|
html_make_id(const struct roff_node *n, int unique) |
{ |
{ |
const char *preamble, *css, *trail; |
const struct roff_node *nch; |
char buf[512]; |
struct id_entry *entry; |
size_t res; |
char *buf, *cp; |
|
size_t len; |
|
unsigned int slot; |
|
|
preamble = |
if (n->tag != NULL) |
"<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01//EN\"\n" |
buf = mandoc_strdup(n->tag); |
" \"http://www.w3.org/TR/html4/strict.dtd\">\n" |
else { |
"<html>\n" |
switch (n->tok) { |
"<head>\n" |
case MDOC_Sh: |
" <meta http-equiv=\"Content-Type\"\n" |
case MDOC_Ss: |
" content=\"text/html;charset=utf-8\">\n" |
case MDOC_Sx: |
" <meta name=\"resource-type\" content=\"document\">\n" |
case MAN_SH: |
" <title>Manual Page for %s(%s)</title>\n"; |
case MAN_SS: |
|
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; |
|
break; |
|
default: |
|
if (n->child == NULL || n->child->type != ROFFT_TEXT) |
|
return NULL; |
|
buf = mandoc_strdup(n->child->string); |
|
break; |
|
} |
|
} |
|
|
css = |
/* |
" <link rel=\"stylesheet\" type=\"text/css\"\n" |
* In ID attributes, only use ASCII characters that are |
" href=\"%s\">\n"; |
* permitted in URL-fragment strings according to the |
trail = |
* explicit list at: |
"</head>\n" |
* https://url.spec.whatwg.org/#url-fragment-string |
"<body>\n" |
* In addition, reserve '~' for ordinal suffixes. |
"<div class=\"mdoc\">"; |
*/ |
|
|
res = 0; |
for (cp = buf; *cp != '\0'; cp++) |
|
if (isalnum((unsigned char)*cp) == 0 && |
|
strchr("!$&'()*+,-./:;=?@_", *cp) == NULL) |
|
*cp = '_'; |
|
|
(void)snprintf(buf, sizeof(buf) - 1, |
if (unique == 0) |
preamble, title, ml_section(section)); |
return buf; |
|
|
if ( ! ml_puts(mbuf, buf, &res)) |
/* Avoid duplicate HTML id= attributes. */ |
return(0); |
|
|
|
assert(args->params.html.css); |
slot = ohash_qlookup(&id_unique, buf); |
if (HTML_CSS_EMBED & args->params.html.flags) { |
if ((entry = ohash_find(&id_unique, slot)) == NULL) { |
if ( ! ml_puts(mbuf, " <style type=\"text/css\"><!--\n", &res)) |
len = strlen(buf) + 1; |
return(0); |
entry = mandoc_malloc(sizeof(*entry) + len); |
if ( ! html_loadcss(mbuf, args->params.html.css)) |
entry->ord = 1; |
return(0); |
memcpy(entry->id, buf, len); |
if ( ! ml_puts(mbuf, " --!></style>\n", &res)) |
ohash_insert(&id_unique, slot, entry); |
return(0); |
} else if (unique == 1) |
} else { |
entry->ord++; |
(void)snprintf(buf, sizeof(buf) - 1, css, |
|
args->params.html.css); |
|
if ( ! ml_puts(mbuf, buf, &res)) |
|
return(0); |
|
} |
|
|
|
if ( ! ml_puts(mbuf, trail, &res)) |
if (entry->ord > 1) { |
return(0); |
cp = buf; |
|
mandoc_asprintf(&buf, "%s~%d", cp, entry->ord); |
return(1); |
free(cp); |
|
} |
|
return buf; |
} |
} |
|
|
|
|
/* ARGSUSED */ |
|
static int |
|
html_end(struct md_mbuf *mbuf, const struct md_args *args) |
|
{ |
|
|
|
return(ml_puts(mbuf, "</div></body>\n</html>", NULL)); |
|
} |
|
|
|
|
|
/* ARGSUSED */ |
|
static int |
static int |
html_bodytagname(struct md_mbuf *mbuf, |
print_escape(struct html *h, char c) |
const struct md_args *args, int tok, struct htmlq *q, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
{ |
|
|
switch (tok) { |
switch (c) { |
case (ROFF_Bl): |
case '<': |
return(html_Bl_bodytagname(mbuf, q, argc, argv, res)); |
