mirror of
https://github.com/VCMP-SqMod/SqMod.git
synced 2024-11-08 16:57:16 +01:00
4a6bfc086c
Switched to POCO library for unified platform/library interface. Deprecated the external module API. It was creating more problems than solving. Removed most built-in libraries in favor of system libraries for easier maintenance. Cleaned and secured code with help from static analyzers.
62 lines
1.6 KiB
C++
62 lines
1.6 KiB
C++
//
|
|
// UTF8Encoding.h
|
|
//
|
|
// Library: Foundation
|
|
// Package: Text
|
|
// Module: UTF8Encoding
|
|
//
|
|
// Definition of the UTF8Encoding class.
|
|
//
|
|
// Copyright (c) 2004-2007, Applied Informatics Software Engineering GmbH.
|
|
// and Contributors.
|
|
//
|
|
// SPDX-License-Identifier: BSL-1.0
|
|
//
|
|
|
|
|
|
#ifndef Foundation_UTF8Encoding_INCLUDED
|
|
#define Foundation_UTF8Encoding_INCLUDED
|
|
|
|
|
|
#include "Poco/Foundation.h"
|
|
#include "Poco/TextEncoding.h"
|
|
|
|
|
|
namespace Poco {
|
|
|
|
|
|
class Foundation_API UTF8Encoding: public TextEncoding
|
|
/// UTF-8 text encoding, as defined in RFC 2279.
|
|
{
|
|
public:
|
|
UTF8Encoding();
|
|
~UTF8Encoding();
|
|
const char* canonicalName() const;
|
|
bool isA(const std::string& encodingName) const;
|
|
const CharacterMap& characterMap() const;
|
|
int convert(const unsigned char* bytes) const;
|
|
int convert(int ch, unsigned char* bytes, int length) const;
|
|
int queryConvert(const unsigned char* bytes, int length) const;
|
|
int sequenceLength(const unsigned char* bytes, int length) const;
|
|
|
|
static bool isLegal(const unsigned char *bytes, int length);
|
|
/// Utility routine to tell whether a sequence of bytes is legal UTF-8.
|
|
/// This must be called with the length pre-determined by the first byte.
|
|
/// The sequence is illegal right away if there aren't enough bytes
|
|
/// available. If presented with a length > 4, this function returns false.
|
|
/// The Unicode definition of UTF-8 goes up to 4-byte sequences.
|
|
///
|
|
/// Adapted from ftp://ftp.unicode.org/Public/PROGRAMS/CVTUTF/ConvertUTF.c
|
|
/// Copyright 2001-2004 Unicode, Inc.
|
|
|
|
private:
|
|
static const char* _names[];
|
|
static const CharacterMap _charMap;
|
|
};
|
|
|
|
|
|
} // namespace Poco
|
|
|
|
|
|
#endif // Foundation_UTF8Encoding_INCLUDED
|