BaseTools/Scripts/ConvertUni.py

   1 ## @file
   2 #  Check a patch for various format issues
   3 #
   4 #  Copyright (c) 2015, Intel Corporation. All rights reserved.<BR>
   5 #
   6 #  This program and the accompanying materials are licensed and made
   7 #  available under the terms and conditions of the BSD License which
   8 #  accompanies this distribution. The full text of the license may be
   9 #  found at http://opensource.org/licenses/bsd-license.php
  10 #
  11 #  THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS"
  12 #  BASIS, WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER
  13 #  EXPRESS OR IMPLIED.
  14 #
  15
  16 from __future__ import print_function
  17
  18 VersionNumber = '0.1'
  19 __copyright__ = "Copyright (c) 2015, Intel Corporation  All rights reserved."
  20
  21 import argparse
  22 import codecs
  23 import os
  24 import sys
  25
  26 try:
  27     from io import StringIO
  28 except ImportError:
  29     from StringIO import StringIO
  30
  31 class ConvertOneArg:
  32     """Converts utf-16 to utf-8 for one command line argument.
  33
  34        This could be a single file, or a directory.
  35     """
  36
  37     def __init__(self, utf8, source):
  38         self.utf8 = utf8
  39         self.source = source
  40
  41         self.ok = True
  42
  43         if not os.path.exists(source):
  44             self.ok = False
  45         elif os.path.isdir(source):
  46             for (root, dirs, files) in os.walk(source):
  47                 files = filter(lambda a: a.endswith('.uni'), files)
  48                 for filename in files:
  49                     path = os.path.join(root, filename)
  50                     self.ok &= self.convert_one_file(path)
  51                     if not self.ok:
  52                         break
  53
  54                 if not self.ok:
  55                     break
  56         else:
  57             self.ok &= self.convert_one_file(source)
  58
  59     def convert_one_file(self, source):
  60         if self.utf8:
  61             new_enc, old_enc = 'utf-8', 'utf-16'
  62         else:
  63             new_enc, old_enc = 'utf-16', 'utf-8'
  64         #
  65         # Read file
  66         #
  67         f = open(source, mode='rb')
  68         file_content = f.read()
  69         f.close()
  70
  71         #
  72         # Detect UTF-16 Byte Order Mark at beginning of file.
  73         #
  74         bom = (file_content.startswith(codecs.BOM_UTF16_BE) or
  75                file_content.startswith(codecs.BOM_UTF16_LE))
  76         if bom != self.utf8:
  77             print("%s: already %s" % (source, new_enc))
  78             return True
  79
  80         #
  81         # Decode old string data
  82         #
  83         str_content = file_content.decode(old_enc, 'ignore')
  84
  85         #
  86         # Encode new string data
  87         #
  88         new_content = str_content.encode(new_enc, 'ignore')
  89
  90         #
  91         # Write converted data back to file
  92         #
  93         f = open(source, mode='wb')
  94         f.write(new_content)
  95         f.close()
  96
  97         print(source + ": converted, size", len(file_content), '=>', len(new_content))
  98         return True
  99
 100
 101 class ConvertUniApp:
 102     """Converts .uni files between utf-16 and utf-8."""
 103
 104     def __init__(self):
 105         self.parse_options()
 106         sources = self.args.source
 107
 108         self.ok = True
 109         for patch in sources:
 110             self.process_one_arg(patch)
 111
 112         if self.ok:
 113             self.retval = 0
 114         else:
 115             self.retval = -1
 116
 117     def process_one_arg(self, arg):
 118         self.ok &= ConvertOneArg(self.utf8, arg).ok
 119
 120     def parse_options(self):
 121         parser = argparse.ArgumentParser(description=__copyright__)
 122         parser.add_argument('--version', action='version',
 123                             version='%(prog)s ' + VersionNumber)
 124         parser.add_argument('source', nargs='+',
 125                             help='[uni file | directory]')
 126         group = parser.add_mutually_exclusive_group()
 127         group.add_argument("--utf-8",
 128                            action="store_true",
 129                            help="Convert from utf-16 to utf-8 [default]")
 130         group.add_argument("--utf-16",
 131                            action="store_true",
 132                            help="Convert from utf-8 to utf-16")
 133         self.args = parser.parse_args()
 134         self.utf8 = not self.args.utf_16
 135
 136 if __name__ == "__main__":
 137     sys.exit(ConvertUniApp().retval)