static void md_convert_hashes(cmark_iter *const iter) {
for(;;) {
cmark_event_type const event = cmark_iter_next(iter);
if(CMARK_EVENT_DONE == event) break;
if(CMARK_EVENT_EXIT != event) continue;
cmark_node *const node = cmark_iter_get_node(iter);
cmark_node_type const type = cmark_node_get_type(node);
if(CMARK_NODE_LINK != type && CMARK_NODE_IMAGE != type) continue;
char const *const URI = cmark_node_get_url(node);
if(!URI) continue;
if(0 != strncasecmp(URI, STR_LEN("hash:"))) continue;
cmark_node *sup = superscript("#", HASH_INFO_MSG, URI);
cmark_node_insert_after(node, sup);
cmark_iter_reset(iter, sup, CMARK_EVENT_EXIT);
char *escaped = QSEscape(URI, strlen(URI), true);
size_t const elen = strlen(escaped);
cmark_strbuf rel[1];
char const qpfx[] = "/sources/";
cmark_strbuf_init(&DEFAULT_MEM_ALLOCATOR, rel, sizeof(qpfx)-1+elen);
cmark_strbuf_put(rel, (unsigned char const *)qpfx, sizeof(qpfx)-1);
cmark_strbuf_put(rel, (unsigned char const *)escaped, elen);
free(escaped); escaped = NULL;
cmark_node_set_url(node, cmark_strbuf_cstr(rel));
cmark_strbuf_free(rel);
}
}
void cmark_consolidate_text_nodes(cmark_node *root) {
cmark_iter *iter = cmark_iter_new(root);
cmark_strbuf buf = GH_BUF_INIT;
cmark_event_type ev_type;
cmark_node *cur, *tmp, *next;
while ((ev_type = cmark_iter_next(iter)) != CMARK_EVENT_DONE) {
cur = cmark_iter_get_node(iter);
if (ev_type == CMARK_EVENT_ENTER && cur->type == CMARK_NODE_TEXT &&
cur->next && cur->next->type == CMARK_NODE_TEXT) {
cmark_strbuf_clear(&buf);
cmark_strbuf_put(&buf, cur->as.literal.data, cur->as.literal.len);
tmp = cur->next;
while (tmp && tmp->type == CMARK_NODE_TEXT) {
cmark_iter_next(iter); // advance pointer
cmark_strbuf_put(&buf, tmp->as.literal.data, tmp->as.literal.len);
next = tmp->next;
cmark_node_free(tmp);
tmp = next;
}
cmark_chunk_free(&cur->as.literal);
cur->as.literal = cmark_chunk_buf_detach(&buf);
}
}
cmark_strbuf_free(&buf);
cmark_iter_free(iter);
}
static void
S_parser_feed(cmark_parser *parser, const unsigned char *buffer, size_t len,
bool eof)
{
const unsigned char *end = buffer + len;
while (buffer < end) {
const unsigned char *eol
= (const unsigned char *)memchr(buffer, '\n',
end - buffer);
size_t line_len;
if (eol) {
line_len = eol + 1 - buffer;
} else if (eof) {
line_len = end - buffer;
} else {
cmark_strbuf_put(parser->linebuf, buffer, end - buffer);
break;
}
if (parser->linebuf->size > 0) {
cmark_strbuf_put(parser->linebuf, buffer, line_len);
S_process_line(parser, parser->linebuf->ptr,
parser->linebuf->size);
cmark_strbuf_clear(parser->linebuf);
} else {
S_process_line(parser, buffer, line_len);
}
buffer += line_len;
}
}
int houdini_escape_href(cmark_strbuf *ob, const uint8_t *src, bufsize_t size) {
static const uint8_t hex_chars[] = "0123456789ABCDEF";
bufsize_t i = 0, org;
uint8_t hex_str[3];
hex_str[0] = '%';
while (i < size) {
org = i;
while (i < size && HREF_SAFE[src[i]] != 0)
i++;
if (likely(i > org))
cmark_strbuf_put(ob, src + org, i - org);
/* escaping */
if (i >= size)
break;
switch (src[i]) {
/* amp appears all the time in URLs, but needs
* HTML-entity escaping to be inside an href */
case '&':
cmark_strbuf_puts(ob, "&");
break;
/* the single quote is a valid URL character
* according to the standard; it needs HTML
* entity escaping too */
case '\'':
cmark_strbuf_puts(ob, "'");
break;
/* the space can be escaped to %20 or a plus
* sign. we're going with the generic escape
* for now. the plus thing is more commonly seen
* when building GET strings */
#if 0
case ' ':
cmark_strbuf_putc(ob, '+');
break;
#endif
/* every other character goes with a %XX escaping */
default:
hex_str[1] = hex_chars[(src[i] >> 4) & 0xF];
hex_str[2] = hex_chars[src[i] & 0xF];
cmark_strbuf_put(ob, hex_str, 3);
}
i++;
}
return 1;
}
static void S_parser_feed(cmark_parser *parser, const unsigned char *buffer,
size_t len, bool eof) {
const unsigned char *end = buffer + len;
static const uint8_t repl[] = {239, 191, 189};
while (buffer < end) {
const unsigned char *eol;
bufsize_t chunk_len;
bool process = false;
for (eol = buffer; eol < end; ++eol) {
if (S_is_line_end_char(*eol)) {
process = true;
break;
}
if (*eol == '\0' && eol < end) {
break;
}
}
if (eol >= end && eof) {
process = true;
}
chunk_len = cmark_strbuf_check_bufsize(eol - buffer);
if (process) {
if (parser->linebuf->size > 0) {
cmark_strbuf_put(parser->linebuf, buffer, chunk_len);
S_process_line(parser, parser->linebuf->ptr, parser->linebuf->size);
cmark_strbuf_clear(parser->linebuf);
} else {
S_process_line(parser, buffer, chunk_len);
}
} else {
if (eol < end && *eol == '\0') {
// omit NULL byte
cmark_strbuf_put(parser->linebuf, buffer, chunk_len);
// add replacement character
cmark_strbuf_put(parser->linebuf, repl, 3);
chunk_len += 1; // so we advance the buffer past NULL
} else {
cmark_strbuf_put(parser->linebuf, buffer, chunk_len);
}
}
buffer += chunk_len;
// skip over line ending characters:
if (buffer < end && *buffer == '\r')
buffer++;
if (buffer < end && *buffer == '\n')
buffer++;
}
}
static cmark_node *try_opening_code_block(cmark_syntax_extension *self,
bool indented,
cmark_parser *parser,
cmark_node *parent,
const char *input)
{
cmark_node *ret = NULL;
cmark_bufsize_t matched = scan_open_gtkdoc_code_block(input,
cmark_parser_get_first_nonspace(parser));
if (!indented && matched) {
ret = cmark_parser_add_child(parser, parent,
CMARK_NODE_CODE_BLOCK, cmark_parser_get_offset(parser));
cmark_node_set_syntax_extension(ret, self);
cmark_node_set_fenced(ret, true, 2,
cmark_parser_get_first_nonspace(parser) - cmark_parser_get_offset(parser),
0);
cmark_parser_advance_offset(parser, input, matched, false);
matched = scan_language_comment(input, matched);
if (matched) {
cmark_strbuf *lang = cmark_strbuf_new(32);
cmark_strbuf_put(lang, (unsigned char *) input + 17, matched - 20);
