From 3d39e842eca0913ba5ec385429496ceacc3699f8 Mon Sep 17 00:00:00 2001
From: Vicent Marti <tanoku@gmail.com>
Date: Wed, 10 Sep 2014 16:59:43 +0200
Subject: [PATCH] Load Git Attributes for Linguist-specific overrides

---
 lib/linguist/lazy_blob.rb  | 24 ++++++++++++++++++++++++
 lib/linguist/repository.rb |  6 +++++-
 test/test_repository.rb    | 14 ++++++++++++++
 3 files changed, 43 insertions(+), 1 deletion(-)

diff --git a/lib/linguist/lazy_blob.rb b/lib/linguist/lazy_blob.rb
index bb262241..c7031283 100644
--- a/lib/linguist/lazy_blob.rb
+++ b/lib/linguist/lazy_blob.rb
@@ -1,8 +1,13 @@
 require 'linguist/blob_helper'
+require 'linguist/language'
 require 'rugged'
 
 module Linguist
   class LazyBlob
+    GIT_ATTR = ['linguist-ignore', 'linguist-lang']
+    GIT_ATTR_OPTS = { :priority => [:index], :skip_system => true }
+    GIT_ATTR_FLAGS = Rugged::Repository::Attributes.parse_opts(GIT_ATTR_OPTS)
+
     include BlobHelper
 
     MAX_SIZE = 128 * 1024
@@ -19,6 +24,25 @@ module Linguist
       @mode = mode
     end
 
+    def git_attributes
+      @git_attributes ||= repository.fetch_attributes(
+        name, GIT_ATTR, GIT_ATTR_FLAGS)
+    end
+
+    def ignored?
+      !!git_attributes['linguist-ignore']
+    end
+
+    def overriden_language
+      if lang = git_attributes['linguist-lang']
+        Language.find_by_name(lang)
+      end
+    end
+
+    def language
+      @language ||= (overriden_language || Language.detect(self))
+    end
+
     def data
       load_blob!
       @data
diff --git a/lib/linguist/repository.rb b/lib/linguist/repository.rb
index a89c81e6..9e0b7f1e 100644
--- a/lib/linguist/repository.rb
+++ b/lib/linguist/repository.rb
@@ -123,6 +123,10 @@ module Linguist
 
       diff = Rugged::Tree.diff(repository, old_tree, new_tree)
 
+      attr_index = Rugged::Index.new
+      attr_index.read_tree(new_tree)
+      repository.index = attr_index
+
       diff.each_delta do |delta|
         old = delta.old_file[:path]
         new = delta.new_file[:path]
@@ -138,7 +142,7 @@ module Linguist
           blob = Linguist::LazyBlob.new(repository, delta.new_file[:oid], new, mode.to_s(8))
 
           # Skip vendored or generated blobs
-          next if blob.vendored? || blob.generated? || blob.language.nil?
+          next if blob.ignored? || blob.vendored? || blob.generated? || blob.language.nil?
 
           # Only include programming languages and acceptable markup languages
           if blob.language.type == :programming || Language.detectable_markup.include?(blob.language.name)
diff --git a/test/test_repository.rb b/test/test_repository.rb
index f4c5ad70..4f7b938e 100644
--- a/test/test_repository.rb
+++ b/test/test_repository.rb
@@ -47,4 +47,18 @@ class TestRepository < Test::Unit::TestCase
 
     assert_equal linguist_repo.cache, new_repo.cache
   end
+
+  def test_git_attributes
+    attr_commit = '525304738ebdb7ab3b7d2bf9a7514cc428faa273'
+    repo = linguist_repo(attr_commit)
+
+    assert repo.breakdown_by_file.has_key?("Java")
+    assert repo.breakdown_by_file["Java"].include?("lib/linguist.rb")
+
+    assert repo.breakdown_by_file.has_key?("Ruby")
+    assert !repo.breakdown_by_file["Ruby"].empty?
+    repo.breakdown_by_file["Ruby"].each do |file|
+      assert !file.start_with?("test/")
+    end
+  end
 end