static void remove_trailing_blank_lines(cmark_strbuf *ln)
{
int i;
unsigned char c;
for (i = ln->size - 1; i >= 0; --i) {
c = ln->ptr[i];
if (c != ' ' && c != '\t' && c != '\r' && c != '\n')
break;
}
if (i < 0) {
cmark_strbuf_clear(ln);
return;
}
for(; i < ln->size; ++i) {
c = ln->ptr[i];
if (c != '\r' && c != '\n')
continue;
cmark_strbuf_truncate(ln, i);
break;
}
}
// Destructively unescape a string: remove backslashes before punctuation chars.
extern void cmark_strbuf_unescape(cmark_strbuf *buf) {
bufsize_t r, w;
for (r = 0, w = 0; r < buf->size; ++r) {
if (buf->ptr[r] == '\\' && cmark_ispunct(buf->ptr[r + 1]))
r++;
buf->ptr[w++] = buf->ptr[r];
}
cmark_strbuf_truncate(buf, w);
}
// Destructively modify string, collapsing consecutive
// space and newline characters into a single space.
void cmark_strbuf_normalize_whitespace(cmark_strbuf *s) {
bool last_char_was_space = false;
bufsize_t r, w;
for (r = 0, w = 0; r < s->size; ++r) {
if (cmark_isspace(s->ptr[r])) {
if (!last_char_was_space) {
s->ptr[w++] = ' ';
last_char_was_space = true;
}
} else {
s->ptr[w++] = s->ptr[r];
last_char_was_space = false;
}
}
cmark_strbuf_truncate(s, w);
}
static void remove_trailing_blank_lines(cmark_strbuf *ln)
{
int i;
for (i = ln->size - 1; i >= 0; --i) {
unsigned char c = ln->ptr[i];
if (c != ' ' && c != '\t' && c != '\r' && c != '\n')
break;
}
if (i < 0) {
cmark_strbuf_clear(ln);
return;
}
i = cmark_strbuf_strchr(ln, '\n', i);
if (i >= 0)
cmark_strbuf_truncate(ln, i);
}
static int S_render_node(cmark_renderer *renderer, cmark_node *node,
cmark_event_type ev_type, int options) {
cmark_node *tmp;
int list_number;
cmark_delim_type list_delim;
int numticks;
bool extra_spaces;
int i;
bool entering = (ev_type == CMARK_EVENT_ENTER);
const char *info, *code, *title;
char fencechar[2] = {'\0', '\0'};
size_t info_len, code_len;
char listmarker[LISTMARKER_SIZE];
char *emph_delim;
bool first_in_list_item;
bufsize_t marker_width;
bool allow_wrap = renderer->width > 0 && !(CMARK_OPT_NOBREAKS & options) &&
!(CMARK_OPT_HARDBREAKS & options);
// Don't adjust tight list status til we've started the list.
// Otherwise we loose the blank line between a paragraph and
// a following list.
if (!(node->type == CMARK_NODE_ITEM && node->prev == NULL && entering)) {
tmp = get_containing_block(node);
renderer->in_tight_list_item =
tmp && // tmp might be NULL if there is no containing block
((tmp->type == CMARK_NODE_ITEM &&
cmark_node_get_list_tight(tmp->parent)) ||
(tmp && tmp->parent && tmp->parent->type == CMARK_NODE_ITEM &&
cmark_node_get_list_tight(tmp->parent->parent)));
}
switch (node->type) {
case CMARK_NODE_DOCUMENT:
break;
case CMARK_NODE_BLOCK_QUOTE:
if (entering) {
LIT("> ");
renderer->begin_content = true;
cmark_strbuf_puts(renderer->prefix, "> ");
} else {
cmark_strbuf_truncate(renderer->prefix, renderer->prefix->size - 2);
BLANKLINE();
}
break;
case CMARK_NODE_LIST:
if (!entering && node->next && (node->next->type == CMARK_NODE_CODE_BLOCK ||
node->next->type == CMARK_NODE_LIST)) {
// this ensures that a following indented code block or list will be
// inteprereted correctly.