/* Will be transformed to fence info */
cmark_node_set_string_content(ret, cmark_strbuf_get(lang));
cmark_strbuf_free(lang);
cmark_parser_advance_offset(parser, input, matched, false);
}
}
return ret;
}
int
houdini_escape_html0(cmark_strbuf *ob, const uint8_t *src, size_t size, int secure)
{
size_t i = 0, org, esc = 0;
while (i < size) {
org = i;
while (i < size && (esc = HTML_ESCAPE_TABLE[src[i]]) == 0)
i++;
if (i > org)
cmark_strbuf_put(ob, src + org, i - org);
/* escaping */
if (unlikely(i >= size))
break;
/* The forward slash is only escaped in secure mode */
if ((src[i] == '/' || src[i] == '\'') && !secure) {
cmark_strbuf_putc(ob, src[i]);
} else {
cmark_strbuf_puts(ob, HTML_ESCAPES[esc]);
}
i++;
}
return 1;
}
static void
S_parser_feed(cmark_parser *parser, const unsigned char *buffer, size_t len,
bool eof)
{
const unsigned char *end = buffer + len;
while (buffer < end) {
const unsigned char *eol;
size_t line_len;
for (eol = buffer; eol < end; ++eol) {
if (*eol == '\r' || *eol == '\n')
break;
}
if (eol >= end)
eol = NULL;
if (eol) {
if (eol < end && *eol == '\r')
eol++;
if (eol < end && *eol == '\n')
eol++;
line_len = eol - buffer;
} else if (eof) {
line_len = end - buffer;
} else {
cmark_strbuf_put(parser->linebuf, buffer, end - buffer);
break;
}
if (parser->linebuf->size > 0) {
cmark_strbuf_put(parser->linebuf, buffer, line_len);
S_process_line(parser, parser->linebuf->ptr,
parser->linebuf->size);
cmark_strbuf_clear(parser->linebuf);
} else {
S_process_line(parser, buffer, line_len);
}
buffer += line_len;
}
}
static cmark_node *www_match(cmark_parser *parser, cmark_node *parent,
cmark_inline_parser *inline_parser) {
cmark_chunk *chunk = cmark_inline_parser_get_chunk(inline_parser);
size_t max_rewind = cmark_inline_parser_get_offset(inline_parser);
uint8_t *data = chunk->data + max_rewind;
size_t size = chunk->len - max_rewind;
int start = cmark_inline_parser_get_column(inline_parser);
size_t link_end;
if (max_rewind > 0 && strchr("*_~(", data[-1]) == NULL &&
!cmark_isspace(data[-1]))
return 0;
if (size < 4 || memcmp(data, "www.", strlen("www.")) != 0)
return 0;
link_end = check_domain(data, size, 0);
if (link_end == 0)
return NULL;
while (link_end < size && !cmark_isspace(data[link_end]))
link_end++;
link_end = autolink_delim(data, link_end);
if (link_end == 0)
return NULL;
cmark_inline_parser_set_offset(inline_parser, (int)(max_rewind + link_end));
cmark_node *node = cmark_node_new_with_mem(CMARK_NODE_LINK, parser->mem);
cmark_strbuf buf;
cmark_strbuf_init(parser->mem, &buf, 10);
cmark_strbuf_puts(&buf, "http://");
cmark_strbuf_put(&buf, data, (bufsize_t)link_end);
node->as.link.url = cmark_chunk_buf_detach(&buf);
cmark_node *text = cmark_node_new_with_mem(CMARK_NODE_TEXT, parser->mem);
text->as.literal =
cmark_chunk_dup(chunk, (bufsize_t)max_rewind, (bufsize_t)link_end);
cmark_node_append_child(node, text);
node->start_line = text->start_line =
node->end_line = text->end_line =
cmark_inline_parser_get_line(inline_parser);
node->start_column = text->start_column = start - 1;
node->end_column = text->end_column = cmark_inline_parser_get_column(inline_parser) - 1;
return node;
}
void utf8proc_detab(cmark_strbuf *ob, const uint8_t *line, size_t size)
{
static const uint8_t whitespace[] = " ";
size_t i = 0, tab = 0;
while (i < size) {
size_t org = i;
while (i < size && line[i] != '\t' && line[i] != '\0'
&& line[i] < 0x80) {
i++;
tab++;
}
if (i > org)
cmark_strbuf_put(ob, line + org, i - org);
if (i >= size)
break;
if (line[i] == '\t') {
int numspaces = 4 - (tab % 4);
cmark_strbuf_put(ob, whitespace, numspaces);
i += 1;
tab += numspaces;
} else {
int charlen = utf8proc_valid(line + i, size - i);
if (charlen >= 0) {
cmark_strbuf_put(ob, line + i, charlen);
} else {
encode_unknown(ob);
charlen = -charlen;
}
i += charlen;
tab += 1;
}
}
}
static void filter_html_block(cmark_html_renderer *renderer, uint8_t *data, size_t len) {
cmark_strbuf *html = renderer->html;
cmark_llist *it;
cmark_syntax_extension *ext;
bool filtered;
uint8_t *match;
while (len) {
match = (uint8_t *) memchr(data, '<', len);
if (!match)
break;
if (match != data) {
cmark_strbuf_put(html, data, (bufsize_t)(match - data));
len -= (match - data);
data = match;
}
filtered = false;
for (it = renderer->filter_extensions; it; it = it->next) {
ext = ((cmark_syntax_extension *) it->data);
if (!ext->html_filter_func(ext, data, len)) {
filtered = true;
break;
}
}
if (!filtered) {
cmark_strbuf_putc(html, '<');
} else {
cmark_strbuf_puts(html, "<");
}
++data;
--len;
}
if (len)
cmark_strbuf_put(html, data, (bufsize_t)len);
}
int
houdini_unescape_html(cmark_strbuf *ob, const uint8_t *src, size_t size)
{
size_t i = 0, org, ent;
while (i < size) {
org = i;
while (i < size && src[i] != '&')
i++;
if (likely(i > org)) {
if (unlikely(org == 0)) {
if (i >= size)
return 0;
cmark_strbuf_grow(ob, HOUDINI_UNESCAPED_SIZE(size));
}
cmark_strbuf_put(ob, src + org, i - org);
}
/* escaping */
if (i >= size)
break;
i++;
ent = houdini_unescape_ent(ob, src + i, size - i);
i += ent;
/* not really an entity */
if (ent == 0)
cmark_strbuf_putc(ob, '&');
}
return 1;
}
static void add_line(cmark_node* node, cmark_chunk *ch, int offset)
{
assert(node->open);
cmark_strbuf_put(&node->string_content, ch->data + offset, ch->len - offset);
}
static void encode_unknown(cmark_strbuf *buf)
{
static const uint8_t repl[] = {239, 191, 189};
cmark_strbuf_put(buf, repl, 3);
}
size_t
houdini_unescape_ent(cmark_strbuf *ob, const uint8_t *src, size_t size)
{
size_t i = 0;
if (size >= 3 && src[0] == '#') {
int codepoint = 0;
int num_digits = 0;
if (_isdigit(src[1])) {
for (i = 1; i < size && _isdigit(src[i]); ++i) {
codepoint = (codepoint * 10) + (src[i] - '0');
if (codepoint >= 0x110000) {
// Keep counting digits but
// avoid integer overflow.
