// Free a cmark_node list and any children.
static
void S_free_nodes(cmark_node *e)
{
cmark_node *next;
while (e != NULL) {
cmark_strbuf_free(&e->string_content);
switch (e->type) {
case NODE_CODE_BLOCK:
cmark_chunk_free(&e->as.code.info);
cmark_chunk_free(&e->as.code.literal);
break;
case NODE_TEXT:
case NODE_INLINE_HTML:
case NODE_CODE:
case NODE_HTML:
cmark_chunk_free(&e->as.literal);
break;
case NODE_LINK:
case NODE_IMAGE:
free(e->as.link.url);
free(e->as.link.title);
break;
default:
break;
}
if (e->last_child) {
// Splice children into list
e->last_child->next = e->next;
e->next = e->first_child;
}
next = e->next;
free(e);
e = next;
}
}
static void process_emphasis(subject *subj, delimiter *start_delim)
{
delimiter *closer = subj->last_delim;
delimiter *opener;
// move back to first relevant delim.
while (closer != NULL && closer->previous != start_delim) {
closer = closer->previous;
}
// now move forward, looking for closers, and handling each
while (closer != NULL) {
if (closer->can_close &&
(closer->delim_char == '*' || closer->delim_char == '_' ||
closer->delim_char == '"' || closer->delim_char == '\'')) {
// Now look backwards for first matching opener:
opener = closer->previous;
while (opener != NULL && opener != start_delim) {
if (opener->delim_char == closer->delim_char &&
opener->can_open) {
break;
}
opener = opener->previous;
}
if (closer->delim_char == '*' || closer->delim_char == '_') {
if (opener != NULL && opener != start_delim) {
closer = S_insert_emph(subj, opener, closer);
} else {
closer = closer->next;
}
} else if (closer->delim_char == '\'') {
cmark_chunk_free(&closer->inl_text->as.literal);
closer->inl_text->as.literal =
cmark_chunk_literal(RIGHTSINGLEQUOTE);
if (opener != NULL && opener != start_delim) {
cmark_chunk_free(&opener->inl_text->as.literal);
opener->inl_text->as.literal =
cmark_chunk_literal(LEFTSINGLEQUOTE);
}
closer = closer->next;
} else if (closer->delim_char == '"') {
cmark_chunk_free(&closer->inl_text->as.literal);
closer->inl_text->as.literal =
cmark_chunk_literal(RIGHTDOUBLEQUOTE);
if (opener != NULL && opener != start_delim) {
cmark_chunk_free(&opener->inl_text->as.literal);
opener->inl_text->as.literal =
cmark_chunk_literal(LEFTDOUBLEQUOTE);
}
closer = closer->next;
}
} else {
closer = closer->next;
}
}
// free all delimiters in list until start_delim:
while (subj->last_delim != start_delim) {
remove_delimiter(subj, subj->last_delim);
}
}
static void reference_free(cmark_reference_map *map, cmark_reference *ref) {
cmark_mem *mem = map->mem;
if (ref != NULL) {
mem->free(ref->label);
cmark_chunk_free(mem, &ref->url);
cmark_chunk_free(mem, &ref->title);
mem->free(ref);
}
}
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);
}
// Free an inline list. Avoid recursion to prevent stack overflows
// on deeply nested structures.
void cmark_free_inlines(cmark_node_inl* e)
{
node_inl * next;
while (e != NULL) {
switch (e->tag){
case CMARK_INL_STRING:
case CMARK_INL_RAW_HTML:
case CMARK_INL_CODE:
cmark_chunk_free(&e->content.literal);
break;
case CMARK_INL_LINEBREAK:
case CMARK_INL_SOFTBREAK:
break;
case CMARK_INL_LINK:
case CMARK_INL_IMAGE:
free(e->content.linkable.url);
free(e->content.linkable.title);
splice_into_list(e, e->content.linkable.label);
break;
case CMARK_INL_EMPH:
case CMARK_INL_STRONG:
splice_into_list(e, e->content.inlines);
break;
default:
fprintf(stderr, "[WARN] (%s:%d) Unknown inline tag %d",
__FILE__, __LINE__, e->tag);
break;
}
next = e->next;
free(e);
e = next;
}
}
// Free a cmark_node list and any children.
