unicode.cpp

Go to the documentation of this file.

/*******************************************************************\
 
Module:
 
Author: Daniel Kroening, kroening@kroening.com
 
\*******************************************************************/
 
#include "unicode.h"
 
#include "invariant.h"
 
#include <codecvt>
#include <cstdint>
#include <iomanip>
#include <locale>
#include <sstream>
 
#ifdef _WIN32
#  include <util/pragma_push.def>
#  ifdef _MSC_VER
#    pragma warning(disable : 4668)
// using #if/#elif on undefined macro
#    pragma warning(disable : 5039)
// pointer or reference to potentially throwing function passed to extern C
#  endif
#  include <util/pragma_pop.def>
#  include <windows.h>
#endif
 
static void utf8_append_code(unsigned int c, std::string &);
 
std::string narrow(const wchar_t *s)
{
#ifdef _WIN32
 
  int slength = static_cast<int>(wcslen(s));
  int rlength =
    WideCharToMultiByte(CP_UTF8, 0, s, slength, NULL, 0, NULL, NULL);
  std::string r(rlength, 0);
  WideCharToMultiByte(CP_UTF8, 0, s, slength, &r[0], rlength, NULL, NULL);
  return r;
 
#else
  return narrow(std::wstring(s));
#endif
}
 
std::wstring widen(const char *s)
{
#ifdef _WIN32
 
  int slength = static_cast<int>(strlen(s));
  int rlength = MultiByteToWideChar(CP_UTF8, 0, s, slength, NULL, 0);
  std::wstring r(rlength, 0);
  MultiByteToWideChar(CP_UTF8, 0, s, slength, &r[0], rlength);
  return r;
 
#else
  return widen(std::string(s));
#endif
}
 
std::string narrow(const std::wstring &s)
{
#ifdef _WIN32
 
  int slength = static_cast<int>(s.size());
  int rlength =
    WideCharToMultiByte(CP_UTF8, 0, &s[0], slength, NULL, 0, NULL, NULL);
  std::string r(rlength, 0);
  WideCharToMultiByte(CP_UTF8, 0, &s[0], slength, &r[0], rlength, NULL, NULL);
  return r;
 
#else
  std::string result;
 
  result.reserve(s.size()); // at least that long
 
  for(const auto codepoint : s)
    utf8_append_code(codepoint, result);
 
  return result;
#endif
}
 
std::wstring widen(const std::string &s)
{
#ifdef _WIN32
 
  int slength = static_cast<int>(s.size());
  int rlength = MultiByteToWideChar(CP_UTF8, 0, &s[0], slength, NULL, 0);
  std::wstring r(rlength, 0);
  MultiByteToWideChar(CP_UTF8, 0, &s[0], slength, &r[0], rlength);
  return r;
 
#else
  auto utf32 = utf8_to_utf32(std::string(s));
 
  std::wstring r;
  r.reserve(utf32.size());
  for(auto codepoint : utf32)
    r += codepoint;
  return r;
#endif
}
 
static void utf8_append_code(unsigned int c, std::string &result)
{
  if(c <= 0x7f)
    result += static_cast<char>(c);
  else if(c <= 0x7ff)
  {
    result += static_cast<char>((c >> 6) | 0xc0);
    result += static_cast<char>((c & 0x3f) | 0x80);
  }
  else if(c <= 0xffff)
  {
    result += static_cast<char>((c >> 12) | 0xe0);
    result += static_cast<char>(((c >> 6) & 0x3f) | 0x80);
    result += static_cast<char>((c & 0x3f) | 0x80);
  }
  else
  {
    result += static_cast<char>((c >> 18) | 0xf0);
    result += static_cast<char>(((c >> 12) & 0x3f) | 0x80);
    result += static_cast<char>(((c >> 6) & 0x3f) | 0x80);
    result += static_cast<char>((c & 0x3f) | 0x80);
  }
}
 
std::string utf32_native_endian_to_utf8(const std::basic_string<char32_t> &s)
{
  std::string result;
 
  result.reserve(s.size()); // at least that long
 
  for(const auto c : s)
    utf8_append_code(c, result);
 
  return result;
}
 
std::vector<std::string> narrow_argv(int argc, const wchar_t **argv_wide)
{
  if(argv_wide == nullptr)
    return std::vector<std::string>();
 
  std::vector<std::string> argv_narrow;
  argv_narrow.reserve(argc);
 
  for(int i = 0; i != argc; ++i)
    argv_narrow.push_back(narrow(argv_wide[i]));
 
  return argv_narrow;
}
 
static void utf16_append_code(unsigned int code, std::wstring &result)
{
  // we do not treat 0xD800 to 0xDFFF, although
  // they are not valid unicode symbols
 
  if(code < 0xFFFF)
  {
    // code is encoded as one UTF16 character
    result += static_cast<wchar_t>(code);
  }
  else // code is encoded as two UTF16 characters
  {
    // if this is valid unicode, we have
    // code<0x10FFFF
    // but let's not check it programmatically
 
