</div>
-5. I haven't worked a lot yet on the MMD-specific features, so there may be
-more changes to come. One thing I do anticipate is that if fenced code blocks
-stay, they will work slightly differently. Currently, an opening fence
-doesn't mean anything unless there is a closing fence that follows it. Again,
-this requires backtracking in the parser. I suspect that an opening fence
-will definitely open a code block. If there is no closing fence, then the
-rest of the document will remain inside the code block. This is the approach
-used by CommonMark and it's a reasonable one, IMO.
+5. "Malformed" reference link definitions are handled slightly differently.
+For example, `Reference Footnotes.text` is parsed differently in compatibility
+mode than MMD-5. This started as a side-effect of the parsing algorithm, but
+I actually think it makes sense. This may or may not change in the future.
## Where Does MultiMarkdown 6 Stand? ##
* Automatic cross-reference targets
* Basic Citation support
* CriticMarkup support
+* Definition lists
+* Footnotes
* Inline and reference footnotes
* Image and Link attributes (attributes can now be used with inline links as
well as reference links)
* Locators required?
* CriticMarkup -- need to decide:
* How to handle CM stretches that include blank lines
-* Definition lists
* Fenced code blocks
-* Footnotes -- need support for multiple blocks inside
* Headers -- need support for manual labels
* Metadata
* Full/Snippet modes
print(">");
+ // If we're printing contents of bracket as text, then ensure we include it all
+ if (text && text->child && text->child->len > 1) {
+ text->child->next->start--;
+ text->child->next->len++;
+ }
+
mmd_export_token_tree_html(out, source, text->child, offset, scratch);
print("</a>");
// Parse options
unsigned long extensions = EXT_SMART | EXT_NOTES | EXT_CRITIC;
- if (a_compatibility->count > 0) {
+ if (a_compatibility->count >= 0) {
// Compatibility mode disables certain features
// Reset extensions
extensions = EXT_COMPATIBILITY | EXT_NO_LABELS | EXT_OBFUSCATE;
// Brackets, Parentheses, Angles
token_pair_engine_add_pairing(e->pairings2, BRACKET_LEFT, BRACKET_RIGHT, PAIR_BRACKET, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
- token_pair_engine_add_pairing(e->pairings2, BRACKET_CITATION_LEFT, BRACKET_RIGHT, PAIR_BRACKET_CITATION, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
- token_pair_engine_add_pairing(e->pairings2, BRACKET_FOOTNOTE_LEFT, BRACKET_RIGHT, PAIR_BRACKET_FOOTNOTE, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
- token_pair_engine_add_pairing(e->pairings2, BRACKET_IMAGE_LEFT, BRACKET_RIGHT, PAIR_BRACKET_IMAGE, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
+
+ if (extensions & EXT_NOTES) {
+ token_pair_engine_add_pairing(e->pairings2, BRACKET_CITATION_LEFT, BRACKET_RIGHT, PAIR_BRACKET_CITATION, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
+ token_pair_engine_add_pairing(e->pairings2, BRACKET_FOOTNOTE_LEFT, BRACKET_RIGHT, PAIR_BRACKET_FOOTNOTE, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
+ } else {
+ token_pair_engine_add_pairing(e->pairings2, BRACKET_CITATION_LEFT, BRACKET_RIGHT, PAIR_BRACKET, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
+ token_pair_engine_add_pairing(e->pairings2, BRACKET_FOOTNOTE_LEFT, BRACKET_RIGHT, PAIR_BRACKET, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
+ }
+
token_pair_engine_add_pairing(e->pairings2, BRACKET_VARIABLE_LEFT, BRACKET_RIGHT, PAIR_BRACKET_VARIABLE, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
+