static void S_free_nodes(cmark_node *e) {
cmark_node *next;
while (e != NULL) {
cmark_strbuf_free(&e->content);
switch (e->type) {
case CMARK_NODE_CODE_BLOCK:
cmark_chunk_free(NODE_MEM(e), &e->as.code.info);
cmark_chunk_free(NODE_MEM(e), &e->as.code.literal);
break;
case CMARK_NODE_TEXT:
case CMARK_NODE_HTML_INLINE:
case CMARK_NODE_CODE:
case CMARK_NODE_HTML_BLOCK:
cmark_chunk_free(NODE_MEM(e), &e->as.literal);
break;
case CMARK_NODE_LINK:
case CMARK_NODE_IMAGE:
cmark_chunk_free(NODE_MEM(e), &e->as.link.url);
cmark_chunk_free(NODE_MEM(e), &e->as.link.title);
break;
case CMARK_NODE_CUSTOM_BLOCK:
case CMARK_NODE_CUSTOM_INLINE:
cmark_chunk_free(NODE_MEM(e), &e->as.custom.on_enter);
cmark_chunk_free(NODE_MEM(e), &e->as.custom.on_exit);
break;
default:
break;
}
if (e->last_child) {
// Splice children into list
e->last_child->next = e->next;
e->next = e->first_child;
}
next = e->next;
NODE_MEM(e)->free(e);
e = next;
}
}
// Return a link, an image, or a literal close bracket.
static cmark_node* handle_close_bracket(subject* subj, cmark_node *parent)
{
int initial_pos;
int starturl, endurl, starttitle, endtitle, endall;
int n;
int sps;
cmark_reference *ref;
bool is_image = false;
cmark_chunk url_chunk, title_chunk;
unsigned char *url, *title;
delimiter *opener;
cmark_node *link_text;
cmark_node *inl;
cmark_chunk raw_label;
int found_label;
advance(subj); // advance past ]
initial_pos = subj->pos;
// look through list of delimiters for a [ or !
opener = subj->last_delim;
while (opener) {
if (opener->delim_char == '[' || opener->delim_char == '!') {
break;
}
opener = opener->previous;
}
if (opener == NULL) {
return make_str(cmark_chunk_literal("]"));
}
if (!opener->active) {
// take delimiter off stack
remove_delimiter(subj, opener);
return make_str(cmark_chunk_literal("]"));
}
// If we got here, we matched a potential link/image text.
is_image = opener->delim_char == '!';
link_text = opener->inl_text->next;
// Now we check to see if it's a link/image.
// First, look for an inline link.
if (peek_char(subj) == '(' &&
((sps = scan_spacechars(&subj->input, subj->pos + 1)) > -1) &&
((n = scan_link_url(&subj->input, subj->pos + 1 + sps)) > -1)) {
// try to parse an explicit link:
starturl = subj->pos + 1 + sps; // after (
endurl = starturl + n;
starttitle = endurl + scan_spacechars(&subj->input, endurl);
// ensure there are spaces btw url and title
endtitle = (starttitle == endurl) ? starttitle :
starttitle + scan_link_title(&subj->input, starttitle);
endall = endtitle + scan_spacechars(&subj->input, endtitle);
if (peek_at(subj, endall) == ')') {
subj->pos = endall + 1;
url_chunk = cmark_chunk_dup(&subj->input, starturl, endurl - starturl);
title_chunk = cmark_chunk_dup(&subj->input, starttitle, endtitle - starttitle);
url = cmark_clean_url(&url_chunk);
title = cmark_clean_title(&title_chunk);
cmark_chunk_free(&url_chunk);
cmark_chunk_free(&title_chunk);
goto match;
} else {
goto noMatch;
}
}
// Next, look for a following [link label] that matches in refmap.