codepoint = 0x110000;
}
}
num_digits = i - 1;
}
else if (src[1] == 'x' || src[1] == 'X') {
for (i = 2; i < size && _isxdigit(src[i]); ++i) {
codepoint = (codepoint * 16) + ((src[i] | 32) % 39 - 9);
if (codepoint >= 0x110000) {
// Keep counting digits but
// avoid integer overflow.
codepoint = 0x110000;
}
}
num_digits = i - 2;
}
if (num_digits >= 1 && num_digits <= 8 &&
i < size && src[i] == ';') {
if (codepoint == 0 ||
(codepoint >= 0xD800 && codepoint < 0xE000) ||
codepoint >= 0x110000) {
codepoint = 0xFFFD;
}
utf8proc_encode_char(codepoint, ob);
return i + 1;
}
}
else {
if (size > MAX_WORD_LENGTH)
size = MAX_WORD_LENGTH;
for (i = MIN_WORD_LENGTH; i < size; ++i) {
if (src[i] == ' ')
break;
if (src[i] == ';') {
const struct html_ent *entity = find_entity((char *)src, i);
if (entity != NULL) {
int len = 0;
while (len < 4 && entity->utf8[len] != '\0') {
++len;
}
cmark_strbuf_put(ob, entity->utf8, len);
return i + 1;
}
break;
}
}
}
return 0;
}
void houdini_unescape_html_f(cmark_strbuf *ob, const uint8_t *src, size_t size)
{
if (!houdini_unescape_html(ob, src, size))
cmark_strbuf_put(ob, src, size);
}
static int
S_render_node(cmark_node *node, cmark_event_type ev_type,
struct render_state *state, int options)
{
cmark_node *parent;
cmark_node *grandparent;
cmark_strbuf *html = state->html;
char start_header[] = "<h0";
char end_header[] = "</h0";
bool tight;
bool entering = (ev_type == CMARK_EVENT_ENTER);
if (state->plain == node) { // back at original node
state->plain = NULL;
}
if (state->plain != NULL) {
switch(node->type) {
case CMARK_NODE_TEXT:
case CMARK_NODE_CODE:
case CMARK_NODE_INLINE_HTML:
escape_html(html, node->as.literal.data,
node->as.literal.len);
break;
case CMARK_NODE_LINEBREAK:
case CMARK_NODE_SOFTBREAK:
cmark_strbuf_putc(html, ' ');
break;
default:
break;
}
return 1;
}
switch (node->type) {
case CMARK_NODE_DOCUMENT:
break;
case CMARK_NODE_BLOCK_QUOTE:
if (entering) {
cr(html);
cmark_strbuf_puts(html, "<blockquote");
S_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, ">\n");
} else {
cr(html);
cmark_strbuf_puts(html, "</blockquote>\n");
}
break;
case CMARK_NODE_LIST: {
cmark_list_type list_type = node->as.list.list_type;
int start = node->as.list.start;
if (entering) {
cr(html);
if (list_type == CMARK_BULLET_LIST) {
cmark_strbuf_puts(html, "<ul");
S_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, ">\n");
} else if (start == 1) {
cmark_strbuf_puts(html, "<ol");
S_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, ">\n");
} else {
cmark_strbuf_printf(html,
"<ol start=\"%d\"",
start);
S_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, ">\n");
}
} else {
cmark_strbuf_puts(html,
list_type == CMARK_BULLET_LIST ?
"</ul>\n" : "</ol>\n");
}
break;
}
case CMARK_NODE_ITEM:
if (entering) {
cr(html);
cmark_strbuf_puts(html, "<li");
S_render_sourcepos(node, html, options);
cmark_strbuf_putc(html, '>');
} else {
cmark_strbuf_puts(html, "</li>\n");
}
break;
case CMARK_NODE_HEADER:
if (entering) {
cr(html);
start_header[2] = '0' + node->as.header.level;
cmark_strbuf_puts(html, start_header);
S_render_sourcepos(node, html, options);
cmark_strbuf_putc(html, '>');
} else {
end_header[3] = '0' + node->as.header.level;
cmark_strbuf_puts(html, end_header);
cmark_strbuf_puts(html, ">\n");
}
break;
case CMARK_NODE_CODE_BLOCK:
cr(html);
if (!node->as.code.fenced || node->as.code.info.len == 0) {
cmark_strbuf_puts(html, "<pre");
S_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, "><code>");
} else {
int first_tag = 0;
while (first_tag < node->as.code.info.len &&
node->as.code.info.data[first_tag] != ' ') {
first_tag += 1;
}
cmark_strbuf_puts(html, "<pre");
S_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, "><code class=\"language-");
escape_html(html, node->as.code.info.data, first_tag);
cmark_strbuf_puts(html, "\">");
}
escape_html(html, node->as.code.literal.data,
node->as.code.literal.len);
cmark_strbuf_puts(html, "</code></pre>\n");
break;
case CMARK_NODE_HTML:
cr(html);
cmark_strbuf_put(html, node->as.literal.data, node->as.literal.len);
break;
case CMARK_NODE_HRULE:
cr(html);
cmark_strbuf_puts(html, "<hr");
S_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, " />\n");
break;
case CMARK_NODE_PARAGRAPH:
parent = cmark_node_parent(node);
grandparent = cmark_node_parent(parent);
if (grandparent != NULL &&
grandparent->type == CMARK_NODE_LIST) {
tight = grandparent->as.list.tight;
} else {
tight = false;
}
if (!tight) {
if (entering) {
cr(html);
cmark_strbuf_puts(html, "<p");
S_render_sourcepos(node, html, options);
cmark_strbuf_putc(html, '>');
} else {
cmark_strbuf_puts(html, "</p>\n");
}
}
break;
case CMARK_NODE_TEXT:
escape_html(html, node->as.literal.data, node->as.literal.len);
break;
case CMARK_NODE_LINEBREAK:
cmark_strbuf_puts(html, "<br />\n");
break;
case CMARK_NODE_SOFTBREAK:
if (options & CMARK_OPT_HARDBREAKS) {
cmark_strbuf_puts(html, "<br />\n");
} else {
cmark_strbuf_putc(html, '\n');
}
break;
case CMARK_NODE_CODE:
cmark_strbuf_puts(html, "<code>");
escape_html(html, node->as.literal.data, node->as.literal.len);
cmark_strbuf_puts(html, "</code>");
break;
case CMARK_NODE_INLINE_HTML:
cmark_strbuf_put(html, node->as.literal.data, node->as.literal.len);
break;
case CMARK_NODE_STRONG:
if (entering) {
cmark_strbuf_puts(html, "<strong>");