print_word(h, "<"); |
case (ROFF_Fo): |
|
return(ml_nputs(mbuf, "span", 4, res)); |
|
case (ROFF_It): |
|
return(html_It_bodytagname(mbuf, q, argc, argv, res)); |
|
case (ROFF_Oo): |
|
return(ml_nputs(mbuf, "span", 4, res)); |
|
default: |
|
break; |
break; |
|
case '>': |
|
print_word(h, ">"); |
|
break; |
|
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; |
} |
} |
|
return 1; |
return(ml_puts(mbuf, "div", res)); |
|
} |
} |
|
|
|
|
/* ARGSUSED */ |
|
static int |
static int |
html_headtagname(struct md_mbuf *mbuf, |
print_encode(struct html *h, const char *p, const char *pend, int norecurse) |
const struct md_args *args, int tok, struct htmlq *q, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
{ |
|
char numbuf[16]; |
|
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' }; |
|
|
switch (tok) { |
if (pend == NULL) |
case (ROFF_It): |
pend = strchr(p, '\0'); |
return(html_It_headtagname(mbuf, q, argc, argv, res)); |
|
case (ROFF_Fo): |
|
return(ml_nputs(mbuf, "span", 4, res)); |
|
case (ROFF_Oo): |
|
return(ml_nputs(mbuf, "span", 4, res)); |
|
case (ROFF_Sh): |
|
return(ml_nputs(mbuf, "h1", 2, res)); |
|
case (ROFF_Ss): |
|
return(ml_nputs(mbuf, "h2", 2, res)); |
|
default: |
|
break; |
|
} |
|
|
|
return(ml_nputs(mbuf, "div", 3, res)); |
breakline = 0; |
} |
nospace = 0; |
|
|
|
while (p < pend) { |
|
if (HTML_SKIPCHAR & h->flags && '\\' != *p) { |
|
h->flags &= ~HTML_SKIPCHAR; |
|
p++; |
|
continue; |
|
} |
|
|
/* ARGSUSED */ |
for (sz = strcspn(p, rejs); sz-- && p < pend; p++) |
static int |
print_byte(h, *p); |
html_blocktagname(struct md_mbuf *mbuf, const struct md_args *args, |
|
int tok, struct htmlq *q, const int *argc, |
|
const char **argv, size_t *res) |
|
{ |
|
|
|
switch (tok) { |
if (breakline && |
case (ROFF_Fo): |
(p >= pend || *p == ' ' || *p == ASCII_NBRSP)) { |
return(ml_nputs(mbuf, "span", 4, res)); |
print_otag(h, TAG_BR, ""); |
case (ROFF_Oo): |
breakline = 0; |
return(ml_nputs(mbuf, "span", 4, res)); |
while (p < pend && (*p == ' ' || *p == ASCII_NBRSP)) |
case (ROFF_It): |
p++; |
return(html_It_blocktagname(mbuf, q, argc, argv, res)); |
continue; |
default: |
} |
break; |
|
} |
|
|
|
return(ml_puts(mbuf, "div", res)); |
if (p >= pend) |
} |
break; |
|
|
|
if (*p == ' ') { |
|
print_endword(h); |
|
p++; |
|
continue; |
|
} |
|
|
/* ARGSUSED */ |
if (print_escape(h, *p++)) |
static int |
continue; |
html_printargs(struct md_mbuf *mbuf, int tok, const char *ns, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
|
|
|
if ( ! ml_puts(mbuf, " class=\"", res)) |
esc = mandoc_escape(&p, &seq, &len); |
return(0); |
switch (esc) { |
if ( ! ml_puts(mbuf, ns, res)) |
case ESCAPE_FONT: |
return(0); |
case ESCAPE_FONTPREV: |
if ( ! ml_puts(mbuf, "-", res)) |
case ESCAPE_FONTBOLD: |
return(0); |
case ESCAPE_FONTITALIC: |
if ( ! ml_puts(mbuf, toknames[tok], res)) |
case ESCAPE_FONTBI: |
return(0); |
case ESCAPE_FONTROMAN: |
return(ml_puts(mbuf, "\"", res)); |
case ESCAPE_FONTCR: |
} |
case ESCAPE_FONTCB: |
|
case ESCAPE_FONTCI: |
|
if (0 == norecurse) { |
|
h->flags |= HTML_NOSPACE; |
|
if (html_setfont(h, esc)) |
|
print_metaf(h); |
|
h->flags &= ~HTML_NOSPACE; |
|
} |
|
continue; |
|
case ESCAPE_SKIPCHAR: |
|
h->flags |= HTML_SKIPCHAR; |
|
continue; |
|
