X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;ds=sidebyside;f=src%2FEncoding.h;h=0f6e8ff4386b8b4b6f269f3ecd815cd96ce49f1a;hb=28be7d552f62cc02fa86d7f79201d089bfb2d7b5;hp=8537a221a6ec9e7f9fbbc093a30624e5c723e010;hpb=6b28f08ff2f15fee2697fee4395225d6d9ec8daf;p=lyx.git diff --git a/src/Encoding.h b/src/Encoding.h index 8537a221a6..0f6e8ff438 100644 --- a/src/Encoding.h +++ b/src/Encoding.h @@ -4,7 +4,7 @@ * This file is part of LyX, the document processor. * Licence details can be found in the file COPYING. * - * \author Lars Gullik Bjønnes + * \author Lars Gullik Bjønnes * \author Jean-Marc Lasgouttes * * Full author contact details are available in file CREDITS. @@ -14,6 +14,7 @@ #define ENCODING_H #include "support/docstring.h" +#include "support/trivstring.h" #include "support/types.h" #include @@ -24,42 +25,133 @@ namespace lyx { namespace support { class FileName; } -class LaTeXFeatures; - class EncodingException : public std::exception { public: EncodingException(char_type c); virtual ~EncodingException() throw() {} virtual const char * what() const throw(); - + char_type failed_char; int par_id; pos_type pos; }; -/// +enum CharInfoFlags { + /// + CharInfoCombining = 1, + /// + CharInfoTextFeature = 2, + /// + CharInfoMathFeature = 4, + /// + CharInfoForce = 8, + /// + CharInfoTextNoTermination = 16, + /// + CharInfoMathNoTermination = 32, + /// + CharInfoForceSelected = 64, + /// + CharInfoDeprecated = 128 +}; + + +/// Information about a single UCS4 character +class CharInfo { +public: + CharInfo() : flags_(0) {} + CharInfo( + docstring const & textcommand, docstring const & mathcommand, + std::string const & textpreamble, std::string const & mathpreamble, + std::string const & tipashortcut, unsigned int flags); + // we assume that at least one command is nonempty when using unicodesymbols + bool isUnicodeSymbol() const { return !textcommand_.empty() || !mathcommand_.empty(); } + /// LaTeX command (text mode) for this character + docstring const textcommand() const { return textcommand_; } + /// LaTeX command (math mode) for this character + docstring mathcommand() const { return mathcommand_; } + /// Needed LaTeX preamble (or feature) for text mode + std::string textpreamble() const { return textpreamble_; } + /// Needed LaTeX preamble (or feature) for math mode + std::string mathpreamble() const { return mathpreamble_; } + /// Is this a combining character? + bool combining() const { return flags_ & CharInfoCombining ? true : false; } + /// Is \c textpreamble a feature known by LaTeXFeatures, or a raw LaTeX + /// command? + bool textfeature() const { return flags_ & CharInfoTextFeature ? true : false; } + /// Is \c mathpreamble a feature known by LaTeXFeatures, or a raw LaTeX + /// command? + bool mathfeature() const { return flags_ & CharInfoMathFeature ? true : false; } + /// Always force the LaTeX command, even if the encoding contains + /// this character? + bool force() const { return flags_ & CharInfoForce ? true : false; } + /// Force the LaTeX command for some encodings? + bool forceselected() const { return flags_ & CharInfoForceSelected ? true : false; } + /// Disable LaTeX command => char_type conversion for this deprecated symbol? + bool deprecated() const { return flags_ & CharInfoDeprecated ? true : false; } + /// TIPA shortcut + std::string const tipashortcut() const { return tipashortcut_; } + /// \c textcommand needs no termination (such as {} or space). + bool textnotermination() const { return flags_ & CharInfoTextNoTermination ? true : false; } + /// \c mathcommand needs no termination (such as {} or space). + bool mathnotermination() const { return flags_ & CharInfoMathNoTermination ? true : false; } + /// +private: + /// LaTeX command (text mode) for this character + trivdocstring textcommand_; + /// LaTeX command (math mode) for this character + trivdocstring mathcommand_; + /// Needed LaTeX preamble (or feature) for text mode + trivstring textpreamble_; + /// Needed LaTeX preamble (or feature) for math mode + trivstring mathpreamble_; + /// TIPA shortcut + trivstring tipashortcut_; + /// feature flags + unsigned int flags_; +}; + + +/** + * An encoding as defined in lib/encodings. + * All const methods are thread-safe, so the caller does not need any locking. + * This property must be kept when changing the class. + */ class Encoding { public: /// Which LaTeX package handles this encoding? enum Package { - none, - inputenc, - CJK + none = 1, + inputenc = 2, + CJK = 4, + japanese = 8 }; + /// Represent any of the above packages + static int const any; /// - Encoding() {} + Encoding() : fixedwidth_(true), unsafe_(false), forced_(0), + start_encodable_(0), package_(none), complete_(false) {} /// Encoding(std::string const & n, std::string const & l, - std::string const & i, bool f, Package p); + std::string const & g, std::string const & i, + bool f, bool u, Package p); /// void init() const; /// - std::string const & name() const { return name_; } + std::string const name() const { return name_; } + /// + std::string const latexName() const { return latexName_; } + /// + std::string const guiName() const { return guiName_; } + /// + std::string const iconvName() const { return iconvName_; } /// - std::string const & latexName() const { return latexName_; } + bool hasFixedWidth() const { return fixedwidth_; } /// - std::string const & iconvName() const { return iconvName_; } + bool unsafe() const { return unsafe_; } + /// \p c is representable in this encoding without a LaTeX macro + bool encodable(char_type c) const; /** * Convert \p c to something that LaTeX can understand. * This is either the character itself (if it is representable @@ -67,30 +159,58 @@ public: * If the character is not representable in this encoding, but no * LaTeX macro is known, a warning is given of lyxerr, and the * character is returned. + * \return the converted character and a flag indicating whether + * the command needs to be terminated by {} or a space. + */ + std::pair latexChar(char_type c) const; + /** + * Convert \p input to something that LaTeX can understand. + * This is either the string itself (if it is representable + * in this encoding), or a LaTeX macro. + * If a character is not representable in this encoding, but no + * LaTeX macro is known, a warning is given of lyxerr, and the + * character is returned in the second string of the pair and + * omitted in the first. + * \p dryrun specifies whether the string is used within source + * preview (which yields a special warning). */ - docstring latexChar(char_type c) const; + std::pair latexString(docstring const & input, + bool dryrun = false) const; /// Which LaTeX package handles this encoding? Package package() const { return package_; } /// A list of all characters usable in this encoding std::vector symbolsList() const; private: + /** + * Do we have to output this character as LaTeX command in any case? + * This is true if the "force" flag is set. + * We need this if the inputencoding does not support a certain glyph. + */ + bool isForced(char_type c) const; + /// + trivstring name_; /// - std::string name_; + trivstring latexName_; /// - std::string latexName_; + trivstring guiName_; /// - std::string iconvName_; + trivstring iconvName_; /// Is this a fixed width encoding? bool fixedwidth_; + /// Is this encoding TeX unsafe, e.g. control characters like {, } + /// and \\ may appear in high bytes? + bool unsafe_; /// typedef std::set CharSet; /// Set of UCS4 characters that we can encode (for singlebyte /// encodings only) - mutable CharSet encodable_; + CharSet encodable_; + /// Set of UCS4 characters that we can't encode + CharSet const * forced_; /// All code points below this are encodable. This helps us to avoid /// lokup of ASCII characters in encodable_ and gives about 1 sec /// speedup on export of the Userguide. - mutable char_type start_encodable_; + char_type start_encodable_; /// Which LaTeX package handles this encoding? Package package_; /** @@ -100,13 +220,19 @@ private: * This is needed especially for the multibyte encodings, if we * complete all encoding info on startup it takes 2-3 minutes. */ - mutable bool complete_; + bool complete_; }; class Encodings { public: /// - typedef std::map EncodingList; + typedef std::set MathCommandSet; + /// + typedef std::set TextCommandSet; + /// + typedef std::set MathSymbolSet; + /// + typedef std::map EncodingList; /// iterator to iterate over all encodings. /// We hide the fact that our encoding list is implemented as a map. class const_iterator : public EncodingList::const_iterator { @@ -125,38 +251,26 @@ public: void read(support::FileName const & encfile, support::FileName const & symbolsfile); /// Get encoding from LyX name \p name - Encoding const * fromLyXName(std::string const & name) const; - /// Get encoding from LaTeX name \p name - Encoding const * fromLaTeXName(std::string const & name) const; + Encoding const * + fromLyXName(std::string const & name, bool allowUnsafe = false) const; + /// Get encoding from LaTeX name \p name and package \p package + Encoding const * fromLaTeXName(std::string const & name, + int const & package = Encoding::any, bool allowUnsafe = false) const; + /// Get encoding from iconv name \p name and package \p package + Encoding const * fromIconvName(std::string const & name, + int const & package = Encoding::any, bool allowUnsafe = false) const; /// const_iterator begin() const { return encodinglist.begin(); } /// const_iterator end() const { return encodinglist.end(); } - /// - enum LetterForm { - /// - FORM_ISOLATED, - /// - FORM_FINAL, - /// - FORM_INITIAL, - /// - FORM_MEDIAL - }; - /// - static bool isHebrewComposeChar(char_type c); - /// - static bool isArabicComposeChar(char_type c); - /// - static bool isArabicSpecialChar(char_type c); - /// - static bool isArabicChar(char_type c); - /// - static char_type transformChar(char_type c, LetterForm form); + /// Accessor for the unicode information table. + static CharInfo const & unicodeCharInfo(char_type c); /// Is this a combining char? static bool isCombiningChar(char_type c); + /// Return the TIPA shortcut + static std::string const TIPAShortcut(char_type c); /** * Is this a known char from some language? * If \p preamble is empty and code point \p c is known to belong @@ -167,17 +281,90 @@ public: */ static bool isKnownScriptChar(char_type const c, std::string & preamble); /** - * Add the preamble snippet needed for the output of \p c to - * \p features. - * This does not depend on the used encoding, since the inputenc - * package only maps the code point \p c to a command, it does not - * make this command available. + * Do we have to display in italics this character when in mathmode? + * This is true if the "mathalpha" flag is set. We use this for + * letters and accented characters that are output as math commands. + */ + static bool isMathAlpha(char_type c); + /** + * Do we have to wrap in \text this character when in mathmode? + * This is true if \p c is not ascii and the "mathalpha" flag is not + * set and a mathcommand is not defined in the unicodesymbols file. + */ + static bool isUnicodeTextOnly(char_type c); + /** + * Register \p c as a mathmode command. + */ + static void addMathCmd(char_type c) { mathcmd.insert(c); } + /** + * Register \p c as a textmode command. + */ + static void addTextCmd(char_type c) { textcmd.insert(c); } + /** + * Register \p c as a mathmode symbol. + */ + static void addMathSym(char_type c) { mathsym.insert(c); } + /** + * Tell whether \p c is registered as a mathmode command. */ - static void validate(char_type c, LaTeXFeatures & features); + static bool isMathCmd(char_type c) { return mathcmd.count(c); } + /** + * Tell whether \p c is registered as a textmode command. + */ + static bool isTextCmd(char_type c) { return textcmd.count(c); } + /** + * Tell whether \p c is registered as a mathmode symbol. + */ + static bool isMathSym(char_type c) { return mathsym.count(c); } + /** + * If \p c cannot be encoded in the given \p encoding, convert + * it to something that LaTeX can understand in mathmode. + * \p needsTermination indicates whether the command needs to be + * terminated by {} or a space. + * \return whether \p command is a mathmode command + */ + static bool latexMathChar(char_type c, bool mathmode, + Encoding const * encoding, docstring & command, + bool & needsTermination); + /** + * Convert the LaTeX command in \p cmd to the corresponding unicode + * point and set \p combining to true if it is a combining symbol. + * \p needsTermination indicates whether the command needs to be + * terminated by {} or a space. + */ + static char_type fromLaTeXCommand(docstring const & cmd, int cmdtype, + bool & combining, bool & needsTermination, + std::set * req = 0); + /// + enum LatexCmd { + /// + MATH_CMD = 1, + /// + TEXT_CMD = 2 + }; + /** + * Convert the LaTeX commands in \p cmd and \return a docstring + * of corresponding unicode points. The conversion stops at the + * first command which could not be converted, and the remaining + * unconverted commands are returned in \p rem. + * The \p cmdtype parameter can be used to limit recognized + * commands to math or text mode commands only. + * \p needsTermination indicates whether the command needs to be + * terminated by {} or a space. + */ + static docstring fromLaTeXCommand(docstring const & cmd, int cmdtype, + bool & needsTermination, docstring & rem, + std::set * req = 0); -private: +protected: /// EncodingList encodinglist; + /// + static MathCommandSet mathcmd; + /// + static TextCommandSet textcmd; + /// + static MathSymbolSet mathsym; }; extern Encodings encodings;