Backport Upskirt changes

- Improve quoting/striking through inner spans - Make `put_scaped_char` more sane - Make autolinking case-insensitive
nono · Apr 22, 2011 · e328f33 · e328f33
1 parent 612b142
commit e328f33
Show file tree

Hide file tree

Showing 6 changed files with 69 additions and 104 deletions.
diff --git a/ext/markdown.c b/ext/markdown.c
@@ -137,7 +137,7 @@ is_safe_link(const char *link, size_t link_len)
 	for (i = 0; i < valid_uris_count; ++i) {
 		size_t len = strlen(valid_uris[i]);
 
-		if (link_len > len && memcmp(link, valid_uris[i], len) == 0)
+		if (link_len > len && strncasecmp(link, valid_uris[i], len) == 0)
 			return 1;
 	}
 
@@ -397,7 +397,7 @@ parse_emph1(struct buf *ob, struct render *rndr, char *data, size_t size, char c
 
 			work = rndr_newbuf(rndr);
 			parse_inline(work, rndr, data, i);
-			r = rndr->make.emphasis(ob, work, c, rndr->make.opaque);
+			r = rndr->make.emphasis(ob, work, rndr->make.opaque);
 			rndr_popbuf(rndr);
 			return r ? i + 1 : 0;
 		}
@@ -410,11 +410,14 @@ parse_emph1(struct buf *ob, struct render *rndr, char *data, size_t size, char c
 static size_t
 parse_emph2(struct buf *ob, struct render *rndr, char *data, size_t size, char c)
 {
+	int (*render_method)(struct buf *ob, struct buf *text, void *opaque);
 	size_t i = 0, len;
 	struct buf *work = 0;
 	int r;
 
-	if (!rndr->make.double_emphasis)
+	render_method = (c == '~') ? rndr->make.strikethrough : rndr->make.double_emphasis;
+
+	if (!render_method)
 		return 0;
 
 	while (i < size) {
@@ -425,7 +428,7 @@ parse_emph2(struct buf *ob, struct render *rndr, char *data, size_t size, char c
 		if (i + 1 < size && data[i] == c && data[i + 1] == c && i && !isspace(data[i - 1])) {
 			work = rndr_newbuf(rndr);
 			parse_inline(work, rndr, data, i);
-			r = rndr->make.double_emphasis(ob, work, c, rndr->make.opaque);
+			r = render_method(ob, work, rndr->make.opaque);
 			rndr_popbuf(rndr);
 			return r ? i + 2 : 0;
 		}
@@ -456,7 +459,7 @@ parse_emph3(struct buf *ob, struct render *rndr, char *data, size_t size, char c
 			struct buf *work = rndr_newbuf(rndr);
 
 			parse_inline(work, rndr, data, i);
-			r = rndr->make.triple_emphasis(ob, work, c, rndr->make.opaque);
+			r = rndr->make.triple_emphasis(ob, work, rndr->make.opaque);
 			rndr_popbuf(rndr);
 			return r ? i + 3 : 0;
 
@@ -484,8 +487,9 @@ char_emphasis(struct buf *ob, struct render *rndr, char *data, size_t offset, si
 	size_t ret;
 
 	if (size > 2 && data[1] != c) {
-		/* whitespace cannot follow an opening emphasis */
-		if (isspace(data[1]) || (ret = parse_emph1(ob, rndr, data + 1, size - 1, c)) == 0)
+		/* whitespace cannot follow an opening emphasis;
+		 * strikethrough only takes two characters '~~' */
+		if (c == '~' || isspace(data[1]) || (ret = parse_emph1(ob, rndr, data + 1, size - 1, c)) == 0)
 			return 0;
 
 		return ret + 1;
@@ -499,7 +503,7 @@ char_emphasis(struct buf *ob, struct render *rndr, char *data, size_t offset, si
 	}
 
 	if (size > 4 && data[1] == c && data[2] == c && data[3] != c) {
-		if (isspace(data[3]) || (ret = parse_emph3(ob, rndr, data + 3, size - 3, c)) == 0)
+		if (c == '~' || isspace(data[3]) || (ret = parse_emph3(ob, rndr, data + 3, size - 3, c)) == 0)
 			return 0;
 
 		return ret + 3;
@@ -1970,8 +1974,13 @@ ups_markdown(struct buf *ob, struct buf *ib, const struct mkd_renderer *rndrer,
 
 	if (extensions & MKDEXT_AUTOLINK) {
 		rndr.active_char['h'] = char_autolink; // http, https
+		rndr.active_char['H'] = char_autolink;
+
 		rndr.active_char['f'] = char_autolink; // ftp
+		rndr.active_char['F'] = char_autolink;
+
 		rndr.active_char['m'] = char_autolink; // mailto
+		rndr.active_char['M'] = char_autolink;
 	}
 
 	/* Extension data */

diff --git a/ext/markdown.h b/ext/markdown.h
@@ -61,13 +61,14 @@ struct mkd_renderer {
 	/* span level callbacks - NULL or return 0 prints the span verbatim */
 	int (*autolink)(struct buf *ob, struct buf *link, enum mkd_autolink type, void *opaque);
 	int (*codespan)(struct buf *ob, struct buf *text, void *opaque);
-	int (*double_emphasis)(struct buf *ob, struct buf *text, char c, void *opaque);
-	int (*emphasis)(struct buf *ob, struct buf *text, char c,void *opaque);
+	int (*double_emphasis)(struct buf *ob, struct buf *text, void *opaque);
+	int (*emphasis)(struct buf *ob, struct buf *text, void *opaque);
 	int (*image)(struct buf *ob, struct buf *link, struct buf *title, struct buf *alt, void *opaque);
 	int (*linebreak)(struct buf *ob, void *opaque);
 	int (*link)(struct buf *ob, struct buf *link, struct buf *title, struct buf *content, void *opaque);
 	int (*raw_html_tag)(struct buf *ob, struct buf *tag, void *opaque);
-	int (*triple_emphasis)(struct buf *ob, struct buf *text, char c, void *opaque);
+	int (*triple_emphasis)(struct buf *ob, struct buf *text, void *opaque);
+	int (*strikethrough)(struct buf *ob, struct buf *text, void *opaque);
 
 	/* low level callbacks - NULL copies input directly into the output */
 	void (*entity)(struct buf *ob, struct buf *entity, void *opaque);

diff --git a/ext/xhtml.c b/ext/xhtml.c
@@ -29,18 +29,23 @@ struct xhtml_renderopt {
 		int current_level;
 	} toc_data;
 
+	struct {
+		int in_squote;
+		int in_dquote;
+	} quotes;
+
 	unsigned int flags;
 };
 
-static inline int
+static inline void
 put_scaped_char(struct buf *ob, char c)
 {
 	switch (c) {
-		case '<': BUFPUTSL(ob, "&lt;"); return 1;
-		case '>': BUFPUTSL(ob, "&gt;"); return 1;
-		case '&': BUFPUTSL(ob, "&amp;"); return 1;
-		case '"': BUFPUTSL(ob, "&quot;"); return 1;
-		default: return 0;
+		case '<': BUFPUTSL(ob, "&lt;"); break;
+		case '>': BUFPUTSL(ob, "&gt;"); break;
+		case '&': BUFPUTSL(ob, "&amp;"); break;
+		case '"': BUFPUTSL(ob, "&quot;"); break;
+		default: bufputc(ob, c); break;
 	}
 }
 
@@ -190,28 +195,34 @@ rndr_codespan(struct buf *ob, struct buf *text, void *opaque)
 }
 
 static int
-rndr_double_emphasis(struct buf *ob, struct buf *text, char c, void *opaque)
+rndr_strikethrough(struct buf *ob, struct buf *text, void *opaque)
 {
 	if (!text || !text->size)
 		return 0;
 
-	if (c == '~') {
-		BUFPUTSL(ob, "<del>");
-		bufput(ob, text->data, text->size);
-		BUFPUTSL(ob, "</del>");
-	} else {
-		BUFPUTSL(ob, "<strong>");
-		bufput(ob, text->data, text->size);
-		BUFPUTSL(ob, "</strong>");
-	}
+	BUFPUTSL(ob, "<del>");
+	bufput(ob, text->data, text->size);
+	BUFPUTSL(ob, "</del>");
+	return 1;
+}
+
+static int
+rndr_double_emphasis(struct buf *ob, struct buf *text, void *opaque)
+{
+	if (!text || !text->size)
+		return 0;
+
+	BUFPUTSL(ob, "<strong>");
+	bufput(ob, text->data, text->size);
+	BUFPUTSL(ob, "</strong>");
 
 	return 1;
 }
 
 static int
-rndr_emphasis(struct buf *ob, struct buf *text, char c, void *opaque)
+rndr_emphasis(struct buf *ob, struct buf *text, void *opaque)
 {
-	if (!text || !text->size || c == '~') return 0;
+	if (!text || !text->size) return 0;
 	BUFPUTSL(ob, "<em>");
 	if (text) bufput(ob, text->data, text->size);
 	BUFPUTSL(ob, "</em>");
@@ -311,6 +322,10 @@ rndr_paragraph(struct buf *ob, struct buf *text, void *opaque)
 		bufput(ob, &text->data[i], text->size - i);
 	}
 	BUFPUTSL(ob, "</p>\n");
+
+	/* Close any open quotes at the end of the paragraph */
+	options->quotes.in_squote = 0;
+	options->quotes.in_dquote = 0;
 }
 
 static void
@@ -329,7 +344,7 @@ rndr_raw_block(struct buf *ob, struct buf *text, void *opaque)
 }
 