} else {
cmark_strbuf_puts(html, "</strong>");
}
break;
case CMARK_NODE_EMPH:
if (entering) {
cmark_strbuf_puts(html, "<em>");
} else {
cmark_strbuf_puts(html, "</em>");
}
break;
case CMARK_NODE_LINK:
if (entering) {
cmark_strbuf_puts(html, "<a href=\"");
if (node->as.link.url)
escape_href(html, node->as.link.url, -1);
if (node->as.link.title) {
cmark_strbuf_puts(html, "\" title=\"");
escape_html(html, node->as.link.title, -1);
}
cmark_strbuf_puts(html, "\">");
} else {
cmark_strbuf_puts(html, "</a>");
}
break;
case CMARK_NODE_IMAGE:
if (entering) {
cmark_strbuf_puts(html, "<img src=\"");
if (node->as.link.url)
escape_href(html, node->as.link.url, -1);
cmark_strbuf_puts(html, "\" alt=\"");
state->plain = node;
} else {
if (node->as.link.title) {
cmark_strbuf_puts(html, "\" title=\"");
escape_html(html, node->as.link.title, -1);
}
cmark_strbuf_puts(html, "\" />");
}
break;
default:
assert(false);
break;
}
// cmark_strbuf_putc(html, 'x');
return 1;
}
static void postprocess_text(cmark_parser *parser, cmark_node *text, int offset) {
size_t link_end;
uint8_t *data = text->as.literal.data,
*at;
size_t size = text->as.literal.len;
int rewind, max_rewind,
nb = 0, np = 0, ns = 0;
if (offset < 0 || (size_t)offset >= size)
return;
data += offset;
size -= offset;
at = (uint8_t *)memchr(data, '@', size);
if (!at)
return;
max_rewind = (int)(at - data);
data += max_rewind;
size -= max_rewind;
for (rewind = 0; rewind < max_rewind; ++rewind) {
uint8_t c = data[-rewind - 1];
if (cmark_isalnum(c))
continue;
if (strchr(".+-_", c) != NULL)
continue;
if (c == '/')
ns++;
break;
}
if (rewind == 0 || ns > 0) {
postprocess_text(parser, text, max_rewind + 1 + offset);
return;
}
for (link_end = 0; link_end < size; ++link_end) {
uint8_t c = data[link_end];
if (cmark_isalnum(c))
continue;
if (c == '@')
nb++;
else if (c == '.' && link_end < size - 1 && cmark_isalnum(data[link_end + 1]))
np++;
else if (c != '-' && c != '_')
break;
}
if (link_end < 2 || nb != 1 || np == 0 ||
(!cmark_isalpha(data[link_end - 1]) && data[link_end - 1] != '.')) {
postprocess_text(parser, text, max_rewind + 1 + offset);
return;
}
link_end = autolink_delim(data, link_end);
if (link_end == 0) {
postprocess_text(parser, text, max_rewind + 1 + offset);
return;
}
cmark_chunk_to_cstr(parser->mem, &text->as.literal);
cmark_node *link_node = cmark_node_new_with_mem(CMARK_NODE_LINK, parser->mem);
cmark_strbuf buf;
cmark_strbuf_init(parser->mem, &buf, 10);
cmark_strbuf_puts(&buf, "mailto:");
cmark_strbuf_put(&buf, data - rewind, (bufsize_t)(link_end + rewind));
link_node->as.link.url = cmark_chunk_buf_detach(&buf);
cmark_node *link_text = cmark_node_new_with_mem(CMARK_NODE_TEXT, parser->mem);
cmark_chunk email = cmark_chunk_dup(
&text->as.literal,
offset + max_rewind - rewind,
(bufsize_t)(link_end + rewind));
cmark_chunk_to_cstr(parser->mem, &email);
link_text->as.literal = email;
cmark_node_append_child(link_node, link_text);
cmark_node_insert_after(text, link_node);
cmark_node *post = cmark_node_new_with_mem(CMARK_NODE_TEXT, parser->mem);
post->as.literal = cmark_chunk_dup(&text->as.literal,
(bufsize_t)(offset + max_rewind + link_end),
(bufsize_t)(size - link_end));
cmark_chunk_to_cstr(parser->mem, &post->as.literal);
cmark_node_insert_after(link_node, post);
text->as.literal.len = offset + max_rewind - rewind;
text->as.literal.data[text->as.literal.len] = 0;
postprocess_text(parser, post, 0);
}
static void S_process_line(cmark_parser *parser, const unsigned char *buffer,
bufsize_t bytes) {
cmark_node *last_matched_container;
bufsize_t matched = 0;
int lev = 0;
int i;
cmark_list *data = NULL;
bool all_matched = true;
cmark_node *container;
bool indented;
cmark_chunk input;
bool maybe_lazy;
if (parser->options & CMARK_OPT_VALIDATE_UTF8) {
cmark_utf8proc_check(parser->curline, buffer, bytes);
} else {
cmark_strbuf_put(parser->curline, buffer, bytes);
}
// ensure line ends with a newline:
if (bytes == 0 || !S_is_line_end_char(parser->curline->ptr[bytes - 1])) {
cmark_strbuf_putc(parser->curline, '\n');
}
parser->offset = 0;
parser->column = 0;
parser->blank = false;
input.data = parser->curline->ptr;
input.len = parser->curline->size;
// container starts at the document root.
container = parser->root;
parser->line_number++;
// for each containing node, try to parse the associated line start.
// bail out on failure: container will point to the last matching node.
while (container->last_child && container->last_child->open) {
container = container->last_child;
S_find_first_nonspace(parser, &input);
if (container->type == CMARK_NODE_BLOCK_QUOTE) {
matched =
parser->indent <= 3 && peek_at(&input, parser->first_nonspace) == '>';
if (matched) {
S_advance_offset(parser, &input, parser->indent + 1, true);
if (peek_at(&input, parser->offset) == ' ')
parser->offset++;
} else {
all_matched = false;
}
} else if (container->type == CMARK_NODE_ITEM) {
if (parser->indent >=
container->as.list.marker_offset + container->as.list.padding) {
S_advance_offset(parser, &input, container->as.list.marker_offset +
container->as.list.padding,
true);
} else if (parser->blank && container->first_child != NULL) {
// if container->first_child is NULL, then the opening line
// of the list item was blank after the list marker; in this
// case, we are done with the list item.
