static cmark_node *url_match(cmark_parser *parser, cmark_node *parent,
cmark_inline_parser *inline_parser) {
size_t link_end, domain_len;
int rewind = 0;
cmark_chunk *chunk = cmark_inline_parser_get_chunk(inline_parser);
int max_rewind = cmark_inline_parser_get_offset(inline_parser);
uint8_t *data = chunk->data + max_rewind;
size_t size = chunk->len - max_rewind;
if (size < 4 || data[1] != '/' || data[2] != '/')
return 0;
while (rewind < max_rewind && cmark_isalpha(data[-rewind - 1]))
rewind++;
if (!sd_autolink_issafe(data - rewind, size + rewind))
return 0;
link_end = strlen("://");
domain_len = check_domain(data + link_end, size - link_end, 1);
if (domain_len == 0)
return 0;
link_end += domain_len;
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_unput(parent, rewind);
cmark_node *node = cmark_node_new_with_mem(CMARK_NODE_LINK, parser->mem);
cmark_chunk url = cmark_chunk_dup(chunk, max_rewind - rewind,
(bufsize_t)(link_end + rewind));
node->as.link.url = url;
cmark_node *text = cmark_node_new_with_mem(CMARK_NODE_TEXT, parser->mem);
text->as.literal = url;
cmark_node_append_child(node, text);
return node;
}
static CMARK_INLINE void outc(cmark_renderer *renderer, cmark_escaping escape,
int32_t c, unsigned char nextc) {
bool needs_escaping = false;
bool follows_digit =
renderer->buffer->size > 0 &&
cmark_isdigit(renderer->buffer->ptr[renderer->buffer->size - 1]);
char encoded[ENCODED_SIZE];
needs_escaping =
c < 0x80 && escape != LITERAL &&
((escape == NORMAL &&
(c < 0x20 ||
c == '*' || c == '_' || c == '[' || c == ']' || c == '#' || c == '<' ||
c == '>' || c == '\\' || c == '`' || c == '!' ||
(c == '&' && cmark_isalpha(nextc)) || (c == '!' && nextc == '[') ||
(renderer->begin_content && (c == '-' || c == '+' || c == '=') &&
// begin_content doesn't get set to false til we've passed digits
// at the beginning of line, so...
!follows_digit) ||
(renderer->begin_content && (c == '.' || c == ')') && follows_digit &&
(nextc == 0 || cmark_isspace(nextc))))) ||
(escape == URL &&
(c == '`' || c == '<' || c == '>' || cmark_isspace(c) || c == '\\' ||
c == ')' || c == '(')) ||
(escape == TITLE &&
(c == '`' || c == '<' || c == '>' || c == '"' || c == '\\')));
if (needs_escaping) {
if (escape == URL && cmark_isspace(c)) {
// use percent encoding for spaces
snprintf(encoded, ENCODED_SIZE, "%%%2X", c);
cmark_strbuf_puts(renderer->buffer, encoded);
renderer->column += 3;
} else if (cmark_ispunct(c)) {
cmark_render_ascii(renderer, "\\");
cmark_render_code_point(renderer, c);
} else { // render as entity
snprintf(encoded, ENCODED_SIZE, "&#%d;", c);
cmark_strbuf_puts(renderer->buffer, encoded);
renderer->column += strlen(encoded);
}
} else {
cmark_render_code_point(renderer, c);
}
}
static size_t autolink_delim(uint8_t *data, size_t link_end) {
uint8_t cclose, copen;
size_t i;
for (i = 0; i < link_end; ++i)
if (data[i] == '<') {
link_end = i;
break;
}
while (link_end > 0) {
cclose = data[link_end - 1];
switch (cclose) {
case ')':
copen = '(';
break;
default:
copen = 0;
}
if (strchr("?!.,:*_~'\"", data[link_end - 1]) != NULL)
link_end--;
else if (data[link_end - 1] == ';') {
size_t new_end = link_end - 2;
while (new_end > 0 && cmark_isalpha(data[new_end]))
new_end--;
if (new_end < link_end - 2 && data[new_end] == '&')
link_end = new_end;
else
link_end--;
} else if (copen != 0) {
size_t closing = 0;
size_t opening = 0;
i = 0;
/* Allow any number of matching brackets (as recognised in copen/cclose)
* at the end of the URL. If there is a greater number of closing
* brackets than opening ones, we remove one character from the end of
* the link.
*
* Examples (input text => output linked portion):
*
* http://www.pokemon.com/Pikachu_(Electric)
* => http://www.pokemon.com/Pikachu_(Electric)
*
* http://www.pokemon.com/Pikachu_((Electric)
* => http://www.pokemon.com/Pikachu_((Electric)
*
* http://www.pokemon.com/Pikachu_(Electric))
* => http://www.pokemon.com/Pikachu_(Electric)
*
* http://www.pokemon.com/Pikachu_((Electric))
* => http://www.pokemon.com/Pikachu_((Electric))
*/
while (i < link_end) {
if (data[i] == copen)
opening++;
else if (data[i] == cclose)
closing++;
i++;
}
if (closing <= opening)
break;
link_end--;
} else
break;
}
return link_end;
}
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);
}
