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,
57 bool f, bool u, Package p);
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 hasFixedWidth() const { return fixedwidth_; }
71 bool unsafe() const { return unsafe_; }
72 /// \p c is representable in this encoding without a LaTeX macro
73 bool encodable(char_type c) const;
75 * Convert \p c to something that LaTeX can understand.
76 * This is either the character itself (if it is representable
77 * in this encoding), or a LaTeX macro.
78 * If the character is not representable in this encoding, but no
79 * LaTeX macro is known, a warning is given of lyxerr, and the
80 * character is returned.
81 * \return the converted character and a flag indicating whether
82 * the command needs to be terminated by {} or a space.
84 std::pair<docstring, bool> latexChar(char_type c) const;
86 * Convert \p input to something that LaTeX can understand.
87 * This is either the string itself (if it is representable
88 * in this encoding), or a LaTeX macro.
89 * If a character is not representable in this encoding, but no
90 * LaTeX macro is known, a warning is given of lyxerr, and the
91 * character is returned in the second string of the pair and
92 * omitted in the first.
93 * \p dryrun specifies whether the string is used within source
94 * preview (which yields a special warning).
96 std::pair<docstring, docstring> latexString(docstring const input,
97 bool dryrun = false) const;
98 /// Which LaTeX package handles this encoding?
99 Package package() const { return package_; }
100 /// A list of all characters usable in this encoding
101 std::vector<char_type> symbolsList() const;
106 std::string latexName_;
108 std::string guiName_;
110 std::string iconvName_;
111 /// Is this a fixed width encoding?
113 /// Is this encoding TeX unsafe, e.g. control characters like {, }
114 /// and \\ may appear in high bytes?
117 typedef std::set<char_type> CharSet;
118 /// Set of UCS4 characters that we can encode (for singlebyte
120 mutable CharSet encodable_;
121 /// All code points below this are encodable. This helps us to avoid
122 /// lokup of ASCII characters in encodable_ and gives about 1 sec
123 /// speedup on export of the Userguide.
124 mutable char_type start_encodable_;
125 /// Which LaTeX package handles this encoding?
128 * If this is true the stored information about the encoding covers
129 * all encodable characters. We set this to false initially so that
130 * we only need to query iconv for the actually used encodings.
131 * This is needed especially for the multibyte encodings, if we
132 * complete all encoding info on startup it takes 2-3 minutes.
134 mutable bool complete_;
140 typedef std::set<char_type> MathCommandSet;
142 typedef std::set<char_type> TextCommandSet;
144 typedef std::set<char_type> MathSymbolSet;
146 typedef std::map<std::string, Encoding> EncodingList;
147 /// iterator to iterate over all encodings.
148 /// We hide the fact that our encoding list is implemented as a map.
149 class const_iterator : public EncodingList::const_iterator {
150 typedef EncodingList::const_iterator base;
152 const_iterator() : base() {}
153 const_iterator(base const & b) : base(b) {}
154 Encoding const & operator*() const { return base::operator*().second; }
155 Encoding const * operator->() const { return &(base::operator*().second); }
159 /// Read the encodings.
160 /// \param encfile encodings definition file
161 /// \param symbolsfile unicode->LaTeX mapping file
162 void read(support::FileName const & encfile,
163 support::FileName const & symbolsfile);
164 /// Get encoding from LyX name \p name
166 fromLyXName(std::string const & name, bool allowUnsafe = false) const;
167 /// Get encoding from LaTeX name \p name
169 fromLaTeXName(std::string const & name, bool allowUnsafe = false) const;
172 const_iterator begin() const { return encodinglist.begin(); }
174 const_iterator end() const { return encodinglist.end(); }
188 static bool isHebrewComposeChar(char_type c);
190 static bool isArabicComposeChar(char_type c);
192 static bool isArabicSpecialChar(char_type c);
194 static bool isArabicChar(char_type c);
196 static char_type transformChar(char_type c, LetterForm form);
197 /// Is this a combining char?
198 static bool isCombiningChar(char_type c);
199 /// Return the TIPA shortcut
200 static std::string const TIPAShortcut(char_type c);
202 * Is this a known char from some language?
203 * If \p preamble is empty and code point \p c is known to belong
204 * to a supported script, true is returned and \p preamble is set
205 * to the corresponding entry in the unicodesymbols file.
206 * If \p preamble is not empty, a check is made whether code point
207 * \p c is a known character matching the preamble entry.
209 static bool isKnownScriptChar(char_type const c, std::string & preamble);
211 * Do we have to output this character as LaTeX command in any case?
212 * This is true if the "force" flag is set.
213 * We need this if the inputencoding does not support a certain glyph.
215 static bool isForced(char_type c);
217 * Do we have to display in italics this character when in mathmode?
218 * This is true if the "mathalpha" flag is set. We use this for
219 * letters and accented characters that are output as math commands.
221 static bool isMathAlpha(char_type c);
223 * Register \p c as a mathmode command.
225 static void addMathCmd(char_type c) { mathcmd.insert(c); }
227 * Register \p c as a textmode command.
229 static void addTextCmd(char_type c) { textcmd.insert(c); }
231 * Register \p c as a mathmode symbol.
233 static void addMathSym(char_type c) { mathsym.insert(c); }
235 * Tell whether \p c is registered as a mathmode command.
237 static bool isMathCmd(char_type c) { return mathcmd.count(c); }
239 * Tell whether \p c is registered as a textmode command.
241 static bool isTextCmd(char_type c) { return textcmd.count(c); }
243 * Tell whether \p c is registered as a mathmode symbol.
245 static bool isMathSym(char_type c) { return mathsym.count(c); }
247 * Initialize mathcmd, textcmd, and mathsym sets.
249 static void initUnicodeMath(Buffer const & buffer, bool for_master = true);
251 * If \p c cannot be encoded in the given \p encoding, convert
252 * it to something that LaTeX can understand in mathmode.
253 * \p needsTermination indicates whether the command needs to be
254 * terminated by {} or a space.
255 * \return whether \p command is a mathmode command
257 static bool latexMathChar(char_type c, bool mathmode,
258 Encoding const * encoding, docstring & command,
259 bool & needsTermination);
261 * Convert the LaTeX command in \p cmd to the corresponding unicode
262 * point and set \p combining to true if it is a combining symbol.
263 * \p needsTermination indicates whether the command needs to be
264 * terminated by {} or a space.
266 static char_type fromLaTeXCommand(docstring const & cmd, int cmdtype,
267 bool & combining, bool & needsTermination,
268 std::set<std::string> * req = 0);
277 * Convert the LaTeX commands in \p cmd and \return a docstring
278 * of corresponding unicode points. The conversion stops at the
279 * first command which could not be converted, and the remaining
280 * unconverted commands are returned in \p rem.
281 * The \p cmdtype parameter can be used to limit recognized
282 * commands to math or text mode commands only.
283 * \p needsTermination indicates whether the command needs to be
284 * terminated by {} or a space.
286 static docstring fromLaTeXCommand(docstring const & cmd, int cmdtype,
287 bool & needsTermination, docstring & rem,
288 std::set<std::string> * req = 0);
290 * Add the preamble snippet needed for the output of \p c to
292 * This does not depend on the used encoding, since the inputenc
293 * package only maps the code point \p c to a command, it does not
294 * make this command available.
296 static void validate(char_type c, LaTeXFeatures & features, bool for_mathed = false);
300 EncodingList encodinglist;
302 static MathCommandSet mathcmd;
304 static TextCommandSet textcmd;
306 static MathSymbolSet mathsym;
309 extern Encodings encodings;