S_advance_offset(parser, &input,
parser->first_nonspace - parser->offset, false);
} else {
all_matched = false;
}
} else if (container->type == CMARK_NODE_CODE_BLOCK) {
if (!container->as.code.fenced) { // indented
if (parser->indent >= CODE_INDENT) {
S_advance_offset(parser, &input, CODE_INDENT, true);
} else if (parser->blank) {
S_advance_offset(parser, &input,
parser->first_nonspace - parser->offset, false);
} else {
all_matched = false;
}
} else { // fenced
matched = 0;
if (parser->indent <= 3 && (peek_at(&input, parser->first_nonspace) ==
container->as.code.fence_char)) {
matched = scan_close_code_fence(&input, parser->first_nonspace);
}
if (matched >= container->as.code.fence_length) {
// closing fence - and since we're at
// the end of a line, we can return:
all_matched = false;
S_advance_offset(parser, &input, matched, false);
parser->current = finalize(parser, container);
goto finished;
} else {
// skip opt. spaces of fence parser->offset
i = container->as.code.fence_offset;
while (i > 0 && peek_at(&input, parser->offset) == ' ') {
S_advance_offset(parser, &input, 1, false);
i--;
}
}
}
} else if (container->type == CMARK_NODE_HEADING) {
// a heading can never contain more than one line
all_matched = false;
} else if (container->type == CMARK_NODE_HTML_BLOCK) {
switch (container->as.html_block_type) {
case 1:
case 2:
case 3:
case 4:
case 5:
// these types of blocks can accept blanks
break;
case 6:
case 7:
if (parser->blank) {
all_matched = false;
}
break;
default:
fprintf(stderr, "Error (%s:%d): Unknown HTML block type %d\n", __FILE__,
__LINE__, container->as.html_block_type);
exit(1);
}
} else if (container->type == CMARK_NODE_PARAGRAPH) {
if (parser->blank) {
all_matched = false;
}
}
if (!all_matched) {
container = container->parent; // back up to last matching node
break;
}
}
last_matched_container = container;
// check to see if we've hit 2nd blank line, break out of list:
if (parser->blank && container->last_line_blank) {
break_out_of_lists(parser, &container);
}
maybe_lazy = parser->current->type == CMARK_NODE_PARAGRAPH;
// try new container starts:
while (container->type != CMARK_NODE_CODE_BLOCK &&
container->type != CMARK_NODE_HTML_BLOCK) {
S_find_first_nonspace(parser, &input);
indented = parser->indent >= CODE_INDENT;
if (!indented && peek_at(&input, parser->first_nonspace) == '>') {
S_advance_offset(parser, &input,
parser->first_nonspace + 1 - parser->offset, false);
// optional following character
if (peek_at(&input, parser->offset) == ' ')
S_advance_offset(parser, &input, 1, false);
container = add_child(parser, container, CMARK_NODE_BLOCK_QUOTE,
parser->offset + 1);
} else if (!indented && (matched = scan_atx_heading_start(
&input, parser->first_nonspace))) {
S_advance_offset(parser, &input,
parser->first_nonspace + matched - parser->offset,
false);
container =
add_child(parser, container, CMARK_NODE_HEADING, parser->offset + 1);
bufsize_t hashpos =
cmark_chunk_strchr(&input, '#', parser->first_nonspace);
int level = 0;
while (peek_at(&input, hashpos) == '#') {
level++;
hashpos++;
}
container->as.heading.level = level;
container->as.heading.setext = false;
} else if (!indented && (matched = scan_open_code_fence(
&input, parser->first_nonspace))) {
container = add_child(parser, container, CMARK_NODE_CODE_BLOCK,
parser->first_nonspace + 1);
container->as.code.fenced = true;
container->as.code.fence_char = peek_at(&input, parser->first_nonspace);
container->as.code.fence_length = matched;
container->as.code.fence_offset =
(int8_t)(parser->first_nonspace - parser->offset);
container->as.code.info = cmark_chunk_literal("");
S_advance_offset(parser, &input,
parser->first_nonspace + matched - parser->offset,
false);
} else if (!indented && ((matched = scan_html_block_start(
&input, parser->first_nonspace)) ||
(container->type != CMARK_NODE_PARAGRAPH &&
(matched = scan_html_block_start_7(
&input, parser->first_nonspace))))) {
container = add_child(parser, container, CMARK_NODE_HTML_BLOCK,
parser->first_nonspace + 1);
container->as.html_block_type = matched;
// note, we don't adjust parser->offset because the tag is part of the
// text
} else if (!indented && container->type == CMARK_NODE_PARAGRAPH &&
(lev =
scan_setext_heading_line(&input, parser->first_nonspace))) {
container->type = CMARK_NODE_HEADING;
container->as.heading.level = lev;
container->as.heading.setext = true;
S_advance_offset(parser, &input, input.len - 1 - parser->offset, false);
} else if (!indented &&
!(container->type == CMARK_NODE_PARAGRAPH && !all_matched) &&
(matched =
scan_thematic_break(&input, parser->first_nonspace))) {
// it's only now that we know the line is not part of a setext heading:
container = add_child(parser, container, CMARK_NODE_THEMATIC_BREAK,
parser->first_nonspace + 1);
S_advance_offset(parser, &input, input.len - 1 - parser->offset, false);
} else if ((matched =
parse_list_marker(&input, parser->first_nonspace, &data)) &&
(!indented || container->type == CMARK_NODE_LIST)) {
// Note that we can have new list items starting with >= 4
// spaces indent, as long as the list container is still open.
// compute padding:
S_advance_offset(parser, &input,
parser->first_nonspace + matched - parser->offset,
false);
i = 0;
while (i <= 5 && peek_at(&input, parser->offset + i) == ' ') {
i++;
}
// i = number of spaces after marker, up to 5
if (i >= 5 || i < 1 ||
S_is_line_end_char(peek_at(&input, parser->offset))) {
data->padding = matched + 1;
if (i > 0) {
S_advance_offset(parser, &input, 1, false);
}
} else {
data->padding = matched + i;
S_advance_offset(parser, &input, i, true);
}
// check container; if it's a list, see if this list item
// can continue the list; otherwise, create a list container.
data->marker_offset = parser->indent;
if (container->type != CMARK_NODE_LIST ||
!lists_match(&container->as.list, data)) {
container = add_child(parser, container, CMARK_NODE_LIST,
parser->first_nonspace + 1);
memcpy(&container->as.list, data, sizeof(*data));
}
// add the list item
container = add_child(parser, container, CMARK_NODE_ITEM,
parser->first_nonspace + 1);
/* TODO: static */
memcpy(&container->as.list, data, sizeof(*data));
free(data);
} else if (indented && !maybe_lazy && !parser->blank) {
S_advance_offset(parser, &input, CODE_INDENT, true);
container = add_child(parser, container, CMARK_NODE_CODE_BLOCK,
parser->offset + 1);
container->as.code.fenced = false;
container->as.code.fence_char = 0;
container->as.code.fence_length = 0;
container->as.code.fence_offset = 0;
container->as.code.info = cmark_chunk_literal("");
} else {
break;
}
if (accepts_lines(container->type)) {
// if it's a line container, it can't contain other containers
break;
}
maybe_lazy = false;
}
// what remains at parser->offset is a text line. add the text to the
// appropriate container.
