=================================================================== RCS file: /cvs/mandoc/html.c,v retrieving revision 1.259 retrieving revision 1.270 diff -u -p -r1.259 -r1.270 --- mandoc/html.c 2019/09/03 12:31:05 1.259 +++ mandoc/html.c 2020/04/20 13:07:24 1.270 @@ -1,7 +1,7 @@ -/* $Id: html.c,v 1.259 2019/09/03 12:31:05 schwarze Exp $ */ +/* $Id: html.c,v 1.270 2020/04/20 13:07:24 schwarze Exp $ */ /* + * Copyright (c) 2011-2015, 2017-2020 Ingo Schwarze * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons - * Copyright (c) 2011-2015, 2017-2019 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 @@ -14,6 +14,9 @@ * 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. + * + * Common functions for mandoc(1) HTML formatters. + * For use by individual formatters and by the main program. */ #include "config.h" @@ -88,6 +91,7 @@ static const struct htmldata htmltags[TAG_MAX] = { {"span", HTML_INPHRASE | HTML_TOPHRASE}, {"var", HTML_INPHRASE | HTML_TOPHRASE}, {"br", HTML_INPHRASE | HTML_NOSTACK | HTML_NLALL}, + {"mark", HTML_INPHRASE }, {"math", HTML_INPHRASE | HTML_NLALL | HTML_INDENT}, {"mrow", 0}, {"mi", 0}, @@ -108,6 +112,11 @@ static const struct htmldata htmltags[TAG_MAX] = { }; /* Avoid duplicate HTML id= attributes. */ + +struct id_entry { + 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 *); @@ -142,7 +151,7 @@ html_alloc(const struct manoutput *outopts) if (outopts->toc) h->oflags |= HTML_TOC; - mandoc_ohash_init(&id_unique, 4, 0); + mandoc_ohash_init(&id_unique, 4, offsetof(struct id_entry, id)); return h; } @@ -151,17 +160,17 @@ static void html_reset_internal(struct html *h) { struct tag *tag; - char *cp; + struct id_entry *entry; unsigned int slot; while ((tag = h->tag) != NULL) { h->tag = tag->next; free(tag); } - cp = ohash_first(&id_unique, &slot); - while (cp != NULL) { - free(cp); - cp = ohash_next(&id_unique, &slot); + entry = ohash_first(&id_unique, &slot); + while (entry != NULL) { + free(entry); + entry = ohash_next(&id_unique, &slot); } ohash_delete(&id_unique); } @@ -170,7 +179,7 @@ void html_reset(void *p) { html_reset_internal(p); - mandoc_ohash_init(&id_unique, 4, 0); + mandoc_ohash_init(&id_unique, 4, offsetof(struct id_entry, id)); } void @@ -320,33 +329,66 @@ html_fillmode(struct html *h, enum roff_tok want) return had; } +/* + * Allocate a string to be used for the "id=" attribute of an HTML + * element and/or as a segment identifier for a URI in an element. + * The function may fail and return NULL if the node lacks text data + * to create the attribute from. + * The caller is responsible for free(3)ing the returned string. + * + * 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 struct roff_node *nch; - char *buf, *bufs, *cp; + struct id_entry *entry; + char *buf, *cp; + size_t len; unsigned int slot; - int suffix; - for (nch = n->child; nch != NULL; nch = nch->next) - if (nch->type != ROFFT_TEXT) - return NULL; + if (n->tag != NULL) + buf = mandoc_strdup(n->tag); + else { + switch (n->tok) { + case MDOC_Sh: + case MDOC_Ss: + case MDOC_Sx: + case MAN_SH: + 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; + } + } - 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 + * In addition, reserve '~' for ordinal suffixes. */ for (cp = buf; *cp != '\0'; cp++) if (isalnum((unsigned char)*cp) == 0 && - strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL) + strchr("!$&'()*+,-./:;=?@_", *cp) == NULL) *cp = '_'; if (unique == 0) @@ -354,25 +396,21 @@ html_make_id(const struct roff_node *n, int unique) /* 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; + if ((entry = ohash_find(&id_unique, slot)) == NULL) { + len = strlen(buf) + 1; + entry = mandoc_malloc(sizeof(*entry) + len); + entry->ord = 1; + memcpy(entry->id, buf, len); + ohash_insert(&id_unique, slot, entry); + } else if (unique == 1) + entry->ord++; + + if (entry->ord > 1) { + cp = buf; + mandoc_asprintf(&buf, "%s~%d", cp, entry->ord); + free(cp); } - ohash_insert(&id_unique, slot, buf); return buf; } @@ -590,8 +628,16 @@ print_otag(struct html *h, enum htmltag tag, const cha 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) { @@ -727,6 +773,49 @@ print_otag(struct html *h, enum htmltag tag, const cha return t; } +/* + * Print an element with an optional "id=" attribute. + * If the element has phrasing content and an "id=" attribute, + * also add a permalink: outside if it can be in phrasing context, + * inside otherwise. + */ +struct tag * +print_otag_id(struct html *h, enum htmltag elemtype, const char *cattr, + struct roff_node *n) +{ + struct roff_node *nch; + struct tag *ret, *t; + char *id, *href; + + ret = NULL; + id = href = NULL; + if (n->flags & NODE_ID) + id = html_make_id(n, 1); + if (n->flags & NODE_HREF) + href = id == NULL ? html_make_id(n, 2) : id; + if (href != NULL && htmltags[elemtype].flags & HTML_INPHRASE) + ret = print_otag(h, TAG_A, "chR", "permalink", href); + 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; +} + static void print_ctag(struct html *h, struct tag *tag) { @@ -796,6 +885,25 @@ print_gen_comment(struct html *h, struct roff_node *n) void print_text(struct html *h, const char *word) { + print_tagged_text(h, word, NULL); +} + +void +print_tagged_text(struct html *h, const char *word, struct roff_node *n) +{ + struct tag *t; + char *href; + + /* + * 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) @@ -805,9 +913,21 @@ print_text(struct html *h, const char *word) print_word(h, " "); } + /* + * 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); + + if (n != NULL && (href = html_make_id(n, 2)) != NULL) { + t = print_otag(h, TAG_A, "chR", "permalink", href); + free(href); + } else + t = NULL; + if ( ! print_encode(h, word, NULL, 0)) { if ( ! (h->flags & HTML_NONOSPACE)) h->flags &= ~HTML_NOSPACE; @@ -818,7 +938,8 @@ print_text(struct html *h, const char *word) if (h->metaf != NULL) { print_tagq(h, h->metaf); h->metaf = NULL; - } + } else if (t != NULL) + print_tagq(h, t); h->flags &= ~HTML_IGNDELIM; } @@ -945,15 +1066,12 @@ print_indent(struct html *h) { size_t i; - if (h->col) + if (h->col || h->noindent) return; - if (h->noindent == 0) { - h->col = h->indent * 2; - for (i = 0; i < h->col; i++) - putchar(' '); - } - h->flags &= ~HTML_NOSPACE; + h->col = h->indent * 2; + for (i = 0; i < h->col; i++) + putchar(' '); } /*