CR();
LIT("<!-- end list -->");
BLANKLINE();
}
break;
case CMARK_NODE_ITEM:
if (cmark_node_get_list_type(node->parent) == CMARK_BULLET_LIST) {
marker_width = 4;
} else {
list_number = cmark_node_get_list_start(node->parent);
list_delim = cmark_node_get_list_delim(node->parent);
tmp = node;
while (tmp->prev) {
tmp = tmp->prev;
list_number += 1;
}
// we ensure a width of at least 4 so
// we get nice transition from single digits
// to double
snprintf(listmarker, LISTMARKER_SIZE, "%d%s%s", list_number,
list_delim == CMARK_PAREN_DELIM ? ")" : ".",
list_number < 10 ? " " : " ");
marker_width = strlen(listmarker);
}
if (entering) {
if (cmark_node_get_list_type(node->parent) == CMARK_BULLET_LIST) {
LIT(" - ");
renderer->begin_content = true;
} else {
LIT(listmarker);
renderer->begin_content = true;
}
for (i = marker_width; i--;) {
cmark_strbuf_putc(renderer->prefix, ' ');
}
} else {
cmark_strbuf_truncate(renderer->prefix,
renderer->prefix->size - marker_width);
CR();
}
break;
case CMARK_NODE_HEADING:
if (entering) {
for (i = cmark_node_get_heading_level(node); i > 0; i--) {
LIT("#");
}
LIT(" ");
renderer->begin_content = true;
renderer->no_linebreaks = true;
} else {
renderer->no_linebreaks = false;
BLANKLINE();
}
break;
case CMARK_NODE_CODE_BLOCK:
first_in_list_item = node->prev == NULL && node->parent &&
node->parent->type == CMARK_NODE_ITEM;
if (!first_in_list_item) {
BLANKLINE();
}
info = cmark_node_get_fence_info(node);
info_len = strlen(info);
fencechar[0] = strchr(info, '`') == NULL ? '`' : '~';
code = cmark_node_get_literal(node);
code_len = strlen(code);
// use indented form if no info, and code doesn't
// begin or end with a blank line, and code isn't
// first thing in a list item
if (info_len == 0 && (code_len > 2 && !cmark_isspace(code[0]) &&
!(cmark_isspace(code[code_len - 1]) &&
cmark_isspace(code[code_len - 2]))) &&
!first_in_list_item) {
LIT(" ");
cmark_strbuf_puts(renderer->prefix, " ");
OUT(cmark_node_get_literal(node), false, LITERAL);
cmark_strbuf_truncate(renderer->prefix, renderer->prefix->size - 4);
} else {
numticks = longest_backtick_sequence(code) + 1;
if (numticks < 3) {
numticks = 3;
}
for (i = 0; i < numticks; i++) {
LIT(fencechar);
}
LIT(" ");
OUT(info, false, LITERAL);
CR();
OUT(cmark_node_get_literal(node), false, LITERAL);
CR();
for (i = 0; i < numticks; i++) {
LIT(fencechar);
}
}
BLANKLINE();
break;
case CMARK_NODE_HTML_BLOCK:
BLANKLINE();
OUT(cmark_node_get_literal(node), false, LITERAL);
BLANKLINE();
break;
case CMARK_NODE_CUSTOM_BLOCK:
BLANKLINE();
OUT(entering ? cmark_node_get_on_enter(node) : cmark_node_get_on_exit(node),
false, LITERAL);
BLANKLINE();
break;
case CMARK_NODE_THEMATIC_BREAK:
BLANKLINE();
LIT("-----");
BLANKLINE();
break;
case CMARK_NODE_PARAGRAPH:
if (!entering) {
BLANKLINE();
}
break;
case CMARK_NODE_TEXT:
OUT(cmark_node_get_literal(node), allow_wrap, NORMAL);
break;
case CMARK_NODE_LINEBREAK:
if (!(CMARK_OPT_HARDBREAKS & options)) {
LIT(" ");
}
CR();
break;
case CMARK_NODE_SOFTBREAK:
if (CMARK_OPT_HARDBREAKS & options) {
LIT(" ");
CR();
} else if (!renderer->no_linebreaks && renderer->width == 0 &&
!(CMARK_OPT_HARDBREAKS & options) &&
!(CMARK_OPT_NOBREAKS & options)) {
CR();
} else {
OUT(" ", allow_wrap, LITERAL);
}
break;
case CMARK_NODE_CODE:
code = cmark_node_get_literal(node);
code_len = strlen(code);
numticks = shortest_unused_backtick_sequence(code);
extra_spaces = code_len == 0 ||
code[0] == '`' || code[code_len - 1] == '`' ||
code[0] == ' ' || code[code_len - 1] == ' ';
