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; }
29 class EncodingException : public std::exception {
31 EncodingException(char_type c);
32 virtual ~EncodingException() throw() {}
33 virtual const char * what() const throw();
35 char_type failed_char;
44 /// Which LaTeX package handles this encoding?
54 Encoding(std::string const & n, std::string const & l,
55 std::string const & g, std::string const & i,
60 std::string const & name() const { return name_; }
62 std::string const & latexName() const { return latexName_; }
64 std::string const & guiName() const { return guiName_; }
66 std::string const & iconvName() const { return iconvName_; }
68 bool const & hasFixedWidth() const { return fixedwidth_; }
70 * Convert \p c to something that LaTeX can understand.
71 * This is either the character itself (if it is representable
72 * in this encoding), or a LaTeX macro.
73 * If the character is not representable in this encoding, but no
74 * LaTeX macro is known, a warning is given of lyxerr, and the
75 * character is returned.
77 docstring latexChar(char_type c, bool for_mathed = false) const;
78 /// Which LaTeX package handles this encoding?
79 Package package() const { return package_; }
80 /// A list of all characters usable in this encoding
81 std::vector<char_type> symbolsList() const;
86 std::string latexName_;
90 std::string iconvName_;
91 /// Is this a fixed width encoding?
94 typedef std::set<char_type> CharSet;
95 /// Set of UCS4 characters that we can encode (for singlebyte
97 mutable CharSet encodable_;
98 /// All code points below this are encodable. This helps us to avoid
99 /// lokup of ASCII characters in encodable_ and gives about 1 sec
100 /// speedup on export of the Userguide.
101 mutable char_type start_encodable_;
102 /// Which LaTeX package handles this encoding?
105 * If this is true the stored information about the encoding covers
106 * all encodable characters. We set this to false initially so that
107 * we only need to query iconv for the actually used encodings.
108 * This is needed especially for the multibyte encodings, if we
109 * complete all encoding info on startup it takes 2-3 minutes.
111 mutable bool complete_;
117 typedef std::map<std::string, Encoding> EncodingList;
118 /// iterator to iterate over all encodings.
119 /// We hide the fact that our encoding list is implemented as a map.
120 class const_iterator : public EncodingList::const_iterator {
121 typedef EncodingList::const_iterator base;
123 const_iterator() : base() {}
124 const_iterator(base const & b) : base(b) {}
125 Encoding const & operator*() const { return base::operator*().second; }
126 Encoding const * operator->() const { return &(base::operator*().second); }
130 /// Read the encodings.
131 /// \param encfile encodings definition file
132 /// \param symbolsfile unicode->LaTeX mapping file
133 void read(support::FileName const & encfile,
134 support::FileName const & symbolsfile);
135 /// Get encoding from LyX name \p name
136 Encoding const * fromLyXName(std::string const & name) const;
137 /// Get encoding from LaTeX name \p name
138 Encoding const * fromLaTeXName(std::string const & name) const;
141 const_iterator begin() const { return encodinglist.begin(); }
143 const_iterator end() const { return encodinglist.end(); }
157 static bool isHebrewComposeChar(char_type c);
159 static bool isArabicComposeChar(char_type c);
161 static bool isArabicSpecialChar(char_type c);
163 static bool isArabicChar(char_type c);
165 static char_type transformChar(char_type c, LetterForm form);
166 /// Is this a combining char?
167 static bool isCombiningChar(char_type c);
169 * Is this a known char from some language?
170 * If \p preamble is empty and code point \p c is known to belong
171 * to a supported script, true is returned and \p preamble is set
172 * to the corresponding entry in the unicodesymbols file.
173 * If \p preamble is not empty, a check is made whether code point
174 * \p c is a known character matching the preamble entry.
176 static bool isKnownScriptChar(char_type const c, std::string & preamble);
178 * Do we have to output this character as LaTeX command in any case?
179 * This is true if the "forced" flag is set.
180 * We need this if the inputencoding does not support a certain glyph.
182 static bool isForced(char_type c);
184 * If \p c cannot be encoded in the given \p encoding, convert
185 * it to something that LaTeX can understand in math mode.
186 * \return whether \p command is a math mode command
188 static bool latexMathChar(char_type c, Encoding const * encoding, docstring & command);
191 * Convert the LaTeX command in \p cmd to the corresponding unicode
192 * point and set \p combining to true if it is a combining symbol
194 static char_type fromLaTeXCommand(docstring const & cmd, bool & combining);
196 * Convert the LaTeX commands in \p cmd and \return a docstring
197 * of corresponding unicode points. The conversion stops at the
198 * first command which could not be converted, and the remaining
199 * unconverted commands are returned in \p rem
201 static docstring fromLaTeXCommand(docstring const & cmd, docstring & rem);
203 * Add the preamble snippet needed for the output of \p c to
205 * This does not depend on the used encoding, since the inputenc
206 * package only maps the code point \p c to a command, it does not
207 * make this command available.
209 static void validate(char_type c, LaTeXFeatures & features, bool for_mathed = false);
213 EncodingList encodinglist;
216 extern Encodings encodings;