libbase/utf8.h

   1 // utf8.h: utilities for converting to and from UTF-8
   2 //
   3 //   Copyright (C) 2008, 2009, 2010, 2011, 2012 Free Software Foundation, Inc.
   4 //
   5 // This program is free software; you can redistribute it and/or modify
   6 // it under the terms of the GNU General Public License as published by
   7 // the Free Software Foundation; either version 3 of the License, or
   8 // (at your option) any later version.
   9 //
  10 // This program is distributed in the hope that it will be useful,
  11 // but WITHOUT ANY WARRANTY; without even the implied warranty of
  12 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13 // GNU General Public License for more details.
  14 //
  15 // You should have received a copy of the GNU General Public License
  16 // along with this program; if not, write to the Free Software
  17 // Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
  18 //
  19 // Based on the public domain work of Thatcher Ulrich <tu@tulrich.com> 2004
  20
  21 #ifndef UTF8_H
  22 #define UTF8_H
  23
  24 #include <string>
  25 #include <cstdint> // for C99 int types
  26 #include <vector>
  27
  28 #include "dsodefs.h" // For DSOEXPORT
  29
  30 // Android doesn't have any support for wide characters at all.
  31 #ifdef __ANDROID__
  32 namespace std {
  33 typedef basic_string
  34    <wchar_t
  35    ,std::char_traits<wchar_t>
  36    ,std::allocator<wchar_t> >
  37 wstring;
  38 }
  39 #endif
  40
  41 namespace gnash {
  42
  43 /// Utilities to convert between std::string and std::wstring.
  44 //
  45 /// Strings in Gnash are generally stored as std::strings.
  46 /// We have to deal, however, with characters larger than standard
  47 /// ASCII (128), which can be encoded in two different ways.
  48 ///
  49 /// SWF6 and later use UTF-8, encoded as multibyte characters and
  50 /// allowing many thousands of unique codes. Multibyte characters are
  51 /// difficult to handle, as their length - used for many string
  52 /// operations - is not certain without parsing the string.
  53 /// Converting the string to a wstring (generally a uint32_t - the
  54 /// pp seems only to handle characters up to 65535 - two bytes is
  55 /// the minimum size of a wchar) facilitates string operations, as
  56 /// the length of the string is equal to the number of valid characters.
  57 ///
  58 /// SWF5 and earlier, however, used the ISO-8859 specification,
  59 /// allowing the standard 128 ASCII characters plus 128 extra
  60 /// characters that depend on the particular subset of ISO-8859.
  61 /// Characters are 8 bits, not the ASCII standard 7. SWF5 cannot
  62 /// handle multi-byte characters without special functions.
  63 ///
  64 /// It is important that SWF5 can distinguish between the two encodings,
  65 /// so we cannot convert all strings to UTF-8.
  66 //
  67 /// Please note that, although this is called utf8, what the Adobe
  68 /// player uses is only loosely related to real unicode, so the
  69 /// encoding support here is correspondingly non-standard.
  70 namespace utf8 {
  71
  72     /// Converts a std::string with multibyte characters into a std::wstring.
  73     //
  74     /// @return a version-dependent wstring.
  75     /// @param str the canonical string to convert.
  76     /// @param version the SWF version, used to decide how to decode the string.
  77     //
  78     /// For SWF5, UTF-8 (or any other) multibyte encoded characters are
  79     /// converted char by char, mangling the string.
  80     DSOEXPORT std::wstring decodeCanonicalString(const std::string& str, int version);
  81
  82     /// Converts a std::wstring into canonical std::string.
  83     //
  84     /// @return a version-dependent encoded std::string.
  85     /// @param wstr the wide string to convert.
  86     /// @param version the SWF version, used to decide how to encode the string.
  87     ///
  88     /// For SWF 5, each character is stored as an 8-bit (at least) char, rather
  89     /// than converting it to a canonical UTF-8 byte sequence. Gnash can then
  90     /// distinguish between 8-bit characters, which it handles correctly, and
  91     /// multi-byte characters, which are regarded as multiple characters for
  92     /// string methods.