for (i = 0; i < numticks; i++) {
LIT("`");
}
if (extra_spaces) {
LIT(" ");
}
OUT(cmark_node_get_literal(node), allow_wrap, LITERAL);
if (extra_spaces) {
LIT(" ");
}
for (i = 0; i < numticks; i++) {
LIT("`");
}
break;
case CMARK_NODE_HTML_INLINE:
OUT(cmark_node_get_literal(node), false, LITERAL);
break;
case CMARK_NODE_CUSTOM_INLINE:
OUT(entering ? cmark_node_get_on_enter(node) : cmark_node_get_on_exit(node),
false, LITERAL);
break;
case CMARK_NODE_STRONG:
if (entering) {
LIT("**");
} else {
LIT("**");
}
break;
case CMARK_NODE_EMPH:
// If we have EMPH(EMPH(x)), we need to use *_x_*
// because **x** is STRONG(x):
if (node->parent && node->parent->type == CMARK_NODE_EMPH &&
node->next == NULL && node->prev == NULL) {
emph_delim = "_";
} else {
emph_delim = "*";
}
if (entering) {
LIT(emph_delim);
} else {
LIT(emph_delim);
}
break;
case CMARK_NODE_LINK:
if (is_autolink(node)) {
if (entering) {
LIT("<");
if (strncmp(cmark_node_get_url(node), "mailto:", 7) == 0) {
LIT((const char *)cmark_node_get_url(node) + 7);
} else {
LIT((const char *)cmark_node_get_url(node));
}
LIT(">");
// return signal to skip contents of node...
return 0;
}
} else {
if (entering) {
LIT("[");
} else {
LIT("](");
OUT(cmark_node_get_url(node), false, URL);
title = cmark_node_get_title(node);
if (strlen(title) > 0) {
LIT(" \"");
OUT(title, false, TITLE);
LIT("\"");
}
LIT(")");
}
}
break;
case CMARK_NODE_IMAGE:
if (entering) {
LIT("![");
} else {
LIT("](");
OUT(cmark_node_get_url(node), false, URL);
title = cmark_node_get_title(node);
if (strlen(title) > 0) {
OUT(" \"", allow_wrap, LITERAL);
OUT(title, false, TITLE);
LIT("\"");
}
LIT(")");
}
break;
default:
assert(false);
break;
}
return 1;
}
static void
S_process_line(cmark_parser *parser, const unsigned char *buffer, size_t bytes)
{
cmark_node* last_matched_container;
int offset = 0;
int matched = 0;
int lev = 0;
int i;
cmark_list *data = NULL;
bool all_matched = true;
cmark_node* container;
bool blank = false;
int first_nonspace;
int indent;
bool indented;
cmark_chunk input;
bool maybe_lazy;
int trim = 0;
bool cr = false;
bool lf = false;
utf8proc_detab(parser->curline, buffer, bytes);
// Add a newline to the end if not present:
// TODO this breaks abstraction:
if (parser->curline->size > trim &&
parser->curline->ptr[parser->curline->size - 1 - trim] == '\n') {
trim += 1;
lf = true;
}
if (parser->curline->size > trim &&
parser->curline->ptr[parser->curline->size - 1 - trim] == '\r') {
trim += 1;
cr = true;
}
if (cr) {
cmark_strbuf_truncate(parser->curline, parser->curline->size - trim);
}
if (cr || !lf) {
cmark_strbuf_putc(parser->curline, '\n');
}
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;
first_nonspace = offset;
while (peek_at(&input, first_nonspace) == ' ') {
first_nonspace++;
}
indent = first_nonspace - offset;
blank = peek_at(&input, first_nonspace) == '\n' ||
peek_at(&input, first_nonspace) == '\r';
if (container->type == NODE_BLOCK_QUOTE) {
matched = indent <= 3 && peek_at(&input, first_nonspace) == '>';
if (matched) {
offset = first_nonspace + 1;
if (peek_at(&input, offset) == ' ')
offset++;
} else {
all_matched = false;
}
} else if (container->type == NODE_ITEM) {
if (indent >= container->as.list.marker_offset +
container->as.list.padding) {
offset += container->as.list.marker_offset +
container->as.list.padding;
} else if (blank) {
offset = first_nonspace;
} else {
all_matched = false;
}
} else if (container->type == NODE_CODE_BLOCK) {