// skip spaces
subj->pos = subj->pos + scan_spacechars(&subj->input, subj->pos);
raw_label = cmark_chunk_literal("");
found_label = link_label(subj, &raw_label);
if (!found_label || raw_label.len == 0) {
cmark_chunk_free(&raw_label);
raw_label = cmark_chunk_dup(&subj->input, opener->position,
initial_pos - opener->position - 1);
}
if (!found_label) {
// If we have a shortcut reference link, back up
// to before the spacse we skipped.
subj->pos = initial_pos;
}
ref = cmark_reference_lookup(subj->refmap, &raw_label);
cmark_chunk_free(&raw_label);
if (ref != NULL) { // found
url = bufdup(ref->url);
title = bufdup(ref->title);
goto match;
} else {
goto noMatch;
}
noMatch:
// If we fall through to here, it means we didn't match a link:
remove_delimiter(subj, opener); // remove this opener from delimiter list
subj->pos = initial_pos;
return make_str(cmark_chunk_literal("]"));
match:
inl = opener->inl_text;
inl->type = is_image ? NODE_IMAGE : NODE_LINK;
cmark_chunk_free(&inl->as.literal);
inl->first_child = link_text;
process_emphasis(subj, opener->previous);
inl->as.link.url = url;
inl->as.link.title = title;
inl->next = NULL;
if (link_text) {
cmark_node *tmp;
link_text->prev = NULL;
for (tmp = link_text; tmp->next != NULL; tmp = tmp->next) {
tmp->parent = inl;
}
tmp->parent = inl;
inl->last_child = tmp;
}
parent->last_child = inl;
// process_emphasis will remove this delimiter and all later ones.
// Now, if we have a link, we also want to deactivate earlier link
// delimiters. (This code can be removed if we decide to allow links
// inside links.)
if (!is_image) {
opener = subj->last_delim;
while (opener != NULL) {
if (opener->delim_char == '[') {
if (!opener->active) {
break;
} else {
opener->active = false;
}
}
opener = opener->previous;
}
}
return NULL;
}
static delimiter*
S_insert_emph(subject *subj, delimiter *opener, delimiter *closer)
{
delimiter *delim, *tmp_delim;
int use_delims;
cmark_node *opener_inl = opener->inl_text;
cmark_node *closer_inl = closer->inl_text;
int opener_num_chars = opener_inl->as.literal.len;
int closer_num_chars = closer_inl->as.literal.len;
cmark_node *tmp, *emph, *first_child, *last_child;
// calculate the actual number of characters used from this closer
if (closer_num_chars < 3 || opener_num_chars < 3) {
use_delims = closer_num_chars <= opener_num_chars ?
closer_num_chars : opener_num_chars;
} else { // closer and opener both have >= 3 characters
use_delims = closer_num_chars % 2 == 0 ? 2 : 1;
}
// remove used characters from associated inlines.
opener_num_chars -= use_delims;
closer_num_chars -= use_delims;
opener_inl->as.literal.len = opener_num_chars;
closer_inl->as.literal.len = closer_num_chars;
// free delimiters between opener and closer
delim = closer->previous;
while (delim != NULL && delim != opener) {
tmp_delim = delim->previous;
remove_delimiter(subj, delim);
delim = tmp_delim;
}
first_child = opener_inl->next;
last_child = closer_inl->prev;
// if opener has 0 characters, remove it and its associated inline
if (opener_num_chars == 0) {
// replace empty opener inline with emph
cmark_chunk_free(&(opener_inl->as.literal));
emph = opener_inl;
emph->type = use_delims == 1 ? NODE_EMPH : NODE_STRONG;
// remove opener from list
remove_delimiter(subj, opener);
} else {
// create new emph or strong, and splice it in to our inlines
// between the opener and closer
emph = use_delims == 1 ? make_emph() : make_strong();
emph->parent = opener_inl->parent;
emph->prev = opener_inl;
opener_inl->next = emph;
}
// push children below emph
emph->next = closer_inl;
closer_inl->prev = emph;
emph->first_child = first_child;
emph->last_child = last_child;
// fix children pointers
first_child->prev = NULL;
last_child->next = NULL;
for (tmp = first_child; tmp != NULL; tmp = tmp->next) {
tmp->parent = emph;
}
// if closer has 0 characters, remove it and its associated inline
if (closer_num_chars == 0) {
// remove empty closer inline
cmark_node_free(closer_inl);
// remove closer from list
tmp_delim = closer->next;
remove_delimiter(subj, closer);
closer = tmp_delim;
}
return closer;
}
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;
}
}
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;
}
}
// Return a link, an image, or a literal close bracket.