+ token_pair_engine_add_pairing(e->pairings2, BRACKET_IMAGE_LEFT, BRACKET_RIGHT, PAIR_BRACKET_IMAGE, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
token_pair_engine_add_pairing(e->pairings2, PAREN_LEFT, PAREN_RIGHT, PAIR_PAREN, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
token_pair_engine_add_pairing(e->pairings2, ANGLE_LEFT, ANGLE_RIGHT, PAIR_ANGLE, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
token_pair_engine_add_pairing(e->pairings2, BRACE_DOUBLE_LEFT, BRACE_DOUBLE_RIGHT, PAIR_BRACES, PAIRING_ALLOW_EMPTY | PAIRING_PRUNE_MATCH);
scan_len = scan_ref_citation(&source[line->start]);
line->type = (scan_len) ? LINE_DEF_CITATION : LINE_PLAIN;
} else {
- line->type = LINE_PLAIN;
+ scan_len = scan_ref_link_no_attributes(&source[line->start]);
+ line->type = (scan_len) ? LINE_DEF_LINK : LINE_PLAIN;
}
break;
case BRACKET_FOOTNOTE_LEFT:
scan_len = scan_ref_foot(&source[line->start]);
line->type = (scan_len) ? LINE_DEF_FOOTNOTE : LINE_PLAIN;
} else {
- line->type = LINE_PLAIN;
+ scan_len = scan_ref_link_no_attributes(&source[line->start]);
+ line->type = (scan_len) ? LINE_DEF_LINK : LINE_PLAIN;
}
break;
case PIPE:
-/* Generated by re2c 0.14.3 on Sat Feb 4 17:39:29 2017 */
+/* Generated by re2c 0.14.3 on Tue Feb 7 20:44:00 2017 */
/**
MultiMarkdown 6 -- Lightweight markup processor to produce HTML, LaTeX, and more.
}
+size_t scan_destination(const char * c) {
+ const char * marker = NULL;
+ const char * start = c;
+
+
+{
+ char yych;
+ yych = *c;
+ switch (yych) {
+ case 0x00:
+ case '\t':
+ case '\r':
+ case ' ': goto yy725;
+ case '\n': goto yy721;
+ case '<': goto yy722;
+ default: goto yy724;
+ }
+yy721:
+ { return 0; }
+yy722:
+ ++c;
+ yych = *c;
+ goto yy729;
+yy723:
+ { return (size_t)( c - start ); }
+yy724:
+ yych = *++c;
+ goto yy727;
+yy725:
+ yych = *++c;
+ goto yy721;
+yy726:
+ ++c;
+ yych = *c;
+yy727:
+ switch (yych) {
+ case 0x00:
+ case '\t':
+ case '\n':
+ case '\r':
+ case ' ': goto yy723;
+ default: goto yy726;
+ }
+yy728:
+ ++c;
+ yych = *c;
+yy729:
+ switch (yych) {
+ case 0x00:
+ case '\t':
+ case '\n':
+ case '\r':
+ case ' ': goto yy723;
+ case '>': goto yy726;
+ default: goto yy728;
+ }
+}
+
+}
+
+
#ifdef TEST
void Test_scan_url(CuTest* tc) {
int url_len;
size_t scan_attr(const char * c);
size_t scan_attributes(const char * c);
size_t scan_definition(const char * c);
+size_t scan_destination(const char * c);
size_t scan_email(const char * c);
size_t scan_fence_start(const char * c);
size_t scan_fence_end(const char * c);
}
+size_t scan_destination(const char * c) {
+ const char * marker = NULL;
+ const char * start = c;
+
+/*!re2c
+ destination { return (size_t)( c - start ); }
+ .? { return 0; }
+*/
+}
+
+
#ifdef TEST
void Test_scan_url(CuTest* tc) {
int url_len;
char * result = NULL;
if (source && pair) {
- result = strndup(&source[pair->start + pair->child->len], pair->len - (pair->child->len + 1));
+ if (pair->child->mate) {
+ // [foo], [^foo], [#foo] should give different strings -- use closer len
+ result = strndup(&source[pair->start + pair->child->mate->len], pair->len - (pair->child->mate->len * 2));
+ } else {
+ result = strndup(&source[pair->start + pair->child->len], pair->len - (pair->child->len + 1));
+ }
}
return result;
}
+char * destination_accept(const char * source, token ** remainder, bool validate) {
+ char * url = NULL;
+ char * clean = NULL;
+ token * t = NULL;
+ size_t start;
+ size_t scan_len;
+
+ switch ((*remainder)->type) {
+ case PAIR_PAREN:
+ case PAIR_ANGLE:
+ case PAIR_QUOTE_SINGLE:
+ case PAIR_QUOTE_DOUBLE:
+ t = token_chain_accept_multiple(remainder, 2, PAIR_ANGLE, PAIR_PAREN);
+ url = text_inside_pair(source, t);