case ESCAPE_ERROR: |
|
continue; |
|
default: |
|
break; |
|
} |
|
|
|
if (h->flags & HTML_SKIPCHAR) { |
|
h->flags &= ~HTML_SKIPCHAR; |
|
continue; |
|
} |
|
|
/* ARGSUSED */ |
switch (esc) { |
static int |
case ESCAPE_UNICODE: |
html_headtagargs(struct md_mbuf *mbuf, |
/* Skip past "u" header. */ |
const struct md_args *args, int tok, |
c = mchars_num2uc(seq + 1, len - 1); |
const int *argc, const char **argv, size_t *res) |
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_UNDEF: |
|
c = *seq; |
|
break; |
|
case ESCAPE_DEVICE: |
|
print_word(h, "html"); |
|
continue; |
|
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(html_printargs(mbuf, tok, "head", argc, argv, res)); |
return nospace; |
} |
} |
|
|
|
static void |
/* ARGSUSED */ |
print_href(struct html *h, const char *name, const char *sec, int man) |
static int |
|
html_bodytagargs(struct md_mbuf *mbuf, |
|
const struct md_args *args, int tok, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
{ |
|
struct stat sb; |
|
const char *p, *pp; |
|
char *filename; |
|
|
return(html_printargs(mbuf, tok, "body", argc, argv, res)); |
if (man) { |
|
pp = h->base_man1; |
|
if (h->base_man2 != NULL) { |
|
mandoc_asprintf(&filename, "%s.%s", name, sec); |
|
if (stat(filename, &sb) == -1) |
|
pp = h->base_man2; |
|
free(filename); |
|
} |
|
} else |
|
pp = 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); |
} |
} |
|
|
|
struct tag * |
/* ARGSUSED */ |
print_otag(struct html *h, enum htmltag tag, const char *fmt, ...) |
static int |
|
html_blocktagargs(struct md_mbuf *mbuf, |
|
const struct md_args *args, int tok, |
|
const int *argc, const char **argv, size_t *res) |
|
{ |
{ |
|
va_list ap; |
|
struct tag *t; |
|
const char *attr; |
|
char *arg1, *arg2; |
|
int style_written, tflags; |
|
|
return(html_printargs(mbuf, tok, "block", argc, argv, res)); |
tflags = htmltags[tag].flags; |
} |
|
|
|
|
/* Flow content is not allowed in phrasing context. */ |
|
|
/* ARGSUSED */ |
if ((tflags & HTML_INPHRASE) == 0) { |
static int |
for (t = h->tag; t != NULL; t = t->next) { |
html_inlinetagargs(struct md_mbuf *mbuf, |
if (t->closed) |
const struct md_args *args, int tok, |
continue; |
const int *argc, const char **argv, size_t *res) |
assert((htmltags[t->tag].flags & HTML_TOPHRASE) == 0); |
{ |
break; |
|
} |
|
|
if ( ! html_printargs(mbuf, tok, "inline", argc, argv, res)) |
/* |
return(0); |
* Always wrap phrasing elements in a paragraph |
|
* unless already contained in some flow container; |
|
* never put them directly into a section. |
|
*/ |
|
|
switch (tok) { |
} else if (tflags & HTML_TOPHRASE && h->tag->tag == TAG_SECTION) |
case (ROFF_Sx): |
print_otag(h, TAG_P, "c", "Pp"); |
assert(*argv); |
|
if ( ! ml_nputs(mbuf, " href=\"#", 8, res)) |
|
return(0); |
|
if ( ! ml_putstring(mbuf, *argv, res)) |
|
return(0); |
|
if ( ! ml_nputs(mbuf, "\"", 1, res)) |
|
return(0); |
|
break; |
|
default: |
|
break; |
|
} |
|
|
|
return(1); |
|
} |
|
|
|
|
/* Push this tag onto the stack of open scopes. */ |
|
|
/* ARGSUSED */ |
if ((tflags & HTML_NOSTACK) == 0) { |
static int |
t = mandoc_malloc(sizeof(struct tag)); |
html_inlinetagname(struct md_mbuf *mbuf, |
t->tag = tag; |
const struct md_args *args, int tok, size_t *res) |