static cmark_node *handle_close_bracket(subject *subj) {
bufsize_t initial_pos, after_link_text_pos;
bufsize_t endurl, starttitle, endtitle, endall;
bufsize_t sps, n;
cmark_reference *ref = NULL;
cmark_chunk url_chunk, title_chunk;
cmark_chunk url, title;
bracket *opener;
cmark_node *inl;
cmark_chunk raw_label;
int found_label;
cmark_node *tmp, *tmpnext;
bool is_image;
advance(subj); // advance past ]
initial_pos = subj->pos;
// get last [ or ![
opener = subj->last_bracket;
if (opener == NULL) {
return make_str(subj->mem, cmark_chunk_literal("]"));
}
if (!opener->active) {
// take delimiter off stack
pop_bracket(subj);
return make_str(subj->mem, cmark_chunk_literal("]"));
}
// If we got here, we matched a potential link/image text.
// Now we check to see if it's a link/image.
is_image = opener->image;
after_link_text_pos = subj->pos;
// First, look for an inline link.
if (peek_char(subj) == '(' &&
((sps = scan_spacechars(&subj->input, subj->pos + 1)) > -1) &&
((n = manual_scan_link_url(&subj->input, subj->pos + 1 + sps, &url_chunk)) > -1)) {
// try to parse an explicit link:
endurl = subj->pos + 1 + sps + n;
starttitle = endurl + scan_spacechars(&subj->input, endurl);
// ensure there are spaces btw url and title
endtitle = (starttitle == endurl)
? starttitle
: starttitle + scan_link_title(&subj->input, starttitle);
endall = endtitle + scan_spacechars(&subj->input, endtitle);
if (peek_at(subj, endall) == ')') {
subj->pos = endall + 1;
title_chunk =
cmark_chunk_dup(&subj->input, starttitle, endtitle - starttitle);
url = cmark_clean_url(subj->mem, &url_chunk);
title = cmark_clean_title(subj->mem, &title_chunk);
cmark_chunk_free(subj->mem, &url_chunk);
cmark_chunk_free(subj->mem, &title_chunk);
goto match;
} else {
// it could still be a shortcut reference link
subj->pos = after_link_text_pos;
}
}
// Next, look for a following [link label] that matches in refmap.
// skip spaces
raw_label = cmark_chunk_literal("");
found_label = link_label(subj, &raw_label);
if (!found_label) {
// If we have a shortcut reference link, back up
// to before the spacse we skipped.
subj->pos = initial_pos;
}
if ((!found_label || raw_label.len == 0) && !opener->bracket_after) {
cmark_chunk_free(subj->mem, &raw_label);
raw_label = cmark_chunk_dup(&subj->input, opener->position,
initial_pos - opener->position - 1);
found_label = true;
}
if (found_label) {
ref = cmark_reference_lookup(subj->refmap, &raw_label);
cmark_chunk_free(subj->mem, &raw_label);
}
if (ref != NULL) { // found
url = chunk_clone(subj->mem, &ref->url);
title = chunk_clone(subj->mem, &ref->title);
goto match;
} else {
goto noMatch;
}
noMatch:
// If we fall through to here, it means we didn't match a link:
pop_bracket(subj); // remove this opener from delimiter list
subj->pos = initial_pos;
return make_str(subj->mem, cmark_chunk_literal("]"));
match:
inl = make_simple(subj->mem, is_image ? CMARK_NODE_IMAGE : CMARK_NODE_LINK);
inl->as.link.url = url;
inl->as.link.title = title;
cmark_node_insert_before(opener->inl_text, inl);
// Add link text:
tmp = opener->inl_text->next;
while (tmp) {
tmpnext = tmp->next;
cmark_node_append_child(inl, tmp);
tmp = tmpnext;
}
// Free the bracket [:
cmark_node_free(opener->inl_text);
process_emphasis(subj, opener->previous_delimiter);
pop_bracket(subj);
// Now, if we have a link, we also want to deactivate earlier link
// delimiters. (This code can be removed if we decide to allow links
// inside links.)
if (!is_image) {
opener = subj->last_bracket;
while (opener != NULL) {
if (!opener->image) {
if (!opener->active) {
break;
} else {
opener->active = false;
}
}
opener = opener->previous;
}
}
return NULL;
}
static void process_emphasis(subject *subj, delimiter *stack_bottom) {
delimiter *closer = subj->last_delim;
delimiter *opener;
delimiter *old_closer;
bool opener_found;
int openers_bottom_index;
delimiter *openers_bottom[6] = {stack_bottom, stack_bottom, stack_bottom,
stack_bottom, stack_bottom, stack_bottom};
// move back to first relevant delim.
