1#!/usr/bin/env python
2# Copyright (c) 2012 The Chromium Authors. All rights reserved.
3# Use of this source code is governed by a BSD-style license that can be
4# found in the LICENSE file.
5
6'''Support for formatting a data pack file used for platform agnostic resource
7files.
8'''
9
10import collections
11import exceptions
12import os
13import struct
14import sys
15if __name__ == '__main__':
16  sys.path.append(os.path.join(os.path.dirname(__file__), '../..'))
17
18from grit import util
19from grit.node import include
20from grit.node import message
21from grit.node import structure
22from grit.node import misc
23
24
25PACK_FILE_VERSION = 4
26HEADER_LENGTH = 2 * 4 + 1  # Two uint32s. (file version, number of entries) and
27                           # one uint8 (encoding of text resources)
28BINARY, UTF8, UTF16 = range(3)
29
30
31class WrongFileVersion(Exception):
32  pass
33
34
35DataPackContents = collections.namedtuple(
36    'DataPackContents', 'resources encoding')
37
38
39def Format(root, lang='en', output_dir='.'):
40  '''Writes out the data pack file format (platform agnostic resource file).'''
41  data = {}
42  for node in root.ActiveDescendants():
43    with node:
44      if isinstance(node, (include.IncludeNode, message.MessageNode,
45                           structure.StructureNode)):
46        id, value = node.GetDataPackPair(lang, UTF8)
47        if value is not None:
48          data[id] = value
49  return WriteDataPackToString(data, UTF8)
50
51
52def ReadDataPack(input_file):
53  """Reads a data pack file and returns a dictionary."""
54  data = util.ReadFile(input_file, util.BINARY)
55  original_data = data
56
57  # Read the header.
58  version, num_entries, encoding = struct.unpack("<IIB", data[:HEADER_LENGTH])
59  if version != PACK_FILE_VERSION:
60    print "Wrong file version in ", input_file
61    raise WrongFileVersion
62
63  resources = {}
64  if num_entries == 0:
65    return DataPackContents(resources, encoding)
66
67  # Read the index and data.
68  data = data[HEADER_LENGTH:]
69  kIndexEntrySize = 2 + 4  # Each entry is a uint16 and a uint32.
70  for _ in range(num_entries):
71    id, offset = struct.unpack("<HI", data[:kIndexEntrySize])
72    data = data[kIndexEntrySize:]
73    next_id, next_offset = struct.unpack("<HI", data[:kIndexEntrySize])
74    resources[id] = original_data[offset:next_offset]
75
76  return DataPackContents(resources, encoding)
77
78
79def WriteDataPackToString(resources, encoding):
80  """Write a map of id=>data into a string in the data pack format and return
81  it."""
82  ids = sorted(resources.keys())
83  ret = []
84
85  # Write file header.
86  ret.append(struct.pack("<IIB", PACK_FILE_VERSION, len(ids), encoding))
87  HEADER_LENGTH = 2 * 4 + 1            # Two uint32s and one uint8.
88
89  # Each entry is a uint16 + a uint32s. We have one extra entry for the last
90  # item.
91  index_length = (len(ids) + 1) * (2 + 4)
92
93  # Write index.
94  data_offset = HEADER_LENGTH + index_length
95  for id in ids:
96    ret.append(struct.pack("<HI", id, data_offset))
97    data_offset += len(resources[id])
98
99  ret.append(struct.pack("<HI", 0, data_offset))
100
101  # Write data.
102  for id in ids:
103    ret.append(resources[id])
104  return ''.join(ret)
105
106
107def WriteDataPack(resources, output_file, encoding):
108  """Write a map of id=>data into output_file as a data pack."""
109  content = WriteDataPackToString(resources, encoding)
110  with open(output_file, "wb") as file:
111    file.write(content)
112
113
114def RePack(output_file, input_files):
115  """Write a new data pack to |output_file| based on a list of filenames
116  (|input_files|)"""
117  resources = {}
118  encoding = None
119  for filename in input_files:
120    new_content = ReadDataPack(filename)
121
122    # Make sure we have no dups.
123    duplicate_keys = set(new_content.resources.keys()) & set(resources.keys())
124    if len(duplicate_keys) != 0:
125      raise exceptions.KeyError("Duplicate keys: " + str(list(duplicate_keys)))
126
127    # Make sure encoding is consistent.
128    if encoding in (None, BINARY):
129      encoding = new_content.encoding
130    elif new_content.encoding not in (BINARY, encoding):
131        raise exceptions.KeyError("Inconsistent encodings: " +
132                                  str(encoding) + " vs " +
133                                  str(new_content.encoding))
134
135    resources.update(new_content.resources)
136
137  # Encoding is 0 for BINARY, 1 for UTF8 and 2 for UTF16
138  if encoding is None:
139    encoding = BINARY
140  WriteDataPack(resources, output_file, encoding)
141
142
143# Temporary hack for external programs that import data_pack.
144# TODO(benrg): Remove this.
145class DataPack(object):
146  pass
147DataPack.ReadDataPack = staticmethod(ReadDataPack)
148DataPack.WriteDataPackToString = staticmethod(WriteDataPackToString)
149DataPack.WriteDataPack = staticmethod(WriteDataPack)
150DataPack.RePack = staticmethod(RePack)
151
152
153def main():
154  if len(sys.argv) > 1:
155    # When an argument is given, read and explode the file to text
156    # format, for easier diffing.
157    data = ReadDataPack(sys.argv[1])
158    print data.encoding
159    for (resource_id, text) in data.resources.iteritems():
160      print "%s: %s" % (resource_id, text)
161  else:
162    # Just write a simple file.
163    data = { 1: "", 4: "this is id 4", 6: "this is id 6", 10: "" }
164    WriteDataPack(data, "datapack1.pak", UTF8)
165    data2 = { 1000: "test", 5: "five" }
166    WriteDataPack(data2, "datapack2.pak", UTF8)
167    print "wrote datapack1 and datapack2 to current directory."
168
169
170if __name__ == '__main__':
171  main()
172