-/* $Id: html.c,v 1.93 2010/01/29 14:39:37 kristaps Exp $ */
+/* $Id: html.c,v 1.112 2010/09/04 20:18:53 kristaps Exp $ */
/*
- * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
+ * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
*
* Permission to use, copy, modify, and distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
#include <string.h>
#include <unistd.h>
+#include "mandoc.h"
#include "out.h"
#include "chars.h"
#include "html.h"
#include "main.h"
-#define UNCONST(a) ((void *)(uintptr_t)(const void *)(a))
-
struct htmldata {
const char *name;
int flags;
{"h1", 0}, /* TAG_H1 */
{"h2", 0}, /* TAG_H2 */
{"span", 0}, /* TAG_SPAN */
- {"link", HTML_CLRLINE | HTML_NOSTACK}, /* TAG_LINK */
+ {"link", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_LINK */
{"br", HTML_CLRLINE | HTML_NOSTACK | HTML_AUTOCLOSE}, /* TAG_BR */
{"a", 0}, /* TAG_A */
{"table", HTML_CLRLINE}, /* TAG_TABLE */
"summary",
};
-static void print_spec(struct html *, const char *, size_t);
+static void print_spec(struct html *, enum roffdeco,
+ const char *, size_t);
static void print_res(struct html *, const char *, size_t);
static void print_ctag(struct html *, enum htmltag);
static void print_doctype(struct html *);
static void print_xmltype(struct html *);
static int print_encode(struct html *, const char *, int);
static void print_metaf(struct html *, enum roffdeco);
+static void print_attr(struct html *,
+ const char *, const char *);
static void *ml_alloc(char *, enum htmltype);
h = calloc(1, sizeof(struct html));
if (NULL == h) {
perror(NULL);
- exit(EXIT_FAILURE);
+ exit((int)MANDOCLEVEL_SYSERR);
}
h->type = type;
static void
-print_spec(struct html *h, const char *p, size_t len)
+print_spec(struct html *h, enum roffdeco d, const char *p, size_t len)
{
+ int cp;
const char *rhs;
size_t sz;
- rhs = chars_a2ascii(h->symtab, p, len, &sz);
-
- if (NULL == rhs)
+ if ((cp = chars_spec2cp(h->symtab, p, len)) > 0) {
+ printf("&#%d;", cp);
+ return;
+ } else if (-1 == cp && DECO_SSPECIAL == d) {
+ fwrite(p, 1, len, stdout);
return;
- fwrite(rhs, 1, sz, stdout);
+ } else if (-1 == cp)
+ return;
+
+ if (NULL != (rhs = chars_spec2str(h->symtab, p, len, &sz)))
+ fwrite(rhs, 1, sz, stdout);
}
static void
print_res(struct html *h, const char *p, size_t len)
{
+ int cp;
const char *rhs;
size_t sz;
- rhs = chars_a2res(h->symtab, p, len, &sz);
-
- if (NULL == rhs)
+ if ((cp = chars_res2cp(h->symtab, p, len)) > 0) {
+ printf("&#%d;", cp);
return;
- fwrite(rhs, 1, sz, stdout);
+ } else if (-1 == cp)
+ return;
+
+ if (NULL != (rhs = chars_res2str(h->symtab, p, len, &sz)))
+ fwrite(rhs, 1, sz, stdout);
}
int len, nospace;
const char *seq;
enum roffdeco deco;
+ static const char rejs[6] = { '\\', '<', '>', '&', ASCII_HYPH, '\0' };
nospace = 0;
for (; *p; p++) {
- sz = strcspn(p, "\\<>&");
+ sz = strcspn(p, rejs);
fwrite(p, 1, sz, stdout);
p += /* LINTED */
} else if ('&' == *p) {
printf("&");
continue;
+ } else if (ASCII_HYPH == *p) {
+ /*
+ * Note: "soft hyphens" aren't graphically
+ * displayed when not breaking the text; we want
+ * them to be displayed.
