+ this = h->tag;
+ for (;;) {
+ next = this->next;
+ flags = htmltags[this->tag].flags;
+ if (flags & (HTML_INPHRASE | HTML_TOPHRASE))
+ print_ctag(h, this);
+ if ((flags & HTML_INPHRASE) == 0)
+ break;
+ this = next;
+ }
+}
+
+/*
+ * 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;
+
+ for (t = h->tag; t != NULL; t = t->next)
+ if (t->tag == TAG_PRE)
+ break;
+
+ had = t == NULL ? ROFF_fi : ROFF_nf;
+
+ 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;
+}
+
+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;
+
+ 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;
+
+ /*
+ * 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
+ */
+
+ for (cp = buf; *cp != '\0'; cp++)
+ if (isalnum((unsigned char)*cp) == 0 &&
+ strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL)
+ *cp = '_';
+
+ if (unique == 0)
+ return buf;
+
+ /* 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;
+}
+
+static int
+print_escape(struct html *h, char c)
+{
+
+ switch (c) {
+ case '<':
+ print_word(h, "<");
+ 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;
+}
+
+static int
+print_encode(struct html *h, const char *p, const char *pend, int norecurse)
+{
+ 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' };
+
+ if (pend == NULL)
+ pend = strchr(p, '\0');
+
+ breakline = 0;
+ nospace = 0;
+
+ while (p < pend) {
+ if (HTML_SKIPCHAR & h->flags && '\\' != *p) {
+ h->flags &= ~HTML_SKIPCHAR;
+ p++;
+ continue;
+ }
+
+ for (sz = strcspn(p, rejs); sz-- && p < pend; p++)
+ print_byte(h, *p);
+
+ if (breakline &&
+ (p >= pend || *p == ' ' || *p == ASCII_NBRSP)) {
+ print_otag(h, TAG_BR, "");
+ breakline = 0;
+ while (p < pend && (*p == ' ' || *p == ASCII_NBRSP))
+ p++;
+ continue;
+ }
+
+ if (p >= pend)
+ break;
+
+ if (*p == ' ') {
+ print_endword(h);
+ p++;
+ continue;
+ }
+
+ if (print_escape(h, *p++))
+ continue;
+
+ esc = mandoc_escape(&p, &seq, &len);
+ switch (esc) {
+ case ESCAPE_FONT:
+ case ESCAPE_FONTPREV:
+ case ESCAPE_FONTBOLD:
+ case ESCAPE_FONTITALIC:
+ case ESCAPE_FONTBI:
+ case ESCAPE_FONTCW:
+ case ESCAPE_FONTROMAN:
+ 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;
+ }
+
+ 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_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 nospace;
+}
+
+static void
+print_href(struct html *h, const char *name, const char *sec, int man)
+{
+ struct stat sb;
+ const char *p, *pp;
+ char *filename;
+
+ 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 *
+print_otag(struct html *h, enum htmltag tag, const char *fmt, ...)
+{
+ va_list ap;
+ struct tag *t;
+ const char *attr;
+ char *arg1, *arg2;
+ int style_written, tflags;
+
+ tflags = htmltags[tag].flags;
+
+ /* Flow content is not allowed in phrasing context. */
+
+ if ((tflags & HTML_INPHRASE) == 0) {
+ for (t = h->tag; t != NULL; t = t->next) {
+ if (t->closed)
+ continue;
+ assert((htmltags[t->tag].flags & HTML_TOPHRASE) == 0);
+ break;
+ }
+
+ /*
+ * Always wrap phrasing elements in a paragraph
+ * unless already contained in some flow container;
+ * never put them directly into a section.
+ */
+
+ } else if (tflags & HTML_TOPHRASE && h->tag->tag == TAG_SECTION)
+ print_otag(h, TAG_P, "c", "Pp");
+
+ /* Push this tag onto the stack of open scopes. */
+
+ if ((tflags & HTML_NOSTACK) == 0) {
+ t = mandoc_malloc(sizeof(struct tag));
+ t->tag = tag;
+ t->next = h->tag;
+ t->refcnt = 0;
+ t->closed = 0;
+ h->tag = t;
+ } else
+ t = NULL;
+
+ 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);
+ }
+ }
+
+ if ( ! (h->flags & HTML_NONOSPACE))
+ h->flags &= ~HTML_NOSPACE;
+ else
+ h->flags |= HTML_NOSPACE;
+
+ /* Print out the tag name and attributes. */
+
+ print_byte(h, '<');
+ print_word(h, htmltags[tag].name);
+
+ va_start(ap, fmt);
+
+ while (*fmt != '\0' && *fmt != 's') {
+
+ /* Parse attributes and arguments. */
+
+ arg1 = va_arg(ap, char *);
+ arg2 = NULL;
+ 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();
+ }
+ if (*fmt == 'M')
+ arg2 = va_arg(ap, char *);
+ if (arg1 == NULL)
+ continue;
+
+ /* Print the attributes. */
+
+ 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;
+ 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, '"');
+
+ va_end(ap);
+
+ /* Accommodate for "well-formed" singleton escaping. */
+
+ if (htmltags[tag].flags & HTML_NOSTACK)
+ 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_ctag(struct html *h, struct tag *tag)
+{
+ int tflags;
+
+ if (tag->closed == 0) {
+ tag->closed = 1;
+ if (tag == h->metaf)
+ h->metaf = NULL;
+ if (tag == h->tblt)
+ h->tblt = NULL;
+
+ 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);
+}
+
+void
+print_gen_comment(struct html *h, struct roff_node *n)
+{
+ int wantblank;
+
+ print_word(h, "<!-- This is an automatically generated file."
+ " Do not edit.");
+ 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;
+ }
+ if (wantblank)
+ print_endline(h);
+ print_word(h, " -->");
+ print_endline(h);
+ h->indent = 0;
+}
+
+void
+print_text(struct html *h, const char *word)
+{
+ /*
+ * Always wrap text in a paragraph unless already contained in
+ * some flow container; never put it directly into a section.
+ */
+
+ if (h->tag->tag == TAG_SECTION)
+ print_otag(h, TAG_P, "c", "Pp");
+
+ /* 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, " ");
+ }
+
+ /*
+ * Print the text, optionally surrounded by HTML whitespace,
+ * optionally manually switching fonts before and after.
+ */
+
+ assert(h->metaf == NULL);
+ print_metaf(h);
+ print_indent(h);
+ 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 != NULL) {
+ print_tagq(h, h->metaf);
+ h->metaf = NULL;
+ }
+
+ h->flags &= ~HTML_IGNDELIM;
+}
+
+void
+print_tagq(struct html *h, const struct tag *until)
git.ckatri.com / mandoc.git / blobdiff