if (!container->as.code.fenced) { // indented
if (indent >= CODE_INDENT) {
offset += CODE_INDENT;
} else if (blank) {
offset = first_nonspace;
} else {
all_matched = false;
}
} else { // fenced
matched = 0;
if (indent <= 3 &&
(peek_at(&input, first_nonspace) ==
container->as.code.fence_char)) {
matched = scan_close_code_fence(&input,
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;
offset += matched;
parser->current = finalize(parser, container);
goto finished;
} else {
// skip opt. spaces of fence offset
i = container->as.code.fence_offset;
while (i > 0 &&
peek_at(&input, offset) == ' ') {
offset++;
i--;
}
}
}
} else if (container->type == NODE_HEADER) {
// a header can never contain more than one line
all_matched = false;
} else if (container->type == NODE_HTML) {
if (blank) {
all_matched = false;
}
} else if (container->type == NODE_PARAGRAPH) {
if (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 (blank && container->last_line_blank) {
break_out_of_lists(parser, &container);
}
maybe_lazy = parser->current->type == NODE_PARAGRAPH;
// try new container starts:
while (container->type != NODE_CODE_BLOCK &&
container->type != NODE_HTML) {
first_nonspace = offset;
while (peek_at(&input, first_nonspace) == ' ')
first_nonspace++;
indent = first_nonspace - offset;
indented = indent >= CODE_INDENT;
blank = peek_at(&input, first_nonspace) == '\n' ||
peek_at(&input, first_nonspace) == '\r';
if (indented && !maybe_lazy && !blank) {
offset += CODE_INDENT;
container = add_child(parser, container, NODE_CODE_BLOCK, 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 if (!indented && peek_at(&input, first_nonspace) == '>') {
offset = first_nonspace + 1;
// optional following character
if (peek_at(&input, offset) == ' ')
offset++;
container = add_child(parser, container, NODE_BLOCK_QUOTE, offset + 1);
} else if (!indented && (matched = scan_atx_header_start(&input, first_nonspace))) {
offset = first_nonspace + matched;
container = add_child(parser, container, NODE_HEADER, offset + 1);
int hashpos = cmark_chunk_strchr(&input, '#', first_nonspace);
int level = 0;
while (peek_at(&input, hashpos) == '#') {
level++;
hashpos++;
}
container->as.header.level = level;
container->as.header.setext = false;
} else if (!indented && (matched = scan_open_code_fence(&input, first_nonspace))) {
container = add_child(parser, container, NODE_CODE_BLOCK, first_nonspace + 1);
container->as.code.fenced = true;
container->as.code.fence_char = peek_at(&input, first_nonspace);
container->as.code.fence_length = matched;
container->as.code.fence_offset = first_nonspace - offset;
container->as.code.info = cmark_chunk_literal("");
offset = first_nonspace + matched;
} else if (!indented && (matched = scan_html_block_tag(&input, first_nonspace))) {
container = add_child(parser, container, NODE_HTML, first_nonspace + 1);
// note, we don't adjust offset because the tag is part of the text
} else if (!indented &&
container->type == NODE_PARAGRAPH &&
(lev = scan_setext_header_line(&input, first_nonspace)) &&
// check that there is only one line in the paragraph:
(cmark_strbuf_strrchr(&container->string_content, '\n',
cmark_strbuf_len(&container->string_content) - 2) < 0 &&
cmark_strbuf_strrchr(&container->string_content, '\r',
cmark_strbuf_len(&container->string_content) - 2) < 0)) {
container->type = NODE_HEADER;
container->as.header.level = lev;
container->as.header.setext = true;
offset = input.len - 1;
} else if (!indented &&
!(container->type == NODE_PARAGRAPH &&