    // encode the code in UTF16
    code = code - 0x10000;
    const uint16_t i1 = static_cast<uint16_t>(((code >> 10) & 0x3ff) | 0xD800);
    result += static_cast<wchar_t>(i1);
    const uint16_t i2 = static_cast<uint16_t>((code & 0x3ff) | 0xDC00);
    result += static_cast<wchar_t>(i2);
  }
}
 
std::wstring utf8_to_utf16_native_endian(const std::string &in)
{
  std::wstring result;
  result.reserve(in.size());
 
  for(auto codepoint : utf8_to_utf32(in))
    utf16_append_code(codepoint, result);
 
  return result;
}
 
std::u32string utf8_to_utf32(const std::string &utf8_str)
{
  std::u32string result;
  result.reserve(utf8_str.size());
  std::string::size_type i = 0;
  while(i < utf8_str.size())
  {
    unsigned char c = utf8_str[i++];
    char32_t code = 0;
    // the ifs that follow find out how many UTF8 characters (1-4) store the
    // next unicode character. This is determined by the few most
    // significant bits.
    if(c <= 0x7F)
    {
      // if it's one character, then code is exactly the value
      code = c;
    }
    else if(c <= 0xDF && i < utf8_str.size())
    { // in other cases, we need to read the right number of chars and decode
      // note: if we wanted to make sure that we capture incorrect strings,
      // we should check that whatever follows first character starts with
      // bits 10.
      code = (c & 0x1Fu) << 6;
      c = utf8_str[i++];
      code += c & 0x3Fu;
    }
    else if(c <= 0xEF && i + 1 < utf8_str.size())
    {
      code = (c & 0xFu) << 12;
      c = utf8_str[i++];
      code += (c & 0x3Fu) << 6;
      c = utf8_str[i++];
      code += c & 0x3Fu;
    }
    else if(c <= 0xF7 && i + 2 < utf8_str.size())
    {
      code = (c & 0x7u) << 18;
      c = utf8_str[i++];
      code += (c & 0x3Fu) << 12;
      c = utf8_str[i++];
      code += (c & 0x3Fu) << 6;
      c = utf8_str[i++];
      code += c & 0x3Fu;
    }
    else
    {
      // The string is not a valid UTF8 string! Either it has some characters
      // missing from a multi-character unicode symbol, or it has a char with
      // too high value.
      // For now, let's replace the character with a space
      code = 32;
    }
 
    result.append(1, code);
  }
 
  return result;
}
 
static void utf16_native_endian_to_java_string(
  const wchar_t ch,
  std::ostringstream &result,
  const std::locale &loc)
{
  // \u unicode characters are translated very early by the Java compiler and so
  // \u000a or \u000d would become a newline character in a char constant, which
  // is illegal. Instead use \n or \r.
  if(ch == '\n')
    result << "\\n";
  else if(ch == '\r')
    result << "\\r";
  // \f, \b and \t do not need to be escaped, but this will improve readability
  // of generated tests.
  else if(ch == '\f')
    result << "\\f";
  else if(ch == '\b')
    result << "\\b";
  else if(ch == '\t')
    result << "\\t";
  else if(ch <= 255 && isprint(ch, loc))
  {
    const auto uch = static_cast<unsigned char>(ch);
    // ", and \ need to be escaped, but not ' for java strings
    // e.g. "\"\\" needs escaping but "'" does not.
    if(uch == '"' || uch == '\\')
      result << '\\';
    result << uch;
  }
  else
  {
    // Format ch as a hexadecimal unicode character padded to four digits with
    // zeros.
    result << "\\u" << std::hex << std::setw(4) << std::setfill('0')
           << static_cast<unsigned int>(ch);
  }
}
 
static void utf16_native_endian_to_java(
  const wchar_t ch,
  std::ostringstream &result,
  const std::locale &loc)
{
  if(ch == (wchar_t)'\'')
  {
    const auto uch = static_cast<unsigned char>(ch);
    // ' needs to be escaped for java characters, e.g. '\''
    result << '\\' << uch;
  }
  else
  {
    utf16_native_endian_to_java_string(ch, result, loc);
  }
}
 
std::string utf16_native_endian_to_java(const char16_t ch)
{
  std::ostringstream result;
  const std::locale loc;
  utf16_native_endian_to_java(ch, result, loc);
  return result.str();
}
 
std::string utf16_native_endian_to_java_string(const std::wstring &in)
{
  std::ostringstream result;
  const std::locale loc;
  for(const auto ch : in)
    utf16_native_endian_to_java_string(ch, result, loc);
  return result.str();
}
 
std::string utf16_native_endian_to_utf8(const char16_t utf16_char)
{
  return utf16_native_endian_to_utf8(std::u16string(1, utf16_char));
}
 
std::string utf16_native_endian_to_utf8(const std::u16string &utf16_str)
{
#ifdef _MSC_VER
  // Workaround for Visual Studio bug, see
  // https://stackoverflow.com/questions/32055357
  std::wstring wide_string(utf16_str.begin(), utf16_str.end());
  return std::wstring_convert<std::codecvt_utf8_utf16<wchar_t>, wchar_t>{}
    .to_bytes(wide_string);
#else
  return std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t>{}
    .to_bytes(utf16_str);
#endif
}
 
char16_t codepoint_hex_to_utf16_native_endian(const std::string &hex)
{
  PRECONDITION(hex.length() == 4);
  return std::strtol(hex.c_str(), nullptr, 16);
}
 
std::string codepoint_hex_to_utf8(const std::string &hex)
{
  return utf16_native_endian_to_utf8(codepoint_hex_to_utf16_native_endian(hex));
}