solidity/libdevcore/UTF8.cpp

142 lines
2.8 KiB
C++
Raw Normal View History

2016-08-05 09:56:59 +00:00
/*
This file is part of solidity.
2016-08-05 09:56:59 +00:00
solidity is free software: you can redistribute it and/or modify
2016-08-05 09:56:59 +00:00
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
solidity is distributed in the hope that it will be useful,
2016-08-05 09:56:59 +00:00
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with solidity. If not, see <http://www.gnu.org/licenses/>.
2016-08-05 09:56:59 +00:00
*/
/** @file UTF8.cpp
* @author Alex Beregszaszi
* @date 2016
*
* UTF-8 related helpers
*/
2019-03-28 12:07:52 +00:00
#include <libdevcore/UTF8.h>
2016-08-05 09:56:59 +00:00
namespace dev
{
2017-06-16 16:13:18 +00:00
namespace
{
/// Validate byte sequence against Unicode chapter 3 Table 3-7.
bool isWellFormed(unsigned char byte1, unsigned char byte2)
{
2017-06-22 12:37:38 +00:00
if (byte1 == 0xc0 || byte1 == 0xc1)
2017-06-16 16:13:18 +00:00
return false;
2017-06-22 12:37:38 +00:00
else if (byte1 >= 0xc2 && byte1 <= 0xdf)
return true;
else if (byte1 == 0xe0)
{
2017-06-16 16:13:18 +00:00
if (byte2 < 0xa0)
return false;
2017-06-22 12:37:38 +00:00
else
return true;
}
else if (byte1 >= 0xe1 && byte1 <= 0xec)
return true;
else if (byte1 == 0xed)
{
2017-06-16 16:13:18 +00:00
if (byte2 > 0x9f)
return false;
2017-06-22 12:37:38 +00:00
else
return true;
}
else if (byte1 == 0xee || byte1 == 0xef)
return true;
else if (byte1 == 0xf0)
{
2017-06-16 16:13:18 +00:00
if (byte2 < 0x90)
return false;
2017-06-22 12:37:38 +00:00
else
return true;
}
else if (byte1 >= 0xf1 && byte1 <= 0xf3)
return true;
else if (byte1 == 0xf4)
{
2017-06-16 16:13:18 +00:00
if (byte2 > 0x8f)
return false;
2017-06-22 12:37:38 +00:00
else
return true;
2017-06-16 16:13:18 +00:00
}
2017-06-22 12:37:38 +00:00
/// 0xf5 .. 0xf7 is disallowed
/// Technically anything below 0xc0 or above 0xf7 is
/// not possible to encode using Table 3-6 anyway.
return false;
2017-06-16 16:13:18 +00:00
}
2019-03-28 12:07:52 +00:00
}
2019-02-13 15:22:42 +00:00
bool validateUTF8(unsigned char const* _input, size_t _length, size_t& _invalidPosition)
2016-08-05 09:56:59 +00:00
{
bool valid = true;
2016-08-08 18:12:52 +00:00
size_t i = 0;
2016-08-05 09:56:59 +00:00
2017-06-22 10:24:15 +00:00
for (; i < _length; i++)
2016-08-05 09:56:59 +00:00
{
2017-06-16 16:13:18 +00:00
// Check for Unicode Chapter 3 Table 3-6 conformity.
2017-06-22 10:24:15 +00:00
if (_input[i] < 0x80)
2016-08-05 09:56:59 +00:00
continue;
2016-08-08 18:12:52 +00:00
size_t count = 0;
2017-06-22 12:37:38 +00:00
if (_input[i] >= 0xc0 && _input[i] <= 0xdf)
count = 1;
else if (_input[i] >= 0xe0 && _input[i] <= 0xef)
count = 2;
else if (_input[i] >= 0xf0 && _input[i] <= 0xf7)
count = 3;
2016-08-05 09:56:59 +00:00
if (count == 0)
{
valid = false;
break;
}
2017-06-22 10:24:15 +00:00
if ((i + count) >= _length)
2016-08-05 09:56:59 +00:00
{
valid = false;
break;
}
2016-08-08 18:12:52 +00:00
for (size_t j = 0; j < count; j++)
2016-08-05 09:56:59 +00:00
{
i++;
2016-08-08 18:12:11 +00:00
if ((_input[i] & 0xc0) != 0x80)
2016-08-05 09:56:59 +00:00
{
valid = false;
break;
}
2017-06-16 16:13:18 +00:00
// Check for Unicode Chapter 3 Table 3-7 conformity.
if ((j == 0) && !isWellFormed(_input[i - 1], _input[i]))
{
valid = false;
break;
}
2016-08-05 09:56:59 +00:00
}
}
if (valid)
return true;
2016-08-08 18:12:11 +00:00
_invalidPosition = i;
2016-08-05 09:56:59 +00:00
return false;
}
2017-06-22 10:24:15 +00:00
bool validateUTF8(std::string const& _input, size_t& _invalidPosition)
{
return validateUTF8(reinterpret_cast<unsigned char const*>(_input.c_str()), _input.length(), _invalidPosition);
}
}