!all_matched) &&
(matched = scan_hrule(&input, first_nonspace))) {
// it's only now that we know the line is not part of a setext header:
container = add_child(parser, container, NODE_HRULE, first_nonspace + 1);
container = finalize(parser, container);
offset = input.len - 1;
} else if ((matched = parse_list_marker(&input, first_nonspace, &data))) {
// compute padding:
offset = first_nonspace + matched;
i = 0;
while (i <= 5 && peek_at(&input, offset + i) == ' ') {
i++;
}
// i = number of spaces after marker, up to 5
if (i >= 5 || i < 1 ||
peek_at(&input, offset) == '\n' ||
peek_at(&input, offset) == '\r') {
data->padding = matched + 1;
if (i > 0) {
offset += 1;
}
} else {
data->padding = matched + i;
offset += i;
}
// check container; if it's a list, see if this list item
// can continue the list; otherwise, create a list container.
data->marker_offset = indent;
if (container->type != NODE_LIST ||
!lists_match(&container->as.list, data)) {
container = add_child(parser, container, NODE_LIST,
first_nonspace + 1);
memcpy(&container->as.list, data, sizeof(*data));
}
// add the list item
container = add_child(parser, container, NODE_ITEM,
first_nonspace + 1);
/* TODO: static */
memcpy(&container->as.list, data, sizeof(*data));
free(data);
} 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 offset is a text line. add the text to the
// appropriate container.
first_nonspace = offset;
while (peek_at(&input, first_nonspace) == ' ')
first_nonspace++;
indent = first_nonspace - offset;
blank = peek_at(&input, first_nonspace) == '\n' ||
peek_at(&input, first_nonspace) == '\r';
if (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 = (blank &&
container->type != NODE_BLOCK_QUOTE &&
container->type != NODE_HEADER &&
!(container->type == NODE_CODE_BLOCK &&
container->as.code.fenced) &&
!(container->type == 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 &&
!blank &&
parser->current->type == NODE_PARAGRAPH &&
cmark_strbuf_len(&parser->current->string_content) > 0) {
add_line(parser->current, &input, 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 == NODE_CODE_BLOCK ||
container->type == NODE_HTML) {
add_line(container, &input, offset);
} else if (blank) {
// ??? do nothing
} else if (accepts_lines(container->type)) {
if (container->type == NODE_HEADER &&
container->as.header.setext == false) {
chop_trailing_hashtags(&input);
}
add_line(container, &input, first_nonspace);
} else {
// create paragraph container for line
container = add_child(parser, container, NODE_PARAGRAPH, first_nonspace + 1);
add_line(container, &input, first_nonspace);
}
parser->current = container;
}
finished:
parser->last_line_length = parser->curline->size;
if (parser->last_line_length &&
parser->curline->ptr[parser->last_line_length - 1] == '\n')
parser->last_line_length--;
if (parser->last_line_length &&
parser->curline->ptr[parser->last_line_length - 1] == '\r')
parser->last_line_length--;
cmark_strbuf_clear(parser->curline);
}
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 int
S_render_node(cmark_renderer *renderer,
cmark_node *node,
cmark_event_type ev_type,
int options)
{
cmark_node *tmp;
int list_number;
cmark_delim_type list_delim;
int numticks;
int i;
bool entering = (ev_type == CMARK_EVENT_ENTER);
const char *info, *code, *title;
size_t info_len, code_len;
cmark_strbuf listmarker = GH_BUF_INIT;
char *emph_delim;
bufsize_t marker_width;
// Don't adjust tight list status til we've started the list.