  93     DSOEXPORT std::string encodeCanonicalString(const std::wstring& wstr, int version);
  94
  95     /// Return the next Unicode character in the UTF-8 encoded string.
  96     //
  97     /// Invalid UTF-8 sequences produce a U+FFFD character
  98     /// as output.  Advances string iterator past the character
  99     /// returned, unless the returned character is '\0', in which
 100     /// case the iterator does not advance.
 101     DSOEXPORT std::uint32_t decodeNextUnicodeCharacter(std::string::const_iterator& it,
 102                                                      const std::string::const_iterator& e);
 103
 104     /// \brief Encodes the given wide character into a canonical
 105     /// string, theoretically up to 6 chars in length.
 106     DSOEXPORT std::string encodeUnicodeCharacter(std::uint32_t ucs_character);
 107
 108     /// Encodes the given wide character into an at least 8-bit character.
 109     //
 110     /// Allows storage of Latin1 (ISO-8859-1) characters. This
 111     /// is the format of SWF5 and below.
 112     DSOEXPORT std::string encodeLatin1Character(std::uint32_t ucsCharacter);
 113
 114     enum TextEncoding {
 115         encUNSPECIFIED,
 116         encUTF8,
 117         encUTF16BE,
 118         encUTF16LE,
 119         encUTF32BE,
 120         encUTF32LE,
 121         encSCSU,
 122         encUTF7,
 123         encUTFEBCDIC,
 124         encBOCU1
 125     };
 126
 127     /// Interpret (and skip) Byte Order Mark in input stream
 128     //
 129     /// This function takes a pointer to a buffer and returns
 130     /// the start of actual data after an eventual BOM.
 131     /// No conversion is performed, no bytes copy, just skipping of
 132     /// the BOM snippet and interpretation of it returned to the
 133     /// encoding input parameter.
 134     ///
 135     /// See http://en.wikipedia.org/wiki/Byte-order_mark
 136     ///
 137     /// @param in
 138     ///    The input buffer.
 139     ///
 140     /// @param size
 141     ///    Size of the input buffer, will be decremented by the
 142     ///    size of the BOM, if any.
 143     ///
 144     /// @param encoding
 145     ///    Output parameter, will always be set.
 146     ///    encUNSPECIFIED if no BOM is found.
 147     ///
 148     /// @returns
 149     ///    A pointer either equal to 'in' or some bytes inside it.
 150     ///
 151     DSOEXPORT const char* stripBOM(const char* in, size_t& size,
 152                                    TextEncoding& encoding);
 153
 154     /// Return name of a text encoding
 155     DSOEXPORT const char* textEncodingName(TextEncoding enc);
 156
 157     enum EncodingGuess {
 158         ENCGUESS_UNICODE = 0,
 159         ENCGUESS_JIS = 1,
 160         ENCGUESS_OTHER = 2
 161     };
 162
 163     /// Common code for guessing at the encoding of random text, between
 164     // Shift-Jis, UTF8, and other. Puts the DisplayObject count in length,
 165     // and the offsets to the DisplayObjects in offsets, if offsets is not NULL.
 166     // If not NULL, offsets should be at least s.length().
 167     // offsets are not accurate if the return value is GUESSENC_OTHER
 168     //
 169     /// TODO: It's doubtful if this even works, and it may not be useful at
 170     /// all.
 171     DSOEXPORT EncodingGuess guessEncoding(const std::string& s, int& length,
 172             std::vector<int>& offsets);
 173
 174
 175 } // namespace utf8
 176 } // namespace gnash
 177
 178 #endif // UTF8_H
 179
 180
 181 // Local Variables:
 182 // mode: C++
 183 // c-basic-offset: 8
 184 // tab-width: 8
 185 // indent-tabs-mode: t
 186 // End: