Add implementation of reference unfolder using banzai

3 files changed, 78 insertions, 8 deletions

diff --git a/lib/banzai/filter/reference_filter.rb b/lib/banzai/filter/reference_filter.rb
index 3637b1bac94..132f0a4bd93 100644
--- a/lib/banzai/filter/reference_filter.rb
+++ b/lib/banzai/filter/reference_filter.rb
@@ -47,6 +47,7 @@ module Banzai
       # Returns a String
       def data_attribute(attributes = {})
         attributes[:reference_filter] = self.class.name.demodulize
+        attributes.delete(:original) if context[:no_original_data]
         attributes.map { |key, value| %Q(data-#{key.to_s.dasherize}="#{escape_once(value)}") }.join(" ")
       end
 
diff --git a/lib/gitlab/gfm/reference_unfolder.rb b/lib/gitlab/gfm/reference_unfolder.rb
index 4969e8ebe85..57871c36eb4 100644
--- a/lib/gitlab/gfm/reference_unfolder.rb
+++ b/lib/gitlab/gfm/reference_unfolder.rb
@@ -8,23 +8,66 @@ module Gitlab
       def initialize(text, project)
         @text = text
         @project = project
+        @original = markdown(text)
       end
 
       def unfold(from_project)
-        referables.each_with_object(@text.dup) do |referable, text|
-          next unless referable.respond_to?(:to_reference)
+        return @text unless @text =~ references_pattern
 
-          pattern = /#{Regexp.escape(referable.to_reference)}/
-          text.gsub!(pattern, referable.to_reference(from_project))
+        unfolded = @text.gsub(references_pattern) do |reference|
+          unfold_reference(reference, Regexp.last_match, from_project)
         end
+
+        unless substitution_valid?(unfolded)
+          raise StandardError, 'Invalid references unfolding!'
+        end
+
+        unfolded
       end
 
       private
 
+      def unfold_reference(reference, match, from_project)
+        before = @text[0...match.begin(0)]
+        after = @text[match.end(0)...@text.length]
+        referable = find_referable(reference)
+
+        return reference unless referable
+        cross_reference = referable.to_reference(from_project)
+        new_text = before + cross_reference + after
+
+        substitution_valid?(new_text) ? cross_reference : reference
+      end
+
+      def references_pattern
+        return @pattern if @pattern
+
+        patterns = Gitlab::ReferenceExtractor::REFERABLES.map do |ref|
+          ref.to_s.classify.constantize.try(:reference_pattern)
+        end
+
+        @pattern = Regexp.union(patterns.compact)
+      end
+
       def referables
+        return @referables if @referables
+
         extractor = Gitlab::ReferenceExtractor.new(@project)
         extractor.analyze(@text)
-        extractor.all
+        @referables = extractor.all
+      end
+
+      def find_referable(reference)
+        referables.find { |ref| ref.to_reference == reference }
+      end
+
+      def substitution_valid?(substituted)
+        @original == markdown(substituted)
+      end
+
+      def markdown(text)
+        helper = Class.new.extend(GitlabMarkdownHelper)
+        helper.markdown(text, project: @project, no_original_data: true)
       end
     end
   end
diff --git a/spec/lib/gitlab/gfm/reference_unfolder_spec.rb b/spec/lib/gitlab/gfm/reference_unfolder_spec.rb
index 4f8b960350c..40cdb7e1452 100644
--- a/spec/lib/gitlab/gfm/reference_unfolder_spec.rb
+++ b/spec/lib/gitlab/gfm/reference_unfolder_spec.rb
@@ -33,10 +33,36 @@ describe Gitlab::Gfm::ReferenceUnfolder do
       end
 
       context 'description ambigous elements' do
-        let(:url) { 'http://gitlab.com/#1' }
-        let(:text) { "This references #1, but not #{url}" }
+        context 'url' do
+          let(:url) { 'http://gitlab.com/#1' }
+          let(:text) { "This references #1, but not #{url}" }
 
-        it { is_expected.to include url }
+          it { is_expected.to include url }
+        end
+
+        context 'code' do
+          let(:text) { "#1, but not `[#1]`" }
+          it { is_expected.to eq "#{issue_first.to_reference(new_project)}, but not `[#1]`" }
+        end
+
+        context 'code reverse' do
+          let(:text) { "not `#1`, but #1" }
+          it { is_expected.to eq "not `#1`, but #{issue_first.to_reference(new_project)}" }
+        end
+
+        context 'code in random order' do
+          let(:text) { "#1, `#1`, #1, `#1`" }
+          let(:ref) { issue_first.to_reference(new_project) }
+
+          it { is_expected.to eq "#{ref}, `#1`, #{ref}, `#1`" }
+        end
+      end
+
+      context 'reference contains milestone' do
+        let(:milestone) { create(:milestone) }
+        let(:text) { "milestone ref: #{milestone.to_reference}" }
+
+        it { is_expected.to eq text }
       end
     end
   end