Support a single directory traversal in RelativeLinkFilter

Prior, if we were viewing a blob at
`https://example.com/namespace/project/blob/master/doc/some-file.md` and
it contained a relative link such as `[README](../README.md)`, the
resulting link when viewing the blob would be:
`https://example.com/namespace/project/blob/README.md` which omits the
`master` ref, resulting in a 404.

lib/banzai/filter/relative_link_filter.rb

@@ -91,7 +91,7 @@ module Banzai
         parts = request_path.split('/')
         parts.pop if path_type(request_path) != 'tree'
-        while parts.length > 1 && path.start_with?('../')
+        while path.start_with?('../')
           parts.pop
           path.sub!('../', '')
         end

spec/lib/banzai/filter/relative_link_filter_spec.rb

@@ -92,6 +92,14 @@ describe Banzai::Filter::RelativeLinkFilter, lib: true do
         to eq "/#{project_path}/blob/#{ref}/doc/api/README.md"
     end
+    it 'rebuilds relative URL for a file in the repository root' do
+      relative_link = link('../README.md')
+      doc = filter(relative_link, requested_path: 'doc/some-file.md')
+
+      expect(doc.at_css('a')['href']).
+        to eq "/#{project_path}/blob/#{ref}/README.md"
+    end
+
     it 'rebuilds relative URL for a file in the repo with an anchor' do
       doc = filter(link('README.md#section'))
       expect(doc.at_css('a')['href']).