// Otherwise we loose the blank line between a paragraph and
// a following list.
if (!(node->type == CMARK_NODE_ITEM && node->prev == NULL &&
entering)) {
tmp = get_containing_block(node);
renderer->in_tight_list_item =
(tmp->type == CMARK_NODE_ITEM &&
cmark_node_get_list_tight(tmp->parent)) ||
(tmp &&
tmp->parent &&
tmp->parent->type == CMARK_NODE_ITEM &&
cmark_node_get_list_tight(tmp->parent->parent));
}
switch (node->type) {
case CMARK_NODE_DOCUMENT:
break;
case CMARK_NODE_BLOCK_QUOTE:
if (entering) {
LIT("> ");
cmark_strbuf_puts(renderer->prefix, "> ");
} else {
cmark_strbuf_truncate(renderer->prefix,
renderer->prefix->size - 2);
BLANKLINE();
}
break;
case CMARK_NODE_LIST:
if (!entering && node->next &&
(node->next->type == CMARK_NODE_CODE_BLOCK ||
node->next->type == CMARK_NODE_LIST)) {
// this ensures 2 blank lines after list,
// if before code block or list:
LIT("\n");
}
break;
case CMARK_NODE_ITEM:
if (cmark_node_get_list_type(node->parent) ==
CMARK_BULLET_LIST) {
marker_width = 2;
} else {
list_number = cmark_node_get_list_start(node->parent);
list_delim = cmark_node_get_list_delim(node->parent);
tmp = node;
while (tmp->prev) {
tmp = tmp->prev;
list_number += 1;
}
// we ensure a width of at least 4 so
// we get nice transition from single digits
// to double
cmark_strbuf_printf(&listmarker,
"%d%s%s", list_number,
list_delim == CMARK_PAREN_DELIM ?
")" : ".",
list_number < 10 ? " " : " ");
marker_width = listmarker.size;
}
if (entering) {
if (cmark_node_get_list_type(node->parent) ==
CMARK_BULLET_LIST) {
LIT("* ");
cmark_strbuf_puts(renderer->prefix, " ");
} else {
LIT((char *)listmarker.ptr);
for (i = marker_width; i--;) {
cmark_strbuf_putc(renderer->prefix, ' ');
}
}
} else {
cmark_strbuf_truncate(renderer->prefix,
renderer->prefix->size -
marker_width);
CR();
}
cmark_strbuf_free(&listmarker);
break;
case CMARK_NODE_HEADER:
if (entering) {
for (int i = cmark_node_get_header_level(node); i > 0; i--) {
LIT("#");
}
LIT(" ");
renderer->no_wrap = true;
} else {
renderer->no_wrap = false;
BLANKLINE();
}
break;
case CMARK_NODE_CODE_BLOCK:
BLANKLINE();
info = cmark_node_get_fence_info(node);
info_len = safe_strlen(info);
code = cmark_node_get_literal(node);
code_len = safe_strlen(code);
// use indented form if no info, and code doesn't
// begin or end with a blank line, and code isn't
// first thing in a list item
if (info_len == 0 &&
(code_len > 2 &&
!isspace(code[0]) &&
!(isspace(code[code_len - 1]) &&
isspace(code[code_len - 2]))) &&
!(node->prev == NULL && node->parent &&
node->parent->type == CMARK_NODE_ITEM)) {
LIT(" ");
cmark_strbuf_puts(renderer->prefix, " ");
OUT(cmark_node_get_literal(node), false, LITERAL);
cmark_strbuf_truncate(renderer->prefix,
