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 no_commands = 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);
175 /// Return the TIPA shortcut
176 static std::string const TIPAShortcut(char_type c);
178 * Is this a known char from some language?
179 * If \p preamble is empty and code point \p c is known to belong
180 * to a supported script, true is returned and \p preamble is set
181 * to the corresponding entry in the unicodesymbols file.
182 * If \p preamble is not empty, a check is made whether code point
183 * \p c is a known character matching the preamble entry.
185 static bool isKnownScriptChar(char_type const c, std::string & preamble);
187 * Do we have to output this character as LaTeX command in any case?
188 * This is true if the "force" flag is set.
189 * We need this if the inputencoding does not support a certain glyph.
191 static bool isForced(char_type c);
193 * Do we have to display in italics this character when in mathmode?
194 * This is true if the "mathalpha" flag is set. We use this for
195 * letters and accented characters that are output as math commands.
197 static bool isMathAlpha(char_type c);
199 * Do we need to terminate this command (by {} or space)?
200 * This is true if the "notermination" flag is not set.
202 static bool needsTermination(char_type c);
204 * Register \p c as a mathmode command.
206 static void addMathCmd(char_type c) { mathcmd.insert(c); }
208 * Register \p c as a textmode command.
210 static void addTextCmd(char_type c) { textcmd.insert(c); }
212 * Register \p c as a mathmode symbol.
214 static void addMathSym(char_type c) { mathsym.insert(c); }
216 * Tell whether \p c is registered as a mathmode command.
218 static bool isMathCmd(char_type c) { return mathcmd.count(c); }
220 * Tell whether \p c is registered as a textmode command.
222 static bool isTextCmd(char_type c) { return textcmd.count(c); }
224 * Tell whether \p c is registered as a mathmode symbol.
226 static bool isMathSym(char_type c) { return mathsym.count(c); }
228 * Initialize mathcmd, textcmd, and mathsym sets.
230 static void initUnicodeMath(Buffer const & buffer, bool for_master = true);
232 * If \p c cannot be encoded in the given \p encoding, convert
233 * it to something that LaTeX can understand in mathmode.
234 * \return whether \p command is a mathmode command
236 static bool latexMathChar(char_type c, bool mathmode,
237 Encoding const * encoding, docstring & command);
239 * Convert the LaTeX command in \p cmd to the corresponding unicode
240 * point and set \p combining to true if it is a combining symbol
242 static char_type fromLaTeXCommand(docstring const & cmd, int cmdtype,
243 bool & combining, std::set<std::string> * req = 0);
252 * Convert the LaTeX commands in \p cmd and \return a docstring
253 * of corresponding unicode points. The conversion stops at the
254 * first command which could not be converted, and the remaining
255 * unconverted commands are returned in \p rem.
256 * The \p cmdtype parameter can be used to limit recognized
257 * commands to math or text mode commands only.
259 static docstring fromLaTeXCommand(docstring const & cmd, int cmdtype,
260 docstring & rem, std::set<std::string> * req = 0);
262 * Add the preamble snippet needed for the output of \p c to
264 * This does not depend on the used encoding, since the inputenc
265 * package only maps the code point \p c to a command, it does not
266 * make this command available.
268 static void validate(char_type c, LaTeXFeatures & features, bool for_mathed = false);
272 EncodingList encodinglist;
274 static MathCommandSet mathcmd;
276 static TextCommandSet textcmd;
278 static MathSymbolSet mathsym;
281 extern Encodings encodings;