wip: fake its a binary diff

4 files changed, 30 insertions, 7 deletions

diff --git a/lib/api/commits.rb b/lib/api/commits.rb
index ea78737288a..4b8d248f5f7 100644
--- a/lib/api/commits.rb
+++ b/lib/api/commits.rb
@@ -104,7 +104,7 @@ module API
 
         not_found! 'Commit' unless commit
 
-        commit.raw_diffs.to_a
+        present commit.raw_diffs.to_a, with: Entities::RepoDiff
       end
 
       desc "Get a commit's comments" do
diff --git a/lib/api/entities.rb b/lib/api/entities.rb
index 9114b69606b..0c63dc345c2 100644
--- a/lib/api/entities.rb
+++ b/lib/api/entities.rb
@@ -291,10 +291,11 @@ module API
     end
 
     class RepoDiff < Grape::Entity
-      expose :old_path, :new_path, :a_mode, :b_mode, :diff
+      expose :old_path, :new_path, :a_mode, :b_mode
       expose :new_file?, as: :new_file
       expose :renamed_file?, as: :renamed_file
       expose :deleted_file?, as: :deleted_file
+      expose :diff
     end
 
     class ProtectedRefAccess < Grape::Entity
diff --git a/lib/gitlab/encoding_helper.rb b/lib/gitlab/encoding_helper.rb
index 8ddc91e341d..c5e173ba55a 100644
--- a/lib/gitlab/encoding_helper.rb
+++ b/lib/gitlab/encoding_helper.rb
@@ -13,6 +13,8 @@ module Gitlab
     # https://gitlab.com/gitlab-org/gitlab_git/merge_requests/77#note_4754193
     ENCODING_CONFIDENCE_THRESHOLD = 50
 
+    #
+    # 
     def encode!(message)
       return nil unless message.respond_to? :force_encoding
 
@@ -22,20 +24,26 @@ module Gitlab
 
       # return message if message type is binary
       detect = CharlockHolmes::EncodingDetector.detect(message)
-      return message.force_encoding("BINARY") if detect && detect[:type] == :binary
+      return message.force_encoding("BINARY") if binary?(message, detect)
 
-      # force detected encoding if we have sufficient confidence.
       if detect && detect[:encoding] && detect[:confidence] > ENCODING_CONFIDENCE_THRESHOLD
+        # force detected encoding if we have sufficient confidence.
         message.force_encoding(detect[:encoding])
       end
 
       # encode and clean the bad chars
       message.replace clean(message)
-    rescue
+    rescue => e
+      byebug
       encoding = detect ? detect[:encoding] : "unknown"
       "--broken encoding: #{encoding}"
     end
 
+    def binary?(message, detect=nil)
+      detect ||= CharlockHolmes::EncodingDetector.detect(message)
+      detect && detect[:type] == :binary && detect[:confidence] == 100
+    end
+
     def encode_utf8(message)
       detect = CharlockHolmes::EncodingDetector.detect(message)
       if detect && detect[:encoding]
@@ -50,7 +58,7 @@ module Gitlab
         clean(message)
       end
     end
-
+      
     private
 
     def clean(message)
diff --git a/lib/gitlab/git/diff.rb b/lib/gitlab/git/diff.rb
index ce3d65062e8..e544f255a5d 100644
--- a/lib/gitlab/git/diff.rb
+++ b/lib/gitlab/git/diff.rb
@@ -116,6 +116,13 @@ module Gitlab
 
           filtered_opts
         end
+
+        # Return a binary diff message like:
+        # 
+        # "Binary files a/file/path and b/file/path differ\n"
+        def binary_message(old_path, new_path)
+          "Binary files #{old_path} and #{new_path} differ\n"
+        end
       end
 
       def initialize(raw_diff, expanded: true)
@@ -214,7 +221,14 @@ module Gitlab
         # binary we're not going to display anything so we skip the size check.
         return if !patch.delta.binary? && prune_large_patch(patch)
 
-        @diff = encode!(strip_diff_headers(patch.to_s))
+        diff = strip_diff_headers(patch.to_s)
+        @diff = if binary?(diff)
+                  # the diff is binary, let's make a message for it
+                  Diff::binary_message(patch.delta.old_file[:path],
+                                       patch.delta.new_file[:path])
+                else
+                  encode!(diff)
+                end
       end
 
       def init_from_hash(hash)