t->next = h->tag; |
{ |
t->refcnt = 0; |
|
t->closed = 0; |
|
h->tag = t; |
|
} else |
|
t = NULL; |
|
|
switch (tok) { |
if (tflags & HTML_NLBEFORE) |
case (ROFF_Pp): |
print_endline(h); |
return(ml_nputs(mbuf, "div", 3, res)); |
if (h->col == 0) |
case (ROFF_Sx): |
print_indent(h); |
return(ml_nputs(mbuf, "a", 1, res)); |
else if ((h->flags & HTML_NOSPACE) == 0) { |
default: |
if (h->flags & HTML_KEEP) |
break; |
print_word(h, " "); |
|
else { |
|
if (h->flags & HTML_PREKEEP) |
|
h->flags |= HTML_KEEP; |
|
print_endword(h); |
|
} |
} |
} |
|
|
return(ml_puts(mbuf, "span", res)); |
if ( ! (h->flags & HTML_NONOSPACE)) |
} |
h->flags &= ~HTML_NOSPACE; |
|
else |
|
h->flags |= HTML_NOSPACE; |
|
|
|
/* Print out the tag name and attributes. */ |
|
|
static ssize_t |
print_byte(h, '<'); |
html_begintag(struct md_mbuf *mbuf, void *data, |
print_word(h, htmltags[tag].name); |
const struct md_args *args, enum md_ns ns, |
|
int tok, const int *argc, const char **argv) |
|
{ |
|
size_t res; |
|
struct htmlq *q; |
|
struct htmlnode *node; |
|
int i; |
|
|
|
assert(ns != MD_NS_DEFAULT); |
va_start(ap, fmt); |
res = 0; |
|
|
|
assert(data); |
while (*fmt != '\0' && *fmt != 's') { |
q = (struct htmlq *)data; |
|
|
|
if (NULL == (node = calloc(1, sizeof(struct htmlnode)))) { |
/* Parse attributes and arguments. */ |
warn("calloc"); |
|
return(-1); |
|
} |
|
|
|
node->parent = q->last; |
arg1 = va_arg(ap, char *); |
node->tok = tok; |
arg2 = NULL; |
node->ns = ns; |
switch (*fmt++) { |
|
case 'c': |
|
attr = "class"; |
|
break; |
|
case 'h': |
|
attr = "href"; |
|
break; |
|
case 'i': |
|
attr = "id"; |
|
break; |
|
case 'r': |
|
attr = "role"; |
|
break; |
|
case '?': |
|
attr = arg1; |
|
arg1 = va_arg(ap, char *); |
|
break; |
|
default: |
|
abort(); |
|
} |
|
if (*fmt == 'M') |
|
arg2 = va_arg(ap, char *); |
|
if (arg1 == NULL) |
|
continue; |
|
|
if (argc) { |
/* Print the attributes. */ |
/* TODO: argv. */ |
|
|
|
assert(argv); |
print_byte(h, ' '); |
/* LINTED */ |
print_word(h, attr); |
for (i = 0; ROFF_ARGMAX != argc[i] |
print_byte(h, '='); |
&& i < ROFF_MAXLINEARG; i++) |
print_byte(h, '"'); |
node->argc[i] = argc[i]; |
switch (*fmt) { |
assert(i != ROFF_MAXLINEARG); |
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; |
|
default: |
|
print_encode(h, arg1, NULL, 1); |
|
break; |
|
} |
|
print_byte(h, '"'); |
|
} |
|
|
|
style_written = 0; |
|
while (*fmt++ == 's') { |
|
arg1 = va_arg(ap, char *); |
|
arg2 = va_arg(ap, char *); |
|
if (arg2 == NULL) |
|
continue; |
|
print_byte(h, ' '); |
|
if (style_written == 0) { |
|
print_word(h, "style=\""); |
|
style_written = 1; |
|
} |
|
print_word(h, arg1); |
|
print_byte(h, ':'); |
|
print_byte(h, ' '); |
|
print_word(h, arg2); |
|
print_byte(h, ';'); |
|
} |
|
if (style_written) |
|
print_byte(h, '"'); |
|
|
q->last = node; |
va_end(ap); |
|
|
switch (ns) { |
/* Accommodate for "well-formed" singleton escaping. */ |
case (MD_NS_BLOCK): |
|
if ( ! html_blocktagname(mbuf, args, tok, |
|
q, argc, argv, &res)) |
|
return(-1); |
|
if ( ! html_blocktagargs(mbuf, args, tok, |
|
argc, argv, &res)) |
|
return(-1); |
|
break; |
|
case (MD_NS_BODY): |
|
if ( ! html_bodytagname(mbuf, args, tok, |
|
q, argc, argv, &res)) |
|
return(-1); |
|
