add a repack utility for combining .pak files generated

by grit into a single .pak file.

Review URL: http://codereview.chromium.org/18254


git-svn-id: svn://svn.chromium.org/chrome/trunk/src@8063 0039d316-1c4b-4281-b951-d872f2087c98

2 files changed, 72 insertions, 6 deletions

diff --git a/tools/data_pack/data_pack.py b/tools/data_pack/data_pack.py
index c269f21..fde9483 100755
--- a/tools/data_pack/data_pack.py
+++ b/tools/data_pack/data_pack.py
@@ -9,7 +9,32 @@ See base/pack_file* for details.
 
 import struct
 
-version = 1
+FILE_FORMAT_VERSION = 1
+HEADER_LENGTH = 2 * 4  # Two uint32s. (file version and number of entries)
+
+class WrongFileVersion(Exception):
+  pass
+
+def ReadDataPack(input_file):
+  """Reads a data pack file and returns a dictionary."""
+  data = open(input_file, "rb").read()
+  original_data = data
+
+  # Read the header.
+  version, num_entries = struct.unpack("<II", data[:HEADER_LENGTH])
+  if version != FILE_FORMAT_VERSION:
+    raise WrongFileVersion
+
+  resources = {}
+  # Read the index and data.
+  data = data[HEADER_LENGTH:]
+  kIndexEntrySize = 3 * 4  # Each entry is 3 uint32s.
+  for _ in range(num_entries):
+    id, offset, length = struct.unpack("<III", data[:kIndexEntrySize])
+    data = data[kIndexEntrySize:]
+    resources[id] = original_data[offset:offset + length]
+
+  return resources
 
 def WriteDataPack(resources, output_file):
   """Write a map of id=>data into output_file as a data pack."""
@@ -17,13 +42,12 @@ def WriteDataPack(resources, output_file):
   file = open(output_file, "wb")
 
   # Write file header.
-  file.write(struct.pack("<II", version, len(ids)))
-  header_length = 2 * 4             # Two uint32s.
+  file.write(struct.pack("<II", FILE_FORMAT_VERSION, len(ids)))
 
   index_length = len(ids) * 3 * 4   # Each entry is 3 uint32s.
 
   # Write index.
-  data_offset = header_length + index_length
+  data_offset = HEADER_LENGTH + index_length
   for id in ids:
     file.write(struct.pack("<III", id, data_offset, len(resources[id])))
     data_offset += len(resources[id])
@@ -35,8 +59,10 @@ def WriteDataPack(resources, output_file):
 def main():
   # Just write a simple file.
   data = { 1: "", 4: "this is id 4", 6: "this is id 6", 10: "" }
-  WriteDataPack(data, "datapack")
-  print "wrote datapack to current directory."
+  WriteDataPack(data, "datapack1.pak")
+  data2 = { 1000: "test", 5: "five" }
+  WriteDataPack(data2, "datapack2.pak")
+  print "wrote datapack1 and datapack2 to current directory."
 
 if __name__ == '__main__':
   main()
diff --git a/tools/data_pack/repack.py b/tools/data_pack/repack.py
new file mode 100755
index 0000000..5d7a344
--- /dev/null
+++ b/tools/data_pack/repack.py
@@ -0,0 +1,40 @@
+#!/usr/bin/python
+# Copyright (c) 2008 The Chromium Authors. All rights reserved.
+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+"""A simple utility function to merge data pack files into a single data pack.
+See base/pack_file* for details about the file format.
+"""
+
+import exceptions
+import struct
+import sys
+
+import data_pack
+
+def RePack(output_file, input_files):
+  """Write a new data pack to |output_file| based on a list of filenames
+  (|input_files|)"""
+  resources = {}
+  for filename in input_files:
+    new_resources = data_pack.ReadDataPack(filename)
+
+    # Make sure we have no dups.
+    duplicate_keys = set(new_resources.keys()) & set(resources.keys())
+    if len(duplicate_keys) != 0:
+      raise exceptions.KeyError("Duplicate keys: " + str(list(duplicate_keys)))
+
+    resources.update(new_resources)
+
+  data_pack.WriteDataPack(resources, output_file)
+
+def main(argv):
+  if len(argv) < 4:
+    print ("Usage:\n  %s <output_filename> <input_file1> <input_file2> "
+        "[input_file3] ..." % argv[0])
+    sys.exit(-1)
+  RePack(argv[1], argv[2:])
+
+if '__main__' == __name__:
+  main(sys.argv)