while (closer != NULL && closer->previous != stack_bottom) {
closer = closer->previous;
}
// now move forward, looking for closers, and handling each
while (closer != NULL) {
if (closer->can_close) {
switch (closer->delim_char) {
case '"':
openers_bottom_index = 0;
break;
case '\'':
openers_bottom_index = 1;
break;
case '_':
openers_bottom_index = 2;
break;
case '*':
openers_bottom_index = 3 + (closer->length % 3);
break;
default:
assert(false);
}
// Now look backwards for first matching opener:
opener = closer->previous;
opener_found = false;
while (opener != NULL && opener != openers_bottom[openers_bottom_index]) {
if (opener->can_open && opener->delim_char == closer->delim_char) {
// interior closer of size 2 can't match opener of size 1
// or of size 1 can't match 2
if (!(closer->can_open || opener->can_close) ||
((opener->length + closer->length) % 3) != 0) {
opener_found = true;
break;
}
}
opener = opener->previous;
}
old_closer = closer;
if (closer->delim_char == '*' || closer->delim_char == '_') {
if (opener_found) {
closer = S_insert_emph(subj, opener, closer);
} else {
closer = closer->next;
}
} else if (closer->delim_char == '\'') {
cmark_chunk_free(subj->mem, &closer->inl_text->as.literal);
closer->inl_text->as.literal = cmark_chunk_literal(RIGHTSINGLEQUOTE);
if (opener_found) {
cmark_chunk_free(subj->mem, &opener->inl_text->as.literal);
opener->inl_text->as.literal = cmark_chunk_literal(LEFTSINGLEQUOTE);
}
closer = closer->next;
} else if (closer->delim_char == '"') {
cmark_chunk_free(subj->mem, &closer->inl_text->as.literal);
closer->inl_text->as.literal = cmark_chunk_literal(RIGHTDOUBLEQUOTE);
if (opener_found) {
cmark_chunk_free(subj->mem, &opener->inl_text->as.literal);
opener->inl_text->as.literal = cmark_chunk_literal(LEFTDOUBLEQUOTE);
}
closer = closer->next;
}
if (!opener_found) {
// set lower bound for future searches for openers
openers_bottom[openers_bottom_index] = old_closer->previous;
if (!old_closer->can_open) {
// we can remove a closer that can't be an
// opener, once we've seen there's no
// matching opener:
remove_delimiter(subj, old_closer);
}
}
} else {
closer = closer->next;
}
}
// free all delimiters in list until stack_bottom:
while (subj->last_delim != NULL && subj->last_delim != stack_bottom) {
remove_delimiter(subj, subj->last_delim);
}
}
static void process_emphasis(subject *subj, delimiter *stack_bottom) {
delimiter *closer = subj->last_delim;
delimiter *opener;
delimiter *old_closer;
bool opener_found;
bool odd_match;
delimiter *openers_bottom[128];
// initialize openers_bottom:
openers_bottom['*'] = stack_bottom;
openers_bottom['_'] = stack_bottom;
openers_bottom['\''] = stack_bottom;
openers_bottom['"'] = stack_bottom;
// move back to first relevant delim.
