Seperate test file for samples

Rakefile

@@ -16,7 +16,7 @@ end
 
 CLOBBER.include 'lib/linguist/samples.yml'
 
-task :classifier => [:clobber, 'lib/linguist/samples.yml']
+task :samples => [:clobber, 'lib/linguist/samples.yml']
 
 namespace :classifier do
   LIMIT = 1_000

test/test_classifier.rb

@@ -16,13 +16,6 @@ class TestClassifier < Test::Unit::TestCase
     File.read(File.join(samples_path, name))
   end
 
-  def test_instance_freshness
-    # Just warn, it shouldn't scare people off by breaking the build.
-    if Samples::DATA['md5'] != Samples.data['md5']
-      warn "Classifier database is out of date. Run `bundle exec rake classifier`."
-    end
-  end
-
   def test_classify
     db = {}
     Classifier.train! db, "Ruby", fixture("ruby/foo.rb")
@@ -59,14 +52,6 @@ class TestClassifier < Test::Unit::TestCase
     assert_equal [], Classifier.classify(Samples::DATA, nil)
   end
 
-  def test_verify
-    data = Samples::DATA
-
-    assert_equal data['languages_total'], data['languages'].inject(0) { |n, (_, c)| n += c }
-    assert_equal data['tokens_total'], data['language_tokens'].inject(0) { |n, (_, c)| n += c }
-    assert_equal data['tokens_total'], data['tokens'].inject(0) { |n, (_, ts)| n += ts.inject(0) { |m, (_, c)| m += c } }
-  end
-
   def test_classify_ambiguous_languages
     Samples.each do |sample|
       language = Linguist::Language.find_by_alias(sample[:language])

test/test_samples.rb

@@ -0,0 +1,25 @@
+require 'linguist/samples'
+
+require 'test/unit'
+
+class TestSamples < Test::Unit::TestCase
+  include Linguist
+
+  def test_up_to_date
+    assert serialized = Samples::DATA
+    assert latest = Samples.data
+
+    # Just warn, it shouldn't scare people off by breaking the build.
+    if serialized['md5'] != latest['md5']
+      warn "Samples database is out of date. Run `bundle exec rake samples`."
+    end
+  end
+
+  def test_verify
+    assert data = Samples::DATA
+
+    assert_equal data['languages_total'], data['languages'].inject(0) { |n, (_, c)| n += c }
+    assert_equal data['tokens_total'], data['language_tokens'].inject(0) { |n, (_, c)| n += c }
+    assert_equal data['tokens_total'], data['tokens'].inject(0) { |n, (_, ts)| n += ts.inject(0) { |m, (_, c)| m += c } }
+  end
+end