if ( ! html_bodytagargs(mbuf, args, tok, |
|
argc, argv, &res)) |
|
return(-1); |
|
break; |
|
case (MD_NS_HEAD): |
|
if ( ! html_headtagname(mbuf, args, tok, q, |
|
argc, argv, &res)) |
|
return(-1); |
|
if ( ! html_headtagargs(mbuf, args, tok, |
|
argc, argv, &res)) |
|
return(-1); |
|
break; |
|
default: |
|
if ( ! html_inlinetagname(mbuf, args, tok, &res)) |
|
return(-1); |
|
if ( ! html_inlinetagargs(mbuf, args, tok, |
|
argc, argv, &res)) |
|
return(-1); |
|
break; |
|
} |
|
|
|
return((ssize_t)res); |
if (htmltags[tag].flags & HTML_NOSTACK) |
} |
print_byte(h, '/'); |
|
|
|
print_byte(h, '>'); |
|
|
static ssize_t |
if (tflags & HTML_NLBEGIN) |
html_endtag(struct md_mbuf *mbuf, void *data, |
print_endline(h); |
const struct md_args *args, enum md_ns ns, int tok) |
else |
{ |
h->flags |= HTML_NOSPACE; |
size_t res; |
|
struct htmlq *q; |
|
struct htmlnode *node; |
|
|
|
assert(ns != MD_NS_DEFAULT); |
if (tflags & HTML_INDENT) |
res = 0; |
h->indent++; |
|
if (tflags & HTML_NOINDENT) |
|
h->noindent++; |
|
|
assert(data); |
return t; |
q = (struct htmlq *)data; |
} |
node = q->last; |
|
|
|
switch (ns) { |
/* |
case (MD_NS_BLOCK): |
* Print an element with an optional "id=" attribute. |
if ( ! html_blocktagname(mbuf, args, tok, |
* If the element has phrasing content and an "id=" attribute, |
q, node->argc, |
* also add a permalink: outside if it can be in phrasing context, |
(const char **)node->argv, &res)) |
* inside otherwise. |
return(-1); |
*/ |
break; |
struct tag * |
case (MD_NS_BODY): |
print_otag_id(struct html *h, enum htmltag elemtype, const char *cattr, |
if ( ! html_bodytagname(mbuf, args, tok, |
struct roff_node *n) |
q, node->argc, |
{ |
(const char **)node->argv, &res)) |
struct roff_node *nch; |
return(-1); |
struct tag *ret, *t; |
break; |
char *id, *href; |
case (MD_NS_HEAD): |
|
if ( ! html_headtagname(mbuf, args, tok, |
ret = NULL; |
q, node->argc, |
id = href = NULL; |
(const char **)node->argv, &res)) |
if (n->flags & NODE_ID) |
return(-1); |
id = html_make_id(n, 1); |
break; |
if (n->flags & NODE_HREF) |
default: |
href = id == NULL ? html_make_id(n, 2) : id; |
if ( ! html_inlinetagname(mbuf, args, tok, &res)) |
if (href != NULL && htmltags[elemtype].flags & HTML_INPHRASE) |
return(-1); |
ret = print_otag(h, TAG_A, "chR", "permalink", href); |
break; |
t = print_otag(h, elemtype, "ci", cattr, id); |
|
if (ret == NULL) { |
|
ret = t; |
|
if (href != NULL && (nch = n->child) != NULL) { |
|
/* man(7) is safe, it tags phrasing content only. */ |
|
if (n->tok > MDOC_MAX || |
|
htmltags[elemtype].flags & HTML_TOPHRASE) |
|
nch = NULL; |
|
else /* For mdoc(7), beware of nested blocks. */ |
|
while (nch != NULL && nch->type == ROFFT_TEXT) |
|
nch = nch->next; |
|
if (nch == NULL) |
|
print_otag(h, TAG_A, "chR", "permalink", href); |
|
} |
} |
} |
|
free(id); |
|
if (id == NULL) |
|
free(href); |
|
return ret; |
|
} |
|
|
q->last = node->parent; |
static void |
|
print_ctag(struct html *h, struct tag *tag) |
|
{ |
|
int tflags; |
|
|
free(node); |
if (tag->closed == 0) { |
|
tag->closed = 1; |
|
if (tag == h->metaf) |
|
h->metaf = NULL; |
|
if (tag == h->tblt) |
|
h->tblt = NULL; |
|
|
return((ssize_t)res); |
tflags = htmltags[tag->tag].flags; |
|
