3 This file is part of libmspstrings
4 Copyright © 2006-2007 Mikko Rasa
5 Distributed under the LGPL
8 #ifndef MSP_STRINGS_CODEC_H_
9 #define MSP_STRINGS_CODEC_H_
12 #include <msp/core/except.h>
17 typedef int UnicodeChar;
19 typedef std::basic_string<UnicodeChar> ustring;
29 An exception thrown for all kinds of problems encountered while encoding or
32 class CodecError: public Exception
35 CodecError(const std::string &w_): Exception(w_) { }
39 Base class for string codecs. Use one of the derived classes or the function
40 create_codec to create a specific codec.
42 Unicode strings are represented as ustrings. An std::string is considered to
43 be an encoded sequence of bytes. A codec is able to determine if an encoded
44 string could be decoded with it.
50 Base class for string encoder.
52 Each codec class should contain an Encoder class derived from this. The
53 encode_char and transliterate functions must be overloaded. Some encoders
54 may find it useful or necessary to implement some other functions too
55 (particularly sync and reset for stateful codecs).
62 Encoder(ErrorMode em): err_mode(em) { }
64 virtual ~Encoder() { }
66 /** Encodes a single unicode character. If the character can't be
67 represented in this encoding, error() should be called. */
68 virtual void encode_char(UnicodeChar ch, std::string &buf) = 0;
70 /** Encodes a unicode string. This is equivalent to calling encode_char
71 for each character in the string with the same buffer. */
72 virtual void encode(const ustring &str, std::string &buf);
74 std::string encode(const ustring &);
76 /** Procuces a sequence of bytes that will bring the encoder back to the
78 virtual void sync(std::string &buf) { (void)buf; }
80 /** Resets the encoder to the initial state without producing output. */
81 virtual void reset() { }
84 /** Handles an error depending on the error mode.
86 THROW_ON_ERROR: throws CodecError(msg)
87 IGNORE_ERRORS: does nothing
88 TRANSLITERATE: calls transliterate(ch, buf) */
89 void error(UnicodeChar ch, std::string &buf, const std::string &msg);
91 /** Attempts to produce an alternative encoding for a unicode character.
92 Typically this includes dropping accent marks or romanizing letters. */
93 virtual void transliterate(UnicodeChar ch, std::string &buf) = 0;
97 Base class for string decoder.
99 Each codec class should contain an Decoder class derived from this.
106 Decoder(ErrorMode em): err_mode(em) { }
108 virtual ~Decoder() { }
110 /** Decodes a single character from a string. The iterator is advanced
111 to the next character. For stateful codecs, -1 may be returned if a
112 state change sequence was decoded but no character followed it. If
113 invalid input is encountered, the error() function should be called and
114 the iterator advanced only if it doesn't throw. */
115 virtual UnicodeChar decode_char(const std::string &str, std::string::const_iterator &i) = 0;
117 /** Decodes a string. */
118 virtual void decode(const std::string &str, ustring &buf);
120 ustring decode(const std::string &);
122 /** Resets the decoder to the initial state. */
123 virtual void reset() { }
126 /** Handles an error depending on the error mode. The return value is
127 suitable for returning from decode_char.
129 THROW_ON_ERROR: throws CodecError(msg)
130 IGNORE_ERRORS: returns -1
131 TRANSLITERATE: return 0xFFFE */
132 UnicodeChar error(const std::string &msg);
140 /** Returns the name of the encoding handled by this codec. */
141 virtual const char *get_name() const = 0;
143 /** Creates an encoder for this codec. */
144 virtual Encoder *create_encoder(ErrorMode err_mode = THROW_ON_ERROR) const = 0;
146 /** Creates a decoder for this codec. */
147 virtual Decoder *create_decoder(ErrorMode err_mode = THROW_ON_ERROR) const = 0;
149 /** Determines whether the given string can be successfully decoded with
150 this codec. Note that this function returning true does not guarantee that
151 the string was actually encoded with this codec. In particular, many 8-bit
152 encodings are indistinguishable. */
153 virtual bool detect(const std::string &) const;
156 typedef Codec::Encoder Encoder;
157 typedef Codec::Decoder Decoder;
159 /** Convenience function that decodes a string. */
161 ustring decode(const std::string &s)
163 typename C::Decoder dec;
165 dec.decode(s, result);
169 /** Convenience function that encodes a string. */
171 std::string encode(const ustring &s)
173 typename C::Encoder enc;
175 enc.encode(s, result);
180 /** Convenience function that transcodes a string from one codec to another. */
181 template<class F, class T>
182 std::string transcode(const std::string &s)
184 typename F::Decoder from;
185 typename T::Encoder to;
187 from.decode(s, temp);
189 to.encode(temp, result);
194 /** Creates a codec for an encoding by name. The caller is responsible for
195 deleting the codec when it's no longer needed. */
196 Codec *create_codec(const std::string &);
198 /** Automatically detects the encoding of a string and creates a codec for it.
199 The codec must be deleted when it's no longer needed. */
200 Codec *detect_codec(const std::string &);
202 } // namespace Codecs