 static int
-rndr_triple_emphasis(struct buf *ob, struct buf *text, char c, void *opaque)
+rndr_triple_emphasis(struct buf *ob, struct buf *text, void *opaque)
 {
 	if (!text || !text->size) return 0;
 	BUFPUTSL(ob, "<strong><em>");
@@ -546,8 +561,8 @@ rndr_normal_text(struct buf *ob, struct buf *text, void *opaque)
 static void
 rndr_smartypants(struct buf *ob, struct buf *text, void *opaque)
 {
+	struct xhtml_renderopt *options = opaque;
 	size_t i;
-	int open_single = 0, open_double = 0, open_tag = 0;
 
 	if (!text)
 		return;
@@ -572,87 +587,25 @@ rndr_smartypants(struct buf *ob, struct buf *text, void *opaque)
 			continue;
 
 		switch (c) {
-		case '<':
-			open_tag = 1;
-			break;
-
-		case '>':
-			open_tag = 0;
-			break;
-
-#if 0
-		/*
-		 * FIXME: this is bongos.
-		 *
-		 * The markdown spec defines that code blocks can be delimited
-		 * by more than one backtick, e.g.
-		 *
-		 *		``There is a literal backtick (`) here.``
-		 *		<p><code>There is a literal backtick (`) here.</code></p>
-		 *
-		 * Obviously, there's no way to differentiate between the start
-		 * of a code block and the start of a quoted string for smartypants
-		 *
-		 * Look at this piece of Python code:
-		 *
-		 *		``result = ''.join(['this', 'is', 'bongos'])``
-		 *
-		 * This MD expression is clearly ambiguous since it can be parsed as:
-		 *
-		 *		<p>&ldquo;result = &rdquo;.join ...</p>
-		 *
-		 * Or also as:
-		 *
-		 *		<p><code>result = ''.join(['this', 'is', 'bongos'])</code></p>
-		 *
-		 * Fuck everything about this. This is temporarily disabled, because at GitHub
-		 * it's probably smarter to prioritize code blocks than pretty cutesy punctuation.
-		 *
-		 * The equivalent closing tag for the (``), ('') has also been disabled, because
-		 * it makes no sense to have closing tags without opening tags.
-		 */
-		case '`':
-			if (open_tag == 0) {
-				if (i + 1 < text->size && text->data[i + 1] == '`') {
-					BUFPUTSL(ob, "&ldquo;"); i++;
-					continue;
-				}
-			}
-			break;
-#endif
-
 		case '\"':
-			if (open_tag == 0) {
-				if (smartypants_quotes(ob, text, i, open_double)) {
-					open_double = !open_double;
-					continue;
-				}
+			if (smartypants_quotes(ob, text, i, options->quotes.in_dquote)) {
+				options->quotes.in_dquote = !options->quotes.in_dquote;
+				continue;
 			}
 			break;
 
 		case '\'':
-			if (open_tag == 0) {
-
-#if 0 /* temporarily disabled, see previous comment */
-				if (i + 1 < text->size && text->data[i + 1] == '\'') {
-					BUFPUTSL(ob, "&rdquo;"); i++;
-					continue;
-				} 
-#endif
-
-				if (smartypants_quotes(ob, text, i, open_single)) {
-					open_single = !open_single;
-					continue;
-				}
+			if (smartypants_quotes(ob, text, i, options->quotes.in_squote)) {
+				options->quotes.in_squote = !options->quotes.in_squote;
+				continue;
 			}
 			break;
 		}
 
 		/*
 		 * Copy raw character
 		 */
-		if (!put_scaped_char(ob, c))
-			bufputc(ob, c);
+		put_scaped_char(ob, c);
 	}
 }
 
@@ -720,6 +673,7 @@ ups_toc_renderer(struct mkd_renderer *renderer)
 		NULL,
 		NULL,
 		rndr_triple_emphasis,
+		rndr_strikethrough,
 
 		NULL,
 		NULL,
@@ -763,6 +717,7 @@ ups_xhtml_renderer(struct mkd_renderer *renderer, unsigned int render_flags)
 		rndr_link,
 		rndr_raw_html,
 		rndr_triple_emphasis,
+		rndr_strikethrough,
 
 		NULL,
 		rndr_normal_text,

diff --git a/lib/redcarpet.rb b/lib/redcarpet.rb
@@ -26,7 +26,7 @@
 #   end
 #
 class Redcarpet
-  VERSION = '1.10.0'
+  VERSION = '1.10.1'
 
   # Original Markdown formatted text.
   attr_reader :text

diff --git a/redcarpet.gemspec b/redcarpet.gemspec
@@ -1,9 +1,9 @@
 Gem::Specification.new do |s|
   s.name = 'redcarpet'
-  s.version = '1.10.0'
+  s.version = '1.10.1'
   s.summary = "Ruby bindings for libupskirt"
   s.description = 'A fast and safe Markdown to (X)HTML parser'
-  s.date = '2011-04-21'
+  s.date = '2011-04-22'
   s.email = 'vicent@github.com'
   s.homepage = 'http://github.com/tanoku/redcarpet'
   s.has_rdoc = true

diff --git a/upskirt b/upskirt