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); |
|
} |
|
if (tag->refcnt == 0) { |
|
h->tag = tag->next; |
|
free(tag); |
|
} |
} |
} |
|
|
|
void |
|
print_gen_decls(struct html *h) |
|
{ |
|
print_word(h, "<!DOCTYPE html>"); |
|
print_endline(h); |
|
} |
|
|
static int |
void |
html_alloc(void **p) |
print_gen_comment(struct html *h, struct roff_node *n) |
{ |
{ |
|
int wantblank; |
|
|
if (NULL == (*p = calloc(1, sizeof(struct htmlq)))) { |
print_word(h, "<!-- This is an automatically generated file." |
warn("calloc"); |
" Do not edit."); |
return(0); |
h->indent = 1; |
|
wantblank = 0; |
|
while (n != NULL && n->type == ROFFT_COMMENT) { |
|
if (strstr(n->string, "-->") == NULL && |
|
(wantblank || *n->string != '\0')) { |
|
print_endline(h); |
|
print_indent(h); |
|
print_word(h, n->string); |
|
wantblank = *n->string != '\0'; |
|
} |
|
n = n->next; |
} |
} |
return(1); |
if (wantblank) |
|
print_endline(h); |
|
print_word(h, " -->"); |
|
print_endline(h); |
|
h->indent = 0; |
} |
} |
|
|
|
void |
|
print_text(struct html *h, const char *word) |
|
{ |
|
print_tagged_text(h, word, NULL); |
|
} |
|
|
static void |
void |
html_free(void *p) |
print_tagged_text(struct html *h, const char *word, struct roff_node *n) |
{ |
{ |
struct htmlq *q; |
struct tag *t; |
struct htmlnode *n; |
char *href; |
|
|
assert(p); |
/* |
q = (struct htmlq *)p; |
* Always wrap text in a paragraph unless already contained in |
|
* some flow container; never put it directly into a section. |
|
*/ |
|
|
/* LINTED */ |
if (h->tag->tag == TAG_SECTION) |
while ((n = q->last)) { |
print_otag(h, TAG_P, "c", "Pp"); |
q->last = n->parent; |
|
free(n); |
/* Output whitespace before this text? */ |
|
|
|
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, " "); |
} |
} |
|
|
free(q); |
/* |
} |
* Optionally switch fonts, optionally write a permalink, then |
|
* print the text, optionally surrounded by HTML whitespace. |
|
*/ |
|
|
|
assert(h->metaf == NULL); |
|
print_metaf(h); |
|
print_indent(h); |
|
|
static ssize_t |
if (n != NULL && (href = html_make_id(n, 2)) != NULL) { |
html_beginhttp(struct md_mbuf *mbuf, |
t = print_otag(h, TAG_A, "chR", "permalink", href); |
const struct md_args *args, |
free(href); |
const char *buf, size_t sz) |
} else |
{ |
t = NULL; |
size_t res; |
|
|
|
res = 0; |
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 ( ! ml_puts(mbuf, "<a href=\"", &res)) |
if (h->metaf != NULL) { |
return(-1); |
print_tagq(h, h->metaf); |
if (1 != ml_nputstring(mbuf, buf, sz, &res)) |
h->metaf = NULL; |
return(-1); |
} else if (t != NULL) |
if ( ! ml_puts(mbuf, "\">", &res)) |
print_tagq(h, t); |
return(-1); |
|
|
|
return((ssize_t)res); |
h->flags &= ~HTML_IGNDELIM; |
} |
} |
|
|
|
void |
static ssize_t |
print_tagq(struct html *h, const struct tag *until) |
html_endhttp(struct md_mbuf *mbuf, |
|
const struct md_args *args, |
|
const char *buf, size_t sz) |
|
{ |
{ |
size_t res; |
struct tag *this, *next; |
|
|
res = 0; |
for (this = h->tag; this != NULL; this = next) { |
|
next = this == until ? NULL : this->next; |
|
print_ctag(h, this); |
|
} |
|
} |
|
|
if ( ! ml_puts(mbuf, "</a>", &res)) |
/* |
return(-1); |
* Close out all open elements up to but excluding suntil. |
|
