Expose language field in GraphQL blob type

Contributes to https://gitlab.com/gitlab-org/gitlab/-/issues/348145

Return the language of the blob for syntax highligher.

Changelog: added

app/graphql/types/repository/blob_type.rb

@@ -125,6 +125,12 @@ module Types
       field :archived, GraphQL::Types::Boolean, null: true, method: :archived?,
             description: 'Whether the current project is archived.'
 
+      field :language, GraphQL::Types::String,
+            description: 'Blob language.',
+            method: :blob_language,
+            null: true,
+            calls_gitaly: true
+
       def raw_text_blob
         object.data unless object.binary?
       end

app/presenters/blob_presenter.rb

@@ -32,7 +32,7 @@ class BlobPresenter < Gitlab::View::Presenter::Delegated
   end
 
   def blob_language
-    @_blob_language ||= Gitlab::Diff::CustomDiff.transformed_blob_language(blob) || language
+    @_blob_language ||= Gitlab::Diff::CustomDiff.transformed_blob_language(blob) || gitattr_language || detect_language
   end
 
   def raw_plain_data
@@ -166,9 +166,15 @@ class BlobPresenter < Gitlab::View::Presenter::Delegated
     @all_lines ||= blob.data.lines
   end
 
-  def language
+  def gitattr_language
     blob.language_from_gitattributes
   end
+
+  def detect_language
+    return if blob.binary?
+
+    Rouge::Lexer.guess(filename: blob.path, source: blob_data(nil)) { |lex| lex.min_by(&:tag) }.tag
+  end
 end
 
 BlobPresenter.prepend_mod_with('BlobPresenter')

app/presenters/snippet_blob_presenter.rb

@@ -33,7 +33,7 @@ class SnippetBlobPresenter < BlobPresenter
     blob.container
   end
 
-  def language
+  def gitattr_language
     nil
   end
 

doc/api/graphql/reference/index.md

@@ -14632,6 +14632,7 @@ Returns [`Tree`](#tree).
 | <a id="repositoryblobid"></a>`id` | [`ID!`](#id) | ID of the blob. |
 | <a id="repositoryblobideeditpath"></a>`ideEditPath` | [`String`](#string) | Web path to edit this blob in the Web IDE. |
 | <a id="repositoryblobideforkandeditpath"></a>`ideForkAndEditPath` | [`String`](#string) | Web path to edit this blob in the Web IDE using a forked project. |
+| <a id="repositorybloblanguage"></a>`language` | [`String`](#string) | Blob language. |
 | <a id="repositorybloblfsoid"></a>`lfsOid` | [`String`](#string) | LFS OID of the blob. |
 | <a id="repositoryblobmode"></a>`mode` | [`String`](#string) | Blob mode. |
 | <a id="repositoryblobname"></a>`name` | [`String`](#string) | Blob name. |

lib/gitlab/search/found_blob.rb

@@ -80,6 +80,10 @@ module Gitlab
         super(presenter_class: BlobPresenter)
       end
 
+      def binary?
+        false
+      end
+
       def fetch_blob
         path = [ref, blob_path]
         missing_blob = { binary_path: blob_path }

spec/graphql/types/repository/blob_type_spec.rb

@@ -41,7 +41,8 @@ RSpec.describe Types::Repository::BlobType do
       :ide_edit_path,
       :external_storage_url,
       :fork_and_edit_path,
-      :ide_fork_and_edit_path
+      :ide_fork_and_edit_path,
+      :language
     )
   end
 end

spec/presenters/blob_presenter_spec.rb

@@ -170,13 +170,13 @@ RSpec.describe BlobPresenter do
     let(:git_blob) { blob.__getobj__ }
 
     it 'returns highlighted content' do
-      expect(Gitlab::Highlight).to receive(:highlight).with('files/ruby/regex.rb', git_blob.data, plain: nil, language: nil)
+      expect(Gitlab::Highlight).to receive(:highlight).with('files/ruby/regex.rb', git_blob.data, plain: nil, language: 'ruby')
 
       presenter.highlight
     end
 
     it 'returns plain content when :plain is true' do
-      expect(Gitlab::Highlight).to receive(:highlight).with('files/ruby/regex.rb', git_blob.data, plain: true, language: nil)
+      expect(Gitlab::Highlight).to receive(:highlight).with('files/ruby/regex.rb', git_blob.data, plain: true, language: 'ruby')
 
       presenter.highlight(plain: true)
     end
@@ -189,7 +189,7 @@ RSpec.describe BlobPresenter do
       end
 
       it 'returns limited highlighted content' do
-        expect(Gitlab::Highlight).to receive(:highlight).with('files/ruby/regex.rb', "line one\n", plain: nil, language: nil)
+        expect(Gitlab::Highlight).to receive(:highlight).with('files/ruby/regex.rb', "line one\n", plain: nil, language: 'ruby')
 
         presenter.highlight(to: 1)
       end
@@ -247,6 +247,36 @@ RSpec.describe BlobPresenter do
     end
   end
 
+  describe '#blob_language' do
+    subject { presenter.blob_language }
+
+    it { is_expected.to eq('ruby') }
+
+    context 'gitlab-language contains a match' do
+      before do
+        allow(blob).to receive(:language_from_gitattributes).and_return('cpp')
+      end
+
+      it { is_expected.to eq('cpp') }
+    end
+
+    context 'when blob is ipynb' do
+      let(:blob) { repository.blob_at('f6b7a707', 'files/ipython/markdown-table.ipynb') }
+
+      before do
+        allow(Gitlab::Diff::CustomDiff).to receive(:transformed_for_diff?).and_return(true)
+      end
+
+      it { is_expected.to eq('md') }
+    end
+
+    context 'when blob is binary' do
+      let(:blob) { repository.blob_at('HEAD', 'Gemfile.zip') }
+
+      it { is_expected.to be_nil }
+    end
+  end
+
   describe '#raw_plain_data' do
     let(:blob) { repository.blob_at('HEAD', file) }