[GN/iOS/Mac] Decouple copy_bundle_data from gyp mac_tool.py.

build/toolchain/mac/copy_data.py

Index: build/toolchain/mac/copy_data.py
diff --git a/build/toolchain/mac/copy_data.py b/build/toolchain/mac/copy_data.py
new file mode 100644
index 0000000000000000000000000000000000000000..6809b79f3111dc58b2fe3f0c361c9e93e4b343f3
--- /dev/null
+++ b/build/toolchain/mac/copy_data.py
@@ -0,0 +1,95 @@
+# Copyright 2016 The Chromium Authors. All rights reserved.

[Robert Sesek, 2016/05/05 14:38:10]

Why not name this copy_bundle_data.py to match the tool name?

[sdefresne, 2016/05/09 07:47:37]

Renamed (there was no other reason than lazyness for the different name).

+# Use of this source code is governed by a BSD-style license that can be
+# found in the LICENSE file.
+
+import argparse
+import os
+import shutil
+import sys
+
+from CoreFoundation import CFDataCreate, CFPropertyListCreateFromXMLData

[Robert Sesek, 2016/05/05 14:38:10]

Moving this import into copy_strings_file will speed up invocations for
non-.strings significantly.

[sdefresne, 2016/05/09 07:47:37]

Done.

+
+
+def detect_encoding(data, default_encoding='UTF-8'):
+  '''Detects the encoding used by |data| from the Byte-Order-Mark if present.

[Robert Sesek, 2016/05/05 14:38:10]

Docstrings should use """ per PEP-8.

[sdefresne, 2016/05/09 07:47:37]

Done.

+
+  Args:
+    data: string whose encoding needs to be detected
+    default_encoding: encoding returned if no BOM is found.
+
+  Returns:
+    The encoding determined from the BOM if present or |default_encoding| if
+    no BOM was found.
+  '''
+  if data.startswith('\xFE\xFF'):
+    return 'UTF-16BE'
+
+  if data.startswith('\xFF\xFE'):
+    return 'UTF-16LE'
+
+  if data.startswith('\xEF\xBB\xBF'):
+    return 'UTF-8'
+
+  return default_encoding
+
+
+def copy_strings_file(source, dest):
+  '''Copies a .strings file from |source| to |dest| and convert it to UTF-16.
+
+  Args:
+    source: string, path to the source file
+    dest: string, path to the destination file
+  '''
+  with open(source, 'rb') as source_file:
+    data = source_file.read()
+
+  # Xcode's CpyCopyStringsFile / builtin-copyStrings seems to call
+  # CFPropertyListCreateFromXMLData() behind the scenes; at least it prints
+  #     CFPropertyListCreateFromXMLData(): Old-style plist parser: missing
+  #     semicolon in dictionary.
+  # on invalid files. Do the same kind of validation.
+  core_data = CFDataCreate(None, data, len(data))

[Robert Sesek, 2016/05/05 14:38:10]

naming: cfdata? 

[sdefresne, 2016/05/09 07:47:37]

Done.

+  _, error = CFPropertyListCreateFromXMLData(None, core_data, 0, None)
+  if error:
+    sys.exit(1)

[Robert Sesek, 2016/05/05 14:38:10]

Maybe raise instead, so you get a stack trace?

[sdefresne, 2016/05/09 07:47:37]

Done.

+
+  encoding = detect_encoding(data)
+  with open(dest, 'wb') as dest_file:
+    dest_file.write(data.decode(encoding).encode('UTF-16'))
+
+
+def copy_file(source, dest):
+  '''Copies a file or directory from |source| to |dest|.
+
+  Args:
+    source: string, path to the source file
+    dest: string, path to the destination file
+  '''
+  if os.path.isdir(source):

[Robert Sesek, 2016/05/05 14:38:10]

There's a TODO(thakis) about mtimes in mac_tool.py. Does that still apply?

[sdefresne, 2016/05/09 07:47:37]

I think it still apply, copied.

+    if os.path.exists(dest):
+      shutil.rmtree(dest)
+    shutil.copytree(source, dest)
+    return
+
+  if os.path.exists(dest):
+    os.unlink(dest)
+
+  _, extension = os.path.splitext(source)
+  if extension == '.strings':
+    copy_strings_file(source, dest)
+    return
+
+  shutil.copy(source, dest)
+
+
+def main():
+  parser = argparse.ArgumentParser(
+      description='copy source to destination for the creation of a bundle')
+  parser.add_argument('source', help='path to source file or directory')
+  parser.add_argument('dest', help='path to destination')
+  args = parser.parse_args()
+
+  copy_file(args.source, args.dest)
+
+if __name__ == '__main__':
+  main()