S_find_first_nonspace(parser, &input);
if (parser->blank && container->last_child) {
container->last_child->last_line_blank = true;
}
// block quote lines are never blank as they start with >
// and we don't count blanks in fenced code for purposes of tight/loose
// lists or breaking out of lists. we also don't set last_line_blank
// on an empty list item.
container->last_line_blank =
(parser->blank && container->type != CMARK_NODE_BLOCK_QUOTE &&
container->type != CMARK_NODE_HEADING &&
container->type != CMARK_NODE_THEMATIC_BREAK &&
!(container->type == CMARK_NODE_CODE_BLOCK &&
container->as.code.fenced) &&
!(container->type == CMARK_NODE_ITEM && container->first_child == NULL &&
container->start_line == parser->line_number));
cmark_node *cont = container;
while (cont->parent) {
cont->parent->last_line_blank = false;
cont = cont->parent;
}
if (parser->current != last_matched_container &&
container == last_matched_container && !parser->blank &&
parser->current->type == CMARK_NODE_PARAGRAPH &&
cmark_strbuf_len(&parser->current->string_content) > 0) {
add_line(parser->current, &input, parser->offset);
} else { // not a lazy continuation
// finalize any blocks that were not matched and set cur to container:
while (parser->current != last_matched_container) {
parser->current = finalize(parser, parser->current);
assert(parser->current != NULL);
}
if (container->type == CMARK_NODE_CODE_BLOCK) {
add_line(container, &input, parser->offset);
} else if (container->type == CMARK_NODE_HTML_BLOCK) {
add_line(container, &input, parser->offset);
int matches_end_condition;
switch (container->as.html_block_type) {
case 1:
// </script>, </style>, </pre>
matches_end_condition =
scan_html_block_end_1(&input, parser->first_nonspace);
break;
case 2:
// -->
matches_end_condition =
scan_html_block_end_2(&input, parser->first_nonspace);
break;
case 3:
// ?>
matches_end_condition =
scan_html_block_end_3(&input, parser->first_nonspace);
break;
case 4:
// >
matches_end_condition =
scan_html_block_end_4(&input, parser->first_nonspace);
break;
case 5:
// ]]>
matches_end_condition =
scan_html_block_end_5(&input, parser->first_nonspace);
break;
default:
matches_end_condition = 0;
break;
}
if (matches_end_condition) {
container = finalize(parser, container);
assert(parser->current != NULL);
}
} else if (parser->blank) {
// ??? do nothing
} else if (accepts_lines(container->type)) {
if (container->type == CMARK_NODE_HEADING &&
container->as.heading.setext == false) {
chop_trailing_hashtags(&input);
}
add_line(container, &input, parser->first_nonspace);
} else {
// create paragraph container for line
container = add_child(parser, container, CMARK_NODE_PARAGRAPH,
parser->first_nonspace + 1);
add_line(container, &input, parser->first_nonspace);
}
parser->current = container;
}
finished:
parser->last_line_length = input.len;
if (parser->last_line_length &&
input.data[parser->last_line_length - 1] == '\n')
parser->last_line_length -= 1;
if (parser->last_line_length &&
input.data[parser->last_line_length - 1] == '\r')
parser->last_line_length -= 1;
cmark_strbuf_clear(parser->curline);
}
static
void S_out(cmark_renderer *renderer,
const char *source,
bool wrap,
cmark_escaping escape)
{
int length = cmark_strbuf_safe_strlen(source);
unsigned char nextc;
int32_t c;
int i = 0;
int len;
cmark_chunk remainder = cmark_chunk_literal("");
int k = renderer->buffer->size - 1;
wrap = wrap && !renderer->no_wrap;
if (renderer->in_tight_list_item && renderer->need_cr > 1) {
renderer->need_cr = 1;
}
while (renderer->need_cr) {
if (k < 0 || renderer->buffer->ptr[k] == '\n') {
k -= 1;
} else {
cmark_strbuf_putc(renderer->buffer, '\n');
if (renderer->need_cr > 1) {
cmark_strbuf_put(renderer->buffer, renderer->prefix->ptr,
renderer->prefix->size);
}
}
renderer->column = 0;
renderer->begin_line = true;
renderer->need_cr -= 1;
}
while (i < length) {
if (renderer->begin_line) {
cmark_strbuf_put(renderer->buffer, renderer->prefix->ptr,
renderer->prefix->size);
// note: this assumes prefix is ascii:
renderer->column = renderer->prefix->size;
}
len = utf8proc_iterate((const uint8_t *)source + i, length - i, &c);
if (len == -1) { // error condition
return; // return without rendering rest of string
}
nextc = source[i + len];
if (c == 32 && wrap) {
if (!renderer->begin_line) {
cmark_strbuf_putc(renderer->buffer, ' ');
renderer->column += 1;
renderer->begin_line = false;
renderer->last_breakable = renderer->buffer->size -
1;
// skip following spaces
while (source[i + 1] == ' ') {
i++;
}
}
} else if (c == 10) {
cmark_strbuf_putc(renderer->buffer, '\n');
renderer->column = 0;
renderer->begin_line = true;
renderer->last_breakable = 0;
} else if (escape == LITERAL) {
cmark_render_code_point(renderer, c);
renderer->begin_line = false;
} else {
(renderer->outc)(renderer, escape, c, nextc);
renderer->begin_line = false;
}
// If adding the character went beyond width, look for an
// earlier place where the line could be broken:
if (renderer->width > 0 &&
renderer->column > renderer->width &&
!renderer->begin_line &&
renderer->last_breakable > 0) {
// copy from last_breakable to remainder
cmark_chunk_set_cstr(&remainder, (char *) renderer->buffer->ptr + renderer->last_breakable + 1);
// truncate at last_breakable
cmark_strbuf_truncate(renderer->buffer, renderer->last_breakable);
// add newline, prefix, and remainder
cmark_strbuf_putc(renderer->buffer, '\n');
cmark_strbuf_put(renderer->buffer, renderer->prefix->ptr,
renderer->prefix->size);
cmark_strbuf_put(renderer->buffer, remainder.data, remainder.len);
renderer->column = renderer->prefix->size + remainder.len;
cmark_chunk_free(&remainder);
renderer->last_breakable = 0;
renderer->begin_line = false;
}
i += len;
}
}
static int S_render_node(cmark_html_renderer *renderer, cmark_node *node,
cmark_event_type ev_type, int options) {
cmark_node *parent;
cmark_node *grandparent;
cmark_strbuf *html = renderer->html;
cmark_llist *it;
cmark_syntax_extension *ext;
char start_heading[] = "<h0";
char end_heading[] = "</h0";
bool tight;
bool filtered;
char buffer[BUFFER_SIZE];
bool entering = (ev_type == CMARK_EVENT_ENTER);