+ */
+ /*printf("­");*/
+ putchar('-');
+ continue;
} else if ('\0' == *p)
break;
case (DECO_RESERVED):
print_res(h, seq, sz);
break;
+ case (DECO_SSPECIAL):
+ /* FALLTHROUGH */
case (DECO_SPECIAL):
- print_spec(h, seq, sz);
+ print_spec(h, deco, seq, sz);
break;
case (DECO_PREVIOUS):
/* FALLTHROUGH */
}
+static void
+print_attr(struct html *h, const char *key, const char *val)
+{
+ printf(" %s=\"", key);
+ (void)print_encode(h, val, 1);
+ putchar('\"');
+}
+
+
struct tag *
print_otag(struct html *h, enum htmltag tag,
int sz, const struct htmlpair *p)
int i;
struct tag *t;
+ /* Push this tags onto the stack of open scopes. */
+
if ( ! (HTML_NOSTACK & htmltags[tag].flags)) {
t = malloc(sizeof(struct tag));
if (NULL == t) {
perror(NULL);
- exit(EXIT_FAILURE);
+ exit((int)MANDOCLEVEL_SYSERR);
}
t->tag = tag;
t->next = h->tags.head;
t = NULL;
if ( ! (HTML_NOSPACE & h->flags))
- if ( ! (HTML_CLRLINE & htmltags[tag].flags))
- putchar(' ');
+ if ( ! (HTML_CLRLINE & htmltags[tag].flags)) {
+ /* Manage keeps! */
+ if ( ! (HTML_KEEP & h->flags)) {
+ if (HTML_PREKEEP & h->flags)
+ h->flags |= HTML_KEEP;
+ putchar(' ');
+ } else
+ printf(" ");
+ }
+
+ if ( ! (h->flags & HTML_NONOSPACE))
+ h->flags &= ~HTML_NOSPACE;
+ else
+ h->flags |= HTML_NOSPACE;
+
+ /* Print out the tag name and attributes. */
printf("<%s", htmltags[tag].name);
- for (i = 0; i < sz; i++) {
- printf(" %s=\"", htmlattrs[p[i].key]);
- assert(p->val);
- (void)print_encode(h, p[i].val, 1);
- putchar('\"');
+ for (i = 0; i < sz; i++)
+ print_attr(h, htmlattrs[p[i].key], p[i].val);
+
+ /* Add non-overridable attributes. */
+
+ if (TAG_HTML == tag && HTML_XHTML_1_0_STRICT == h->type) {
+ print_attr(h, "xmlns", "http://www.w3.org/1999/xhtml");
+ print_attr(h, "xml:lang", "en");
+ print_attr(h, "lang", "en");
}
+ /* Accomodate for XML "well-formed" singleton escaping. */
+
if (HTML_AUTOCLOSE & htmltags[tag].flags)
switch (h->type) {
case (HTML_XHTML_1_0_STRICT):
static void
print_xmltype(struct html *h)
{
- const char *decl;
-
- switch (h->type) {
- case (HTML_XHTML_1_0_STRICT):
- decl = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>";
- break;
- default:
- decl = NULL;
- break;
- }
-
- if (NULL == decl)
- return;
- printf("%s\n", decl);
+ if (HTML_XHTML_1_0_STRICT == h->type)
+ printf("<?xml version=\"1.0\" encoding=\"UTF-8\"?>");
}
{
const char *doctype;
const char *dtd;
+ const char *name;
switch (h->type) {
case (HTML_HTML_4_01_STRICT):
+ name = "HTML";
doctype = "-//W3C//DTD HTML 4.01//EN";
dtd = "http://www.w3.org/TR/html4/strict.dtd";
break;
default:
+ name = "html";
doctype = "-//W3C//DTD XHTML 1.0 Strict//EN";
dtd = "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd";
break;
}
- printf("<!DOCTYPE HTML PUBLIC \"%s\" \"%s\">\n", doctype, dtd);
+ printf("<!DOCTYPE %s PUBLIC \"%s\" \"%s\">\n",
+ name, doctype, dtd);
}
void
-print_text(struct html *h, const char *p)
+print_text(struct html *h, const char *word)
{
- if (*p && 0 == *(p + 1))
- switch (*p) {
+ if (word[0] && '\0' == word[1])
+ switch (word[0]) {
case('.'):
/* FALLTHROUGH */
case(','):
case(')'):
/* FALLTHROUGH */
case(']'):
- /* FALLTHROUGH */
- case('}'):
if ( ! (HTML_IGNDELIM & h->flags))
h->flags |= HTML_NOSPACE;
break;
break;
}
- if ( ! (h->flags & HTML_NOSPACE))
- putchar(' ');
-
- assert(p);
- if ( ! print_encode(h, p, 0))
- h->flags &= ~HTML_NOSPACE;
+ if ( ! (HTML_NOSPACE & h->flags)) {
+ /* Manage keeps! */
+ if ( ! (HTML_KEEP & h->flags)) {
+ if (HTML_PREKEEP & h->flags)
+ h->flags |= HTML_KEEP;
+ putchar(' ');
+ } else
+ printf(" ");
+ }
- if (*p && 0 == *(p + 1))
- switch (*p) {
+ assert(word);
+ if ( ! print_encode(h, word, 0))
+ if ( ! (h->flags & HTML_NONOSPACE))
+ h->flags &= ~HTML_NOSPACE;
+
+ /*
+ * Note that we don't process the pipe: the parser sees it as
+ * punctuation, but we don't in terms of typography.
+ */
+ if (word[0] && '\0' == word[1])
+ switch (word[0]) {
case('('):
/* FALLTHROUGH */
case('['):
- /* FALLTHROUGH */
- case('{'):
h->flags |= HTML_NOSPACE;
break;
default:
break;
}
- if (su->pt)
- buffmt(h, "%s: %f%s;", p, v, u);
- else
- /* LINTED */
- buffmt(h, "%s: %d%s;", p, (int)v, u);
+ /*
+ * XXX: the CSS spec isn't clear as to which types accept
+ * integer or real numbers, so we just make them all decimals.
+ */
+ buffmt(h, "%s: %.2f%s;", p, v, u);
}
git.ckatri.com / mandoc.git / blobdiff