4 * This file is part of LyX, the document processor.
5 * Licence details can be found in the file COPYING.
7 * \author Lars Gullik Bjønnes
8 * \author Jean-Marc Lasgouttes
10 * Full author contact details are available in file CREDITS.
16 #include "support/docstring.h"
17 #include "support/types.h"
25 namespace support
{ class FileName
; }
30 class EncodingException
: public std::exception
{
32 EncodingException(char_type c
);
33 virtual ~EncodingException() throw() {}
34 virtual const char * what() const throw();
36 char_type failed_char
;
45 /// Which LaTeX package handles this encoding?
55 Encoding(std::string
const & n
, std::string
const & l
,
56 std::string
const & g
, std::string
const & i
,
61 std::string
const & name() const { return name_
; }
63 std::string
const & latexName() const { return latexName_
; }
65 std::string
const & guiName() const { return guiName_
; }
67 std::string
const & iconvName() const { return iconvName_
; }
69 bool const & hasFixedWidth() const { return fixedwidth_
; }
71 * Convert \p c to something that LaTeX can understand.
72 * This is either the character itself (if it is representable
73 * in this encoding), or a LaTeX macro.
74 * If the character is not representable in this encoding, but no
75 * LaTeX macro is known, a warning is given of lyxerr, and the
76 * character is returned.
78 docstring
latexChar(char_type c
, bool for_mathed
= false) const;
79 /// Which LaTeX package handles this encoding?
80 Package
package() const { return package_
; }
81 /// A list of all characters usable in this encoding
82 std::vector
<char_type
> symbolsList() const;
87 std::string latexName_
;
91 std::string iconvName_
;
92 /// Is this a fixed width encoding?
95 typedef std::set
<char_type
> CharSet
;
96 /// Set of UCS4 characters that we can encode (for singlebyte
98 mutable CharSet encodable_
;
99 /// All code points below this are encodable. This helps us to avoid
100 /// lokup of ASCII characters in encodable_ and gives about 1 sec
101 /// speedup on export of the Userguide.
102 mutable char_type start_encodable_
;
103 /// Which LaTeX package handles this encoding?
106 * If this is true the stored information about the encoding covers
107 * all encodable characters. We set this to false initially so that
108 * we only need to query iconv for the actually used encodings.
109 * This is needed especially for the multibyte encodings, if we
110 * complete all encoding info on startup it takes 2-3 minutes.
112 mutable bool complete_
;
118 typedef std::set
<char_type
> MathCommandSet
;
120 typedef std::set
<char_type
> TextCommandSet
;
122 typedef std::set
<char_type
> MathSymbolSet
;
124 typedef std::map
<std::string
, Encoding
> EncodingList
;
125 /// iterator to iterate over all encodings.
126 /// We hide the fact that our encoding list is implemented as a map.
127 class const_iterator
: public EncodingList::const_iterator
{
128 typedef EncodingList::const_iterator base
;
130 const_iterator() : base() {}
131 const_iterator(base
const & b
) : base(b
) {}
132 Encoding
const & operator*() const { return base::operator*().second
; }
133 Encoding
const * operator->() const { return &(base::operator*().second
); }
137 /// Read the encodings.
138 /// \param encfile encodings definition file
139 /// \param symbolsfile unicode->LaTeX mapping file
140 void read(support::FileName
const & encfile
,
141 support::FileName
const & symbolsfile
);
142 /// Get encoding from LyX name \p name
143 Encoding
const * fromLyXName(std::string
const & name
) const;
144 /// Get encoding from LaTeX name \p name
145 Encoding
const * fromLaTeXName(std::string
const & name
) const;
148 const_iterator
begin() const { return encodinglist
.begin(); }
150 const_iterator
end() const { return encodinglist
.end(); }
164 static bool isHebrewComposeChar(char_type c
);
166 static bool isArabicComposeChar(char_type c
);
168 static bool isArabicSpecialChar(char_type c
);
170 static bool isArabicChar(char_type c
);
172 static char_type
transformChar(char_type c
, LetterForm form
);
173 /// Is this a combining char?
174 static bool isCombiningChar(char_type c
);
176 * Is this a known char from some language?
177 * If \p preamble is empty and code point \p c is known to belong
178 * to a supported script, true is returned and \p preamble is set
179 * to the corresponding entry in the unicodesymbols file.
180 * If \p preamble is not empty, a check is made whether code point
181 * \p c is a known character matching the preamble entry.
183 static bool isKnownScriptChar(char_type
const c
, std::string
& preamble
);
185 * Do we have to output this character as LaTeX command in any case?
186 * This is true if the "force" flag is set.
187 * We need this if the inputencoding does not support a certain glyph.
189 static bool isForced(char_type c
);
191 * Do we have to display in italics this character when in mathmode?
192 * This is true if the "mathalpha" flag is set. We use this for
193 * letters and accented characters that are output as math commands.
195 static bool isMathAlpha(char_type c
);
197 * Register \p c as a mathmode command.
199 static void addMathCmd(char_type c
) { mathcmd
.insert(c
); }
201 * Register \p c as a textmode command.
203 static void addTextCmd(char_type c
) { textcmd
.insert(c
); }
205 * Register \p c as a mathmode symbol.
207 static void addMathSym(char_type c
) { mathsym
.insert(c
); }
209 * Tell whether \p c is registered as a mathmode command.
211 static bool isMathCmd(char_type c
) { return mathcmd
.count(c
); }
213 * Tell whether \p c is registered as a textmode command.
215 static bool isTextCmd(char_type c
) { return textcmd
.count(c
); }
217 * Tell whether \p c is registered as a mathmode symbol.
219 static bool isMathSym(char_type c
) { return mathsym
.count(c
); }
221 * Initialize mathcmd, textcmd, and mathsym sets.
223 static void initUnicodeMath(Buffer
const & buffer
, bool clear_sets
= true);
225 * If \p c cannot be encoded in the given \p encoding, convert
226 * it to something that LaTeX can understand in mathmode.
227 * \return whether \p command is a mathmode command
229 static bool latexMathChar(char_type c
, bool mathmode
,
230 Encoding
const * encoding
, docstring
& command
);
232 * Convert the LaTeX command in \p cmd to the corresponding unicode
233 * point and set \p combining to true if it is a combining symbol
235 static char_type
fromLaTeXCommand(docstring
const & cmd
, bool & combining
);
244 * Convert the LaTeX commands in \p cmd and \return a docstring
245 * of corresponding unicode points. The conversion stops at the
246 * first command which could not be converted, and the remaining
247 * unconverted commands are returned in \p rem.
248 * The \p cmdtype parameter can be used to limit recognized
249 * commands to math or text mode commands only.
251 static docstring
fromLaTeXCommand(docstring
const & cmd
,
252 docstring
& rem
, int cmdtype
= MATH_CMD
| TEXT_CMD
);
254 * Add the preamble snippet needed for the output of \p c to
256 * This does not depend on the used encoding, since the inputenc
257 * package only maps the code point \p c to a command, it does not
258 * make this command available.
260 static void validate(char_type c
, LaTeXFeatures
& features
, bool for_mathed
= false);
264 EncodingList encodinglist
;
266 static MathCommandSet mathcmd
;
268 static TextCommandSet textcmd
;
270 static MathSymbolSet mathsym
;
273 extern Encodings encodings
;