Merge pull request #31 from gjtorikian/fixups

Update cmark to 0.27.1

Author: gjtorikian

ext/commonmarker/cmark

@@ -28,6 +28,13 @@ static VALUE sym_image;
 static VALUE sym_bullet_list;
 static VALUE sym_ordered_list;
 
+static VALUE encode_utf8_string(char *c_string) {
+  VALUE string = rb_str_new2(c_string);
+  int enc = rb_enc_find_index("UTF-8");
+  rb_enc_associate_index(string, enc);
+  return string;
+}
+
 static void rb_mark_c_struct(void *data) {
   cmark_node *node = data;
   cmark_node *child;
@@ -231,7 +238,7 @@ static VALUE rb_node_get_string_content(VALUE self) {
     rb_raise(rb_mNodeError, "could not get string content");
   }
 
-  return rb_str_new2(text);
+  return encode_utf8_string(text);
 }
 
 /*

ext/commonmarker/commonmarker.c

@@ -7,6 +7,7 @@
 
 #include "cmark.h"
 #include "ruby.h"
+#include "ruby/encoding.h"
 
 #define CSTR2SYM(s) (ID2SYM(rb_intern((s))))
 

ext/commonmarker/commonmarker.h

@@ -9,4 +9,11 @@ def test_encoding
     render = doc.to_html
     assert_equal render.rstrip, '<p>This curly quote “makes commonmarker throw an exception”.</p>'
   end
+
+  def test_string_content_is_utf8
+    doc = CommonMarker.render_doc('Hi *there*')
+    text = doc.first_child.last_child.first_child
+    assert_equal text.string_content, 'there'
+    assert_equal text.string_content.encoding.name, 'UTF-8'
+  end
 end