renderer->prefix->size - 4);
} else {
numticks = longest_backtick_sequence(code) + 1;
if (numticks < 3) {
numticks = 3;
}
for (i = 0; i < numticks; i++) {
LIT("`");
}
LIT(" ");
OUT(info, false, LITERAL);
CR();
OUT(cmark_node_get_literal(node), false, LITERAL);
CR();
for (i = 0; i < numticks; i++) {
LIT("`");
}
}
BLANKLINE();
break;
case CMARK_NODE_HTML:
BLANKLINE();
OUT(cmark_node_get_literal(node), false, LITERAL);
BLANKLINE();
break;
case CMARK_NODE_HRULE:
BLANKLINE();
LIT("-----");
BLANKLINE();
break;
case CMARK_NODE_PARAGRAPH:
if (!entering) {
BLANKLINE();
}
break;
case CMARK_NODE_TEXT:
OUT(cmark_node_get_literal(node), true, NORMAL);
break;
case CMARK_NODE_LINEBREAK:
if (!(CMARK_OPT_HARDBREAKS & options)) {
LIT("\\");
}
CR();
break;
case CMARK_NODE_SOFTBREAK:
if (renderer->width == 0 &&
!(CMARK_OPT_HARDBREAKS & options)) {
CR();
} else {
OUT(" ", true, LITERAL);
}
break;
case CMARK_NODE_CODE:
code = cmark_node_get_literal(node);
code_len = safe_strlen(code);
numticks = shortest_unused_backtick_sequence(code);
for (i = 0; i < numticks; i++) {
LIT("`");
}
if (code_len == 0 || code[0] == '`') {
LIT(" ");
}
OUT(cmark_node_get_literal(node), true, LITERAL);
if (code_len == 0 || code[code_len - 1] == '`') {
LIT(" ");
}
for (i = 0; i < numticks; i++) {
LIT("`");
}
break;
case CMARK_NODE_INLINE_HTML:
OUT(cmark_node_get_literal(node), false, LITERAL);
break;
case CMARK_NODE_STRONG:
if (entering) {
LIT("**");
} else {
LIT("**");
}
break;
case CMARK_NODE_EMPH:
// If we have EMPH(EMPH(x)), we need to use *_x_*
// because **x** is STRONG(x):
if (node->parent && node->parent->type == CMARK_NODE_EMPH &&
node->next == NULL && node->prev == NULL) {
emph_delim = "_";
} else {
emph_delim = "*";
}
if (entering) {
LIT(emph_delim);
} else {
LIT(emph_delim);
}
break;
case CMARK_NODE_LINK:
if (is_autolink(node)) {
if (entering) {
LIT("<");
if (strncmp(cmark_node_get_url(node),
"mailto:", 7) == 0) {
LIT((char *)cmark_node_get_url(node) + 7);
} else {
LIT((char *)cmark_node_get_url(node));
}
LIT(">");
// return signal to skip contents of node...
return 0;
}
} else {
if (entering) {
LIT("[");
} else {
LIT("](");
OUT(cmark_node_get_url(node), false, URL);
title = cmark_node_get_title(node);
if (safe_strlen(title) > 0) {
LIT(" \"");
OUT(title, false, TITLE);
LIT("\"");
}
LIT(")");
}
}
break;
case CMARK_NODE_IMAGE:
if (entering) {
LIT("![");
} else {
LIT("](");
OUT(cmark_node_get_url(node), false, URL);
title = cmark_node_get_title(node);
if (safe_strlen(title) > 0) {
OUT(" \"", true, LITERAL);
OUT(title, false, TITLE);
LIT("\"");
}
LIT(")");
}
break;
default:
assert(false);
break;
}
return 1;
}
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;
}
}