if (renderer->plain == node) { // back at original node
renderer->plain = NULL;
}
if (renderer->plain != NULL) {
switch (node->type) {
case CMARK_NODE_TEXT:
case CMARK_NODE_CODE:
case CMARK_NODE_HTML_INLINE:
escape_html(html, node->as.literal.data, node->as.literal.len);
break;
case CMARK_NODE_LINEBREAK:
case CMARK_NODE_SOFTBREAK:
cmark_strbuf_putc(html, ' ');
break;
default:
break;
}
return 1;
}
if (node->extension && node->extension->html_render_func) {
node->extension->html_render_func(node->extension, renderer, node, ev_type, options);
return 1;
}
switch (node->type) {
case CMARK_NODE_DOCUMENT:
break;
case CMARK_NODE_BLOCK_QUOTE:
if (entering) {
cmark_html_render_cr(html);
cmark_strbuf_puts(html, "<blockquote");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, ">\n");
} else {
cmark_html_render_cr(html);
cmark_strbuf_puts(html, "</blockquote>\n");
}
break;
case CMARK_NODE_LIST: {
cmark_list_type list_type = node->as.list.list_type;
int start = node->as.list.start;
if (entering) {
cmark_html_render_cr(html);
if (list_type == CMARK_BULLET_LIST) {
cmark_strbuf_puts(html, "<ul");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, ">\n");
} else if (start == 1) {
cmark_strbuf_puts(html, "<ol");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, ">\n");
} else {
snprintf(buffer, BUFFER_SIZE, "<ol start=\"%d\"", start);
cmark_strbuf_puts(html, buffer);
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, ">\n");
}
} else {
cmark_strbuf_puts(html,
list_type == CMARK_BULLET_LIST ? "</ul>\n" : "</ol>\n");
}
break;
}
case CMARK_NODE_ITEM:
if (entering) {
cmark_html_render_cr(html);
cmark_strbuf_puts(html, "<li");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_putc(html, '>');
} else {
cmark_strbuf_puts(html, "</li>\n");
}
break;
case CMARK_NODE_HEADING:
if (entering) {
cmark_html_render_cr(html);
start_heading[2] = (char)('0' + node->as.heading.level);
cmark_strbuf_puts(html, start_heading);
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_putc(html, '>');
} else {
end_heading[3] = (char)('0' + node->as.heading.level);
cmark_strbuf_puts(html, end_heading);
cmark_strbuf_puts(html, ">\n");
}
break;
case CMARK_NODE_CODE_BLOCK:
cmark_html_render_cr(html);
if (node->as.code.info.len == 0) {
cmark_strbuf_puts(html, "<pre");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, "><code>");
} else {
bufsize_t first_tag = 0;
while (first_tag < node->as.code.info.len &&
!cmark_isspace(node->as.code.info.data[first_tag])) {
first_tag += 1;
}
if (options & CMARK_OPT_GITHUB_PRE_LANG) {
cmark_strbuf_puts(html, "<pre");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, " lang=\"");
escape_html(html, node->as.code.info.data, first_tag);
if (first_tag < node->as.code.info.len && (options & CMARK_OPT_FULL_INFO_STRING)) {
cmark_strbuf_puts(html, "\" data-meta=\"");
escape_html(html, node->as.code.info.data + first_tag + 1, node->as.code.info.len - first_tag - 1);
}
cmark_strbuf_puts(html, "\"><code>");
} else {
cmark_strbuf_puts(html, "<pre");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, "><code class=\"language-");
escape_html(html, node->as.code.info.data, first_tag);
if (first_tag < node->as.code.info.len && (options & CMARK_OPT_FULL_INFO_STRING)) {
cmark_strbuf_puts(html, "\" data-meta=\"");
escape_html(html, node->as.code.info.data + first_tag + 1, node->as.code.info.len - first_tag - 1);
}
cmark_strbuf_puts(html, "\">");
}
}
escape_html(html, node->as.code.literal.data, node->as.code.literal.len);
cmark_strbuf_puts(html, "</code></pre>\n");
break;
case CMARK_NODE_HTML_BLOCK:
cmark_html_render_cr(html);
if (!(options & CMARK_OPT_UNSAFE)) {
cmark_strbuf_puts(html, "<!-- raw HTML omitted -->");
} else if (renderer->filter_extensions) {
filter_html_block(renderer, node->as.literal.data, node->as.literal.len);
} else {
cmark_strbuf_put(html, node->as.literal.data, node->as.literal.len);
}
cmark_html_render_cr(html);
break;
case CMARK_NODE_CUSTOM_BLOCK:
cmark_html_render_cr(html);
if (entering) {
cmark_strbuf_put(html, node->as.custom.on_enter.data,
node->as.custom.on_enter.len);
} else {
cmark_strbuf_put(html, node->as.custom.on_exit.data,
node->as.custom.on_exit.len);
}
cmark_html_render_cr(html);
break;
case CMARK_NODE_THEMATIC_BREAK:
cmark_html_render_cr(html);
cmark_strbuf_puts(html, "<hr");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_puts(html, " />\n");
break;
case CMARK_NODE_PARAGRAPH:
parent = cmark_node_parent(node);
grandparent = cmark_node_parent(parent);
if (grandparent != NULL && grandparent->type == CMARK_NODE_LIST) {
tight = grandparent->as.list.tight;
} else {
tight = false;
}
if (!tight) {
if (entering) {
cmark_html_render_cr(html);
cmark_strbuf_puts(html, "<p");
cmark_html_render_sourcepos(node, html, options);
cmark_strbuf_putc(html, '>');
} else {
if (parent->type == CMARK_NODE_FOOTNOTE_DEFINITION && node->next == NULL) {
cmark_strbuf_putc(html, ' ');
S_put_footnote_backref(renderer, html);
}
cmark_strbuf_puts(html, "</p>\n");
}
}
break;
case CMARK_NODE_TEXT:
escape_html(html, node->as.literal.data, node->as.literal.len);
break;
case CMARK_NODE_LINEBREAK:
cmark_strbuf_puts(html, "<br />\n");
break;
case CMARK_NODE_SOFTBREAK:
if (options & CMARK_OPT_HARDBREAKS) {
cmark_strbuf_puts(html, "<br />\n");
} else if (options & CMARK_OPT_NOBREAKS) {
cmark_strbuf_putc(html, ' ');
} else {
cmark_strbuf_putc(html, '\n');
}
break;
case CMARK_NODE_CODE:
cmark_strbuf_puts(html, "<code>");
escape_html(html, node->as.literal.data, node->as.literal.len);
cmark_strbuf_puts(html, "</code>");
break;
case CMARK_NODE_HTML_INLINE:
if (!(options & CMARK_OPT_UNSAFE)) {
cmark_strbuf_puts(html, "<!-- raw HTML omitted -->");
} else {
filtered = false;
for (it = renderer->filter_extensions; it; it = it->next) {
ext = (cmark_syntax_extension *) it->data;
if (!ext->html_filter_func(ext, node->as.literal.data, node->as.literal.len)) {
filtered = true;
break;
}
}
if (!filtered) {
cmark_strbuf_put(html, node->as.literal.data, node->as.literal.len);
} else {
cmark_strbuf_puts(html, "<");
cmark_strbuf_put(html, node->as.literal.data + 1, node->as.literal.len - 1);
}
}
break;
case CMARK_NODE_CUSTOM_INLINE:
if (entering) {