+ break;
+ case TEXT_PLAIN:
+ start = (*remainder)->start;
+
+ // Skip any whitespace
+ while (char_is_whitespace(source[start]))
+ start++;
+
+ scan_len = scan_destination(&source[start]);
+
+ // Grab destination string
+ url = strndup(&source[start], scan_len);
+
+ // Advance remainder
+ while ((*remainder)->start < start + scan_len)
+ *remainder = (*remainder)->next;
+
+
+ t = (*remainder)->prev;
+
+ // Is there a space in a URL concatenated with a title or attribute?
+ // e.g. [foo]: http://foo.bar/ class="foo"
+ // Since only one space between URL and class, they are joined.
+
+ if (t->type == TEXT_PLAIN) {
+ // Trim leading whitespace
+ token_trim_leading_whitespace(t, source);
+ token_split_on_char(t, source, ' ');
+ *remainder = t->next;
+ }
+
+ break;
+ }
+
+ // Is this a valid URL?
+ clean = clean_string(url, false);
+
+ if (validate && !validate_url(clean)) {
+ free(clean);
+ clean = NULL;
+ }
+
+ free(url);
+ return clean;
+}
+
+
char * url_accept(const char * source, token ** remainder, bool validate) {
char * url = NULL;
char * clean = NULL;
// Prepare for parsing
+ // Account for settings
+
switch (label->type) {
+ case PAIR_BRACKET_CITATION:
+ if (e->extensions & EXT_NOTES) {
+ if (!token_chain_accept(remainder, COLON))
+ return false;
+
+ title = *remainder; // Track first token of content in 'title'
+ f = footnote_new(e->dstr->str, label, title);
+
+ // Store citation for later use
+ stack_push(e->citation_stack, f);
+
+ break;
+ }
+ case PAIR_BRACKET_FOOTNOTE:
+ if (e->extensions & EXT_NOTES) {
+ if (!token_chain_accept(remainder, COLON))
+ return false;
+
+ title = *remainder; // Track first token of content in 'title'
+ f = footnote_new(e->dstr->str, label, title);
+
+ // Store footnote for later use
+ stack_push(e->footnote_stack, f);
+
+ break;
+ }
case PAIR_BRACKET:
// Reference Link Definition
// Skip space
whitespace_accept(remainder);
- // Grab URL
- url_char = url_accept(e->dstr->str, remainder, false);
+ // Grab destination
+ url_char = destination_accept(e->dstr->str, remainder, false);
whitespace_accept(remainder);
if (l)
stack_push(e->link_stack, l);
- break;
- case PAIR_BRACKET_CITATION:
- if (!token_chain_accept(remainder, COLON))
- return false;
-
- title = *remainder; // Track first token of content in 'title'
- f = footnote_new(e->dstr->str, label, title);
-
- // Store citation for later use
- stack_push(e->citation_stack, f);
-
- break;
- case PAIR_BRACKET_FOOTNOTE:
- if (!token_chain_accept(remainder, COLON))
- return false;
-
- title = *remainder; // Track first token of content in 'title'
- f = footnote_new(e->dstr->str, label, title);
-
- // Store footnote for later use
- stack_push(e->footnote_stack, f);
-
break;
case PAIR_BRACKET_VARIABLE:
fprintf(stderr, "Process variable:\n");
void process_definition_block(mmd_engine * e, token * block) {
footnote * f;
+
token * label = block->child;
if (label->type == BLOCK_PARA)
label = label->child;
chain = chain->next;
}
-}
\ No newline at end of file
+}
</div>
-5. I haven't worked a lot yet on the MMD-specific features, so there may be
-more changes to come. One thing I do anticipate is that if fenced code blocks
-stay, they will work slightly differently. Currently, an opening fence
-doesn't mean anything unless there is a closing fence that follows it. Again,
-this requires backtracking in the parser. I suspect that an opening fence
-will definitely open a code block. If there is no closing fence, then the
-rest of the document will remain inside the code block. This is the approach
-used by CommonMark and it's a reasonable one, IMO.