* Note that a paragraph just inside stays open together with it |
|
* because paragraphs include subsequent phrasing content. |
|
*/ |
|
void |
|
print_stagq(struct html *h, const struct tag *suntil) |
|
{ |
|
struct tag *this, *next; |
|
|
return((ssize_t)res); |
for (this = h->tag; this != NULL; this = next) { |
|
next = this->next; |
|
if (this == suntil || (next == suntil && |
|
(this->tag == TAG_P || this->tag == TAG_PRE))) |
|
break; |
|
print_ctag(h, this); |
|
} |
} |
} |
|
|
|
|
/* ARGSUSED */ |
/*********************************************************************** |
static ssize_t |
* Low level output functions. |
html_beginstring(struct md_mbuf *mbuf, |
* They implement line breaking using a short static buffer. |
const struct md_args *args, |
***********************************************************************/ |
const char *buf, size_t sz) |
|
|
/* |
|
* 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 |
|
print_byte(struct html *h, char c) |
{ |
{ |
|
if ((h->flags & HTML_BUFFER) == 0) { |
|
putchar(c); |
|
h->col++; |
|
return; |
|
} |
|
|
if (0 == strncmp(buf, "http://", 7)) |
if (h->col + h->bufcol < sizeof(h->buf)) { |
return(html_beginhttp(mbuf, args, buf, sz)); |
h->buf[h->bufcol++] = c; |
|
return; |
|
} |
|
|
return(0); |
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; |
} |
} |
|
|
|
/* |
/* ARGSUSED */ |
* If something was printed on the current output line, end it. |
static ssize_t |
* Not to be called right after print_indent(). |
html_endstring(struct md_mbuf *mbuf, |
*/ |
const struct md_args *args, |
void |
const char *buf, size_t sz) |
print_endline(struct html *h) |
{ |
{ |
|
if (h->col == 0) |
if (0 == strncmp(buf, "http://", 7)) |
return; |
return(html_endhttp(mbuf, args, buf, sz)); |
|
|
|
return(0); |
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; |
} |
} |
|
|
|
/* |
int |
* Flush the HTML output buffer. |
md_line_html(void *data, char *buf) |
* If it is inactive, activate it. |
|
*/ |
|
static void |
|
print_endword(struct html *h) |
{ |
{ |
|
if (h->noindent) { |
|
print_byte(h, ' '); |
|
return; |
|
} |
|
|
return(mlg_line((struct md_mlg *)data, buf)); |
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; |
} |
} |
|
|
|
/* |
int |
* If at the beginning of a new output line, |
md_exit_html(void *data, int flush) |
* 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) |
{ |
{ |
|
size_t i; |
|
|
return(mlg_exit((struct md_mlg *)data, flush)); |
if (h->col || h->noindent) |
|
return; |
|
|
|
h->col = h->indent * 2; |
|
for (i = 0; i < h->col; i++) |
|
putchar(' '); |
} |
} |
|
|
|
/* |
void * |
* Print or buffer some characters |
md_init_html(const struct md_args *args, |
* depending on the current HTML output buffer state. |
struct md_mbuf *mbuf, const struct md_rbuf *rbuf) |
*/ |
|
static void |
|
print_word(struct html *h, const char *cp) |
{ |
{ |
struct ml_cbs cbs; |
while (*cp != '\0') |
|
print_byte(h, *cp++); |
cbs.ml_alloc = html_alloc; |
|
cbs.ml_free = html_free; |
|
cbs.ml_begintag = html_begintag; |
|
cbs.ml_endtag = html_endtag; |
|
cbs.ml_begin = html_begin; |
|
cbs.ml_end = html_end; |
|
cbs.ml_beginstring = html_beginstring; |
|
cbs.ml_endstring = html_endstring; |
|
|
|
return(mlg_alloc(args, rbuf, mbuf, &cbs)); |
|
} |
} |