cmark_strbuf_put(html, node->as.custom.on_enter.data,
node->as.custom.on_enter.len);
} else {
cmark_strbuf_put(html, node->as.custom.on_exit.data,
node->as.custom.on_exit.len);
}
break;
case CMARK_NODE_STRONG:
if (entering) {
cmark_strbuf_puts(html, "<strong>");
} else {
cmark_strbuf_puts(html, "</strong>");
}
break;
case CMARK_NODE_EMPH:
if (entering) {
cmark_strbuf_puts(html, "<em>");
} else {
cmark_strbuf_puts(html, "</em>");
}
break;
case CMARK_NODE_LINK:
if (entering) {
cmark_strbuf_puts(html, "<a href=\"");
if (!(!(options & CMARK_OPT_UNSAFE) &&
scan_dangerous_url(&node->as.link.url, 0))) {
houdini_escape_href(html, node->as.link.url.data,
node->as.link.url.len);
}
if (node->as.link.title.len) {
cmark_strbuf_puts(html, "\" title=\"");
escape_html(html, node->as.link.title.data, node->as.link.title.len);
}
cmark_strbuf_puts(html, "\">");
} else {
cmark_strbuf_puts(html, "</a>");
}
break;
case CMARK_NODE_IMAGE:
if (entering) {
cmark_strbuf_puts(html, "<img src=\"");
if (!(!(options & CMARK_OPT_UNSAFE) &&
scan_dangerous_url(&node->as.link.url, 0))) {
houdini_escape_href(html, node->as.link.url.data,
node->as.link.url.len);
}
cmark_strbuf_puts(html, "\" alt=\"");
renderer->plain = node;
} else {
if (node->as.link.title.len) {
cmark_strbuf_puts(html, "\" title=\"");
escape_html(html, node->as.link.title.data, node->as.link.title.len);
}
cmark_strbuf_puts(html, "\" />");
}
break;
case CMARK_NODE_FOOTNOTE_DEFINITION:
if (entering) {
if (renderer->footnote_ix == 0) {
cmark_strbuf_puts(html, "<section class=\"footnotes\">\n<ol>\n");
}
++renderer->footnote_ix;
cmark_strbuf_puts(html, "<li id=\"fn");
char n[32];
snprintf(n, sizeof(n), "%d", renderer->footnote_ix);
cmark_strbuf_puts(html, n);
cmark_strbuf_puts(html, "\">\n");
} else {
if (S_put_footnote_backref(renderer, html)) {
cmark_strbuf_putc(html, '\n');
}
cmark_strbuf_puts(html, "</li>\n");
}
break;
case CMARK_NODE_FOOTNOTE_REFERENCE:
if (entering) {
cmark_strbuf_puts(html, "<sup class=\"footnote-ref\"><a href=\"#fn");
cmark_strbuf_put(html, node->as.literal.data, node->as.literal.len);
cmark_strbuf_puts(html, "\" id=\"fnref");
cmark_strbuf_put(html, node->as.literal.data, node->as.literal.len);
cmark_strbuf_puts(html, "\">");
cmark_strbuf_put(html, node->as.literal.data, node->as.literal.len);
cmark_strbuf_puts(html, "</a></sup>");
}
break;
default:
assert(false);
break;
}
return 1;
}
void cmark_strbuf_puts(cmark_strbuf *buf, const char *string) {
cmark_strbuf_put(buf, (const unsigned char *)string,
cmark_strbuf_safe_strlen(string));
}
static void S_out(cmark_renderer *renderer, cmark_node *node,
const char *source, bool wrap,
cmark_escaping escape) {
int length = (int)strlen(source);
unsigned char nextc;
int32_t c;
int i = 0;
int last_nonspace;
int len;
cmark_chunk remainder = cmark_chunk_literal("");
int k = renderer->buffer->size - 1;
cmark_syntax_extension *ext = NULL;
cmark_node *n = node;
while (n && !ext) {
ext = n->extension;
if (!ext)
n = n->parent;
}
if (ext && !ext->commonmark_escape_func)
ext = NULL;
wrap = wrap && !renderer->no_linebreaks;
if (renderer->in_tight_list_item && renderer->need_cr > 1) {
renderer->need_cr = 1;
}
while (renderer->need_cr) {
if (k < 0 || renderer->buffer->ptr[k] == '\n') {
k -= 1;
} else {
cmark_strbuf_putc(renderer->buffer, '\n');
if (renderer->need_cr > 1) {
cmark_strbuf_put(renderer->buffer, renderer->prefix->ptr,
renderer->prefix->size);
}
}
renderer->column = 0;
renderer->begin_line = true;
renderer->begin_content = true;
renderer->need_cr -= 1;
}
while (i < length) {
if (renderer->begin_line) {
cmark_strbuf_put(renderer->buffer, renderer->prefix->ptr,
renderer->prefix->size);
// note: this assumes prefix is ascii:
renderer->column = renderer->prefix->size;
}
len = cmark_utf8proc_iterate((const uint8_t *)source + i, length - i, &c);
if (len == -1) { // error condition
return; // return without rendering rest of string
}
if (ext && ext->commonmark_escape_func(ext, node, c))
cmark_strbuf_putc(renderer->buffer, '\\');
nextc = source[i + len];
if (c == 32 && wrap) {
if (!renderer->begin_line) {
last_nonspace = renderer->buffer->size;
cmark_strbuf_putc(renderer->buffer, ' ');
renderer->column += 1;
renderer->begin_line = false;
renderer->begin_content = false;
// skip following spaces
while (source[i + 1] == ' ') {
i++;
}
// We don't allow breaks that make a digit the first character
// because this causes problems with commonmark output.
if (!cmark_isdigit(source[i + 1])) {
renderer->last_breakable = last_nonspace;
}
}
} else if (c == 10) {
cmark_strbuf_putc(renderer->buffer, '\n');
renderer->column = 0;
renderer->begin_line = true;
renderer->begin_content = true;
renderer->last_breakable = 0;
} else if (escape == LITERAL) {
cmark_render_code_point(renderer, c);
renderer->begin_line = false;
// we don't set 'begin_content' to false til we've
// finished parsing a digit. Reason: in commonmark
// we need to escape a potential list marker after
// a digit:
renderer->begin_content =
renderer->begin_content && cmark_isdigit((char)c) == 1;
} else {
(renderer->outc)(renderer, node, escape, c, nextc);
renderer->begin_line = false;
renderer->begin_content =
renderer->begin_content && cmark_isdigit((char)c) == 1;
}
// If adding the character went beyond width, look for an
// earlier place where the line could be broken:
if (renderer->width > 0 && renderer->column > renderer->width &&
!renderer->begin_line && renderer->last_breakable > 0) {
// copy from last_breakable to remainder
cmark_chunk_set_cstr(renderer->mem, &remainder,
(char *)renderer->buffer->ptr +
renderer->last_breakable + 1);
// truncate at last_breakable
cmark_strbuf_truncate(renderer->buffer, renderer->last_breakable);
// add newline, prefix, and remainder
cmark_strbuf_putc(renderer->buffer, '\n');
cmark_strbuf_put(renderer->buffer, renderer->prefix->ptr,
renderer->prefix->size);
cmark_strbuf_put(renderer->buffer, remainder.data, remainder.len);
renderer->column = renderer->prefix->size + remainder.len;
cmark_chunk_free(renderer->mem, &remainder);
renderer->last_breakable = 0;
renderer->begin_line = false;
renderer->begin_content = false;
}
i += len;
}
}
