Add deduplication logic to .pak files

tools/grit/grit/format/data_pack_unittest.py

Index: tools/grit/grit/format/data_pack_unittest.py
diff --git a/tools/grit/grit/format/data_pack_unittest.py b/tools/grit/grit/format/data_pack_unittest.py
index f6e9edcfcb6c99799e76d8c4230c3fddb66793aa..4fae80c7ea8bc85e9bf1e812f1040bfb149dac36 100755
--- a/tools/grit/grit/format/data_pack_unittest.py
+++ b/tools/grit/grit/format/data_pack_unittest.py
@@ -17,8 +17,8 @@ from grit.format import data_pack
 
 
 class FormatDataPackUnittest(unittest.TestCase):
-  def testWriteDataPack(self):
-    expected = (
+  def testReadDataPackV4(self):
+    expected_data = (
         '\x04\x00\x00\x00'                  # header(version
         '\x04\x00\x00\x00'                  #        no. entries,
         '\x01'                              #        encoding)
@@ -28,9 +28,40 @@ class FormatDataPackUnittest(unittest.TestCase):
         '\x0a\x00\x3f\x00\x00\x00'          # index entry 10
         '\x00\x00\x3f\x00\x00\x00'          # extra entry for the size of last
         'this is id 4this is id 6')         # data
-    input = {1: '', 4: 'this is id 4', 6: 'this is id 6', 10: ''}
-    output = data_pack.WriteDataPackToString(input, data_pack.UTF8)
-    self.failUnless(output == expected)
+    expected_resources = {
+        1: '',
+        4: 'this is id 4',
+        6: 'this is id 6',
+        10: '',
+    }
+    expected_data_pack = data_pack.DataPackContents(
+        expected_resources, data_pack.UTF8)
+    loaded = data_pack.ReadDataPackFromString(expected_data)
+    self.assertEquals(loaded, expected_data_pack)
+
+  def testReadWriteDataPackV5(self):
+    expected_data = (
+        '\x05\x00\x01\x00'                  # version, encoding
+        '\x03\x00\x01\x00'                  # resource_count, alias_count,

[flackr, 2017/07/07 18:54:12]

nit: I think the test will read more cleanly with one value per line than
putting both (version, encoding) and (resource_count, alias_count) into single
lines.

[agrieve, 2017/07/07 20:47:08]

Done.

+        '\x01\x00\x24\x00\x00\x00'          # index entry 1
+        '\x04\x00\x24\x00\x00\x00'          # index entry 4
+        '\x06\x00\x30\x00\x00\x00'          # index entry 6
+        '\x00\x00\x3c\x00\x00\x00'          # extra entry for the size of last
+        '\x0a\x00\x01\x00'                  # alias table
+        'this is id 4this is id 6')         # data
+    expected_resources = {
+        1: '',
+        4: 'this is id 4',
+        6: 'this is id 6',
+        10: 'this is id 4',
+    }
+    data = data_pack.WriteDataPackToString(expected_resources, data_pack.UTF8)
+    self.assertEquals(data, expected_data)
+
+    expected_data_pack = data_pack.DataPackContents(
+        expected_resources, data_pack.UTF8)
+    loaded = data_pack.ReadDataPackFromString(expected_data)
+    self.assertEquals(loaded, expected_data_pack)
 
   def testRePackUnittest(self):
     expected_with_whitelist = {
@@ -50,12 +81,14 @@ class FormatDataPackUnittest(unittest.TestCase):
               in inputs]
 
     # RePack using whitelist
-    output, _ = data_pack.RePackFromDataPackStrings(inputs, whitelist)
+    output, _ = data_pack.RePackFromDataPackStrings(
+        inputs, whitelist, suppress_removed_key_output=True)
     self.assertDictEqual(expected_with_whitelist, output,
                          'Incorrect resource output')
 
     # RePack a None whitelist
-    output, _ = data_pack.RePackFromDataPackStrings(inputs, None)
+    output, _ = data_pack.RePackFromDataPackStrings(
+        inputs, None, suppress_removed_key_output=True)
     self.assertDictEqual(expected_without_whitelist, output,
                          'Incorrect resource output')