+5. "Malformed" reference link definitions are handled slightly differently.
+For example, `Reference Footnotes.text` is parsed differently in compatibility
+mode than MMD-5. This started as a side-effect of the parsing algorithm, but
+I actually think it makes sense. This may or may not change in the future.
## Where Does MultiMarkdown 6 Stand? ##
* Automatic cross-reference targets
* Basic Citation support
* CriticMarkup support
+* Definition lists
+* Footnotes
* Inline and reference footnotes
* Image and Link attributes (attributes can now be used with inline links as
well as reference links)
* Locators required?
* CriticMarkup -- need to decide:
* How to handle CM stretches that include blank lines
-* Definition lists
* Fenced code blocks
-* Footnotes -- need support for multiple blocks inside
* Headers -- need support for manual labels
* Metadata
* Full/Snippet modes
-<p>
Reference.<a href="#fn:1" id="fnref:1" title="see footnote" class="footnote">[1]</a></p>
+<p>
foo.<a href="#fn:1" id="fnref:1" title="see footnote" class="footnote">[1]</a></p>
-<p>Reference.<a href="#fn:2" id="fnref:2" title="see footnote" class="footnote">[2]</a></p>
+<p>foo.<a href="#fn:2" id="fnref:2" title="see footnote" class="footnote">[2]</a></p>
+
+<p>foo.<a href="#fn:3" id="fnref:3" title="see footnote" class="footnote">[3]</a></p>
<div class="footnotes">
<hr />
<ol>
<li id="fn:1">
-<p>
This is a <em>short</em> footnote. <a href="#fnref:1" title="return to body" class="reversefootnote"> ↩</a></p>
+<p>
<em>foo</em> <a href="#fnref:1" title="return to body" class="reversefootnote"> ↩</a></p>
</li>
<li id="fn:2">
-<p>This is a longer footnote.
-With two lines. <a href="#fnref:2" title="return to body" class="reversefootnote"> ↩</a></p>
+<p><em>foo
+bar</em> <a href="#fnref:2" title="return to body" class="reversefootnote"> ↩</a></p>
+</li>
+
+<li id="fn:3">
+<p>foo</p>
+
+<p><em>bar</em> <a href="#fnref:3" title="return to body" class="reversefootnote"> ↩</a></p>
+
+<ul>
+<li>bat</li>
+</ul>
</li>
</ol>
</div>
-
-<p>Reference.[^foo]</p>
+<p>foo.<a href="*foo*">^bar</a></p>
-<p>Reference.[^foo2]</p>
+<p>foo.<a href="*foo">^bar2</a></p>
-<p>[^foo]: This is a <em>short</em> footnote.
-[^foo2]: This is a longer footnote.
-With two lines.</p>
+<p>foo.<a href="foo">^bar3</a></p>
+
+<pre><code>*bar*
+
+* bat
+</code></pre>
-Reference.[^foo]
+foo.[^bar]
-Reference.[^foo2]
+foo.[^bar2]
+foo.[^bar3]
-[^foo]: This is a *short* footnote.
-[^foo2]: This is a longer footnote.
-With two lines.
+[^bar]: *foo*
+[^bar2]: *foo
+bar*
+[^bar3]: foo
+
+ *bar*
+
+ * bat
projects / multimarkdown / commitdiff