xbmc/utils/Utf8Utils.h

   1 /*
   2  *  Copyright (C) 2013-2018 Team Kodi
   3  *  This file is part of Kodi - https://kodi.tv
   4  *
   5  *  SPDX-License-Identifier: GPL-2.0-or-later
   6  *  See LICENSES/README.md for more information.
   7  */
   8
   9 #pragma once
  10
  11 #include <string>
  12
  13 class CUtf8Utils
  14 {
  15 public:
  16   enum utf8CheckResult
  17   {
  18     plainAscii = -1, // only US-ASCII characters (valid for UTF-8 too)
  19     hiAscii    =  0, // non-UTF-8 sequence with high ASCII characters
  20                      // (possible single-byte national encoding like WINDOWS-1251, multi-byte encoding like UTF-32 or invalid UTF-8)
  21     utf8string =  1  // valid UTF-8 sequences, but not US-ASCII only
  22   };
  23
  24   /**
  25    * Check given string for valid UTF-8 sequences
  26    * @param str string to check
  27    * @return result of check, "plainAscii" for empty string
  28    */
  29   static utf8CheckResult checkStrForUtf8(const std::string& str);
  30
  31   static inline bool isValidUtf8(const std::string& str)
  32   {
  33     return checkStrForUtf8(str) != hiAscii;
  34   }
  35
  36   static size_t FindValidUtf8Char(const std::string& str, const size_t startPos = 0);
  37   static size_t RFindValidUtf8Char(const std::string& str, const size_t startPos);
  38
  39   static size_t SizeOfUtf8Char(const std::string& str, const size_t charStart = 0);
  40 private:
  41   static size_t SizeOfUtf8Char(const char* const str);
  42 };