while (closer != NULL && closer->previous != stack_bottom) {
closer = closer->previous;
}
// now move forward, looking for closers, and handling each
while (closer != NULL) {
if (closer->can_close) {
// Now look backwards for first matching opener:
opener = closer->previous;
opener_found = false;
odd_match = false;
while (opener != NULL && opener != stack_bottom &&
opener != openers_bottom[closer->delim_char]) {
// interior closer of size 2 can't match opener of size 1
// or of size 1 can't match 2
odd_match = (closer->can_open || opener->can_close) &&
((opener->inl_text->as.literal.len +
closer->inl_text->as.literal.len) %
3 ==
0);
if (opener->delim_char == closer->delim_char && opener->can_open &&
!odd_match) {
opener_found = true;
break;
}
opener = opener->previous;
}
old_closer = closer;
if (closer->delim_char == '*' || closer->delim_char == '_') {
if (opener_found) {
closer = S_insert_emph(subj, opener, closer);
} else {
closer = closer->next;
}
} else if (closer->delim_char == '\'') {
cmark_chunk_free(subj->mem, &closer->inl_text->as.literal);
closer->inl_text->as.literal = cmark_chunk_literal(RIGHTSINGLEQUOTE);
if (opener_found) {
cmark_chunk_free(subj->mem, &opener->inl_text->as.literal);
opener->inl_text->as.literal = cmark_chunk_literal(LEFTSINGLEQUOTE);
}
closer = closer->next;
} else if (closer->delim_char == '"') {
cmark_chunk_free(subj->mem, &closer->inl_text->as.literal);
closer->inl_text->as.literal = cmark_chunk_literal(RIGHTDOUBLEQUOTE);
if (opener_found) {
cmark_chunk_free(subj->mem, &opener->inl_text->as.literal);
opener->inl_text->as.literal = cmark_chunk_literal(LEFTDOUBLEQUOTE);
}
closer = closer->next;
}
if (!opener_found && !odd_match) {
// set lower bound for future searches for openers
// (we don't do this with 'odd_match' set because
// a ** that didn't match an earlier * might turn into
// an opener, and the * might be matched by something
// else.
openers_bottom[old_closer->delim_char] = old_closer->previous;
if (!old_closer->can_open) {
// we can remove a closer that can't be an
// opener, once we've seen there's no
// matching opener:
remove_delimiter(subj, old_closer);
}
}
} else {
closer = closer->next;
}
}
// free all delimiters in list until stack_bottom:
while (subj->last_delim != stack_bottom) {
remove_delimiter(subj, subj->last_delim);
}
}
static void process_emphasis(subject *subj, delimiter *stack_bottom)
{
delimiter *closer = subj->last_delim;
delimiter *opener;
delimiter *old_closer;
bool opener_found;
delimiter *openers_bottom[128];
// initialize openers_bottom:
openers_bottom['*'] = stack_bottom;
openers_bottom['_'] = stack_bottom;
openers_bottom['\''] = stack_bottom;
openers_bottom['"'] = stack_bottom;
// move back to first relevant delim.
while (closer != NULL && closer->previous != stack_bottom) {
closer = closer->previous;
}
// now move forward, looking for closers, and handling each
while (closer != NULL) {
if (closer->can_close &&
(closer->delim_char == '*' || closer->delim_char == '_' ||
closer->delim_char == '"' || closer->delim_char == '\'')) {
// Now look backwards for first matching opener:
opener = closer->previous;
opener_found = false;
while (opener != NULL && opener != stack_bottom &&
opener != openers_bottom[closer->delim_char]) {
if (opener->delim_char == closer->delim_char &&
opener->can_open) {
opener_found = true;
break;
}
opener = opener->previous;
}
old_closer = closer;
if (closer->delim_char == '*' || closer->delim_char == '_') {
if (opener_found) {
closer = S_insert_emph(subj, opener, closer);
} else {
closer = closer->next;
}
} else if (closer->delim_char == '\'') {
cmark_chunk_free(&closer->inl_text->as.literal);
closer->inl_text->as.literal =
cmark_chunk_literal(RIGHTSINGLEQUOTE);
if (opener_found) {
cmark_chunk_free(&opener->inl_text->as.literal);
opener->inl_text->as.literal =
cmark_chunk_literal(LEFTSINGLEQUOTE);
}
closer = closer->next;
} else if (closer->delim_char == '"') {
cmark_chunk_free(&closer->inl_text->as.literal);
closer->inl_text->as.literal =
cmark_chunk_literal(RIGHTDOUBLEQUOTE);
if (opener_found) {
cmark_chunk_free(&opener->inl_text->as.literal);
opener->inl_text->as.literal =
cmark_chunk_literal(LEFTDOUBLEQUOTE);
}
closer = closer->next;
}
if (!opener_found) {
// set lower bound for future searches for openers:
openers_bottom[old_closer->delim_char] = old_closer->previous;
if (!old_closer->can_open) {
// we can remove a closer that can't be an
// opener, once we've seen there's no
// matching opener:
remove_delimiter(subj, old_closer);
}
}
} else {
closer = closer->next;
}
}
// free all delimiters in list until stack_bottom:
while (subj->last_delim != stack_bottom) {
remove_delimiter(subj, subj->last_delim);
}
}
