rndr_header_anchor: use djb2 hash for non-ascii text

vmg · Dec 21, 2016 · f2d0ad9 · f2d0ad9
1 parent 8d8e1ec
commit f2d0ad9
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 1 deletion.
diff --git a/ext/redcarpet/html.c b/ext/redcarpet/html.c
@@ -307,6 +307,15 @@ rndr_header_anchor(struct buf *out, const struct buf *anchor)
 	// replace the last dash if there was anything added
 	if (stripped && inserted)
 		out->size--;
+
+	// if anchor found empty, use djb2 hash for it
+	if (!inserted && anchor->size) {
+	        unsigned long hash = 5381;
+		for (i = 0; i < size; ++i) {
+			hash = ((hash << 5) + hash) + a[i]; /* h * 33 + c */
+		}
+		bufprintf(out, "part-%lx", hash);
+	}
 }
 
 static void

diff --git a/test/html_render_test.rb b/test/html_render_test.rb
@@ -254,7 +254,7 @@ def test_non_ascii_removal_in_header_anchors
 
   def test_utf8_only_header_anchors
     markdown = "# 見出し"
-    html = "<h1 id=\"\">見出し</h1>"
+    html = "<h1 id=\"part-37870bfa194139f\">見出し</h1>"
 
     assert_equal html, render(markdown, with: [:with_toc_data])
   end