From 1fd59361b5c4ce323d633261377a9175d3f8f1f9 Mon Sep 17 00:00:00 2001
From: Vicent Marti <tanoku@gmail.com>
Date: Wed, 25 Jun 2014 20:26:44 +0200
Subject: [PATCH] Proper incremental diffing

---
 lib/linguist/repository.rb | 18 +++++++++---------
 test/test_repository.rb    | 33 +++++++++++++++++++++++++++------
 2 files changed, 36 insertions(+), 15 deletions(-)

diff --git a/lib/linguist/repository.rb b/lib/linguist/repository.rb
index 9fb6ebc7..e11d2b23 100644
--- a/lib/linguist/repository.rb
+++ b/lib/linguist/repository.rb
@@ -14,10 +14,10 @@ module Linguist
     # Public: Initialize a new Repository
     #
     # Returns a Repository
-    def initialize(repo, sha1, existing_stats = nil)
+    def initialize(repo, commit_oid, existing_stats = nil)
       @repository = repo
-      @current_sha1 = sha1
-      @old_sha1, @old_stats = existing_stats if existing_stats
+      @commit_oid = commit_oid
+      @old_commit_oid, @old_stats = existing_stats if existing_stats
     end
 
     # Public: Returns a breakdown of language stats.
@@ -66,12 +66,12 @@ module Linguist
       end
     end
 
-    def incremental_stats(old_sha1, new_sha1, cache = nil)
+    def compute_stats(old_commit_oid, commit_oid, cache = nil)
       file_map = cache ? cache.dup : {}
-      old_commit = old_sha1 && Rugged::Commit.lookup(repository, old_sha1)
-      new_commit = Rugged::Commit.lookup(repository, new_sha1)
+      old_tree = old_commit_oid && Rugged::Commit.lookup(repository, old_commit_oid).tree
+      new_tree = Rugged::Commit.lookup(repository, commit_oid).tree
 
-      diff = Rugged::Tree.diff(repository, old_commit, new_commit)
+      diff = Rugged::Tree.diff(repository, old_tree, new_tree)
 
       diff.each_delta do |delta|
         old = delta.old_file[:path]
@@ -99,10 +99,10 @@ module Linguist
 
     def cache
       @cache ||= begin
-        if @old_sha1 == @current_sha1
+        if @old_commit_oid == @commit_oid
           @old_stats
         else
-          incremental_stats(@old_sha1, @current_sha1, @old_stats)
+          compute_stats(@old_commit_oid, @commit_oid, @old_stats)
         end
       end
     end
diff --git a/test/test_repository.rb b/test/test_repository.rb
index 13dbdaef..bd96e66d 100644
--- a/test/test_repository.rb
+++ b/test/test_repository.rb
@@ -3,19 +3,24 @@ require 'linguist/repository'
 require 'test/unit'
 
 class TestRepository < Test::Unit::TestCase
-  include Linguist
+  def rugged_repository
+    @rugged ||= Rugged::Repository.new(File.expand_path("../../.git", __FILE__))
+  end
 
-  def linguist_repo
-    r = Rugged::Repository.new(File.expand_path("../../.git", __FILE__))
-    Linguist::Repository.new(r, '31921838cdc252536ec07668f73d4b64d8022750')
+  def master_oid
+    @master_oid ||= Rugged::Object.rev_parse_oid(rugged_repository, 'master')
+  end
+
+  def linguist_repo(oid = master_oid)
+    Linguist::Repository.new(rugged_repository, oid)
   end
 
   def test_linguist_language
-    # assert_equal Language['Ruby'], linguist_repo.language
+    assert_equal 'Ruby', linguist_repo.language
   end
 
   def test_linguist_languages
-    # assert linguist_repo.languages[Language['Ruby']] > 10_000
+    assert linguist_repo.languages['Ruby'] > 10_000
   end
 
   def test_linguist_size
@@ -27,4 +32,20 @@ class TestRepository < Test::Unit::TestCase
     assert linguist_repo.breakdown_by_file["Ruby"].include?("bin/linguist")
     assert linguist_repo.breakdown_by_file["Ruby"].include?("lib/linguist/language.rb")
   end
+
+  def test_incremental_stats
+    old_commit = Rugged::Object.rev_parse_oid(rugged_repository, 'v2.0.0')
+    old_repo = linguist_repo(old_commit)
+
+    assert old_repo.languages['Ruby'] > 10_000
+    assert old_repo.size > 30_000
+
+    old_cache = [old_commit, old_repo.cache]
+    new_repo = Linguist::Repository.new(rugged_repository, master_oid, old_cache)
+
+    assert new_repo.languages['Ruby'] > old_repo.languages['Ruby']
+    assert new_repo.size > old_repo.size
+
+    assert_equal linguist_repo.cache, new_repo.cache
+  end
 end