4 * This file is part of LyX, the document processor.
5 * Licence details can be found in the file COPYING.
9 * Full author contact details are available in file CREDITS.
19 #include "support/docstream.h"
24 enum mode_type {UNDECIDED_MODE, TEXT_MODE, MATH_MODE, MATHTEXT_MODE, TABLE_MODE};
26 mode_type asMode(mode_type oldmode, std::string const & str);
29 // These are TeX's catcodes
31 catEscape, // 0 backslash
42 catLetter, // 11 a-zA-Z
43 catOther, // 12 none of the above
46 catInvalid // 15 <delete>
51 FLAG_BRACE_LAST = 1 << 1, // last closing brace ends the parsing
52 FLAG_RIGHT = 1 << 2, // next \\right ends the parsing process
53 FLAG_END = 1 << 3, // next \\end ends the parsing process
54 FLAG_BRACK_LAST = 1 << 4, // next closing bracket ends the parsing
55 FLAG_TEXTMODE = 1 << 5, // we are in a box
56 FLAG_ITEM = 1 << 6, // read a (possibly braced token)
57 FLAG_LEAVE = 1 << 7, // leave the loop at the end
58 FLAG_SIMPLE = 1 << 8, // next $ leaves the loop
59 FLAG_EQUATION = 1 << 9, // next \] leaves the loop
60 FLAG_SIMPLE2 = 1 << 10, // next \) leaves the loop
61 FLAG_OPTION = 1 << 11, // read [...] style option
62 FLAG_BRACED = 1 << 12, // read {...} style argument
63 FLAG_CELL = 1 << 13, // read table cell
64 FLAG_TABBING = 1 << 14 // We are inside a tabbing environment
70 // Helper class for parsing
76 Token() : cs_(), cat_(catIgnore) {}
78 Token(docstring const & cs, CatCode cat) : cs_(to_utf8(cs)), cat_(cat) {}
80 /// Returns the token as string
81 std::string const & cs() const { return cs_; }
82 /// Returns the catcode of the token
83 CatCode cat() const { return cat_; }
85 char character() const { return cs_.empty() ? 0 : cs_[0]; }
86 /// Returns the token verbatim
87 std::string asInput() const;
88 /// Is the token an alphanumerical character?
89 bool isAlnumASCII() const;
98 std::ostream & operator<<(std::ostream & os, Token const & t);
102 * Actual parser class
104 * The parser parses every character of the inputstream into a token
105 * and classifies the token.
106 * The following transformations are done:
107 * - Consecutive spaces are combined into one single token with CatCode catSpace
108 * - Consecutive newlines are combined into one single token with CatCode catNewline
109 * - Comments and %\n combinations are parsed into one token with CatCode catComment
116 Parser(idocstream & is);
118 Parser(std::string const & s);
122 /// change the latex encoding of the input stream
123 void setEncoding(std::string const & encoding);
124 /// get the current latex encoding of the input stream
125 std::string getEncoding() const { return encoding_latex_; }
128 int lineno() const { return lineno_; }
131 /// dump contents to screen
135 typedef std::pair<bool, std::string> Arg;
137 * Get an argument enclosed by \p left and \p right.
138 * \returns wether an argument was found in \p Arg.first and the
139 * argument in \p Arg.second. \see getArg().
141 Arg getFullArg(char left, char right);
143 * Get an argument enclosed by \p left and \p right.
144 * \returns the argument (without \p left and \p right) or the empty
145 * string if the next non-space token is not \p left. Use
146 * getFullArg() if you need to know wether there was an empty
147 * argument or no argument at all.
149 std::string getArg(char left, char right);
151 * \returns getFullArg('[', ']') including the brackets or the
152 * empty string if there is no such argument.
154 std::string getFullOpt();
156 * \returns getArg('[', ']') including the brackets or the
157 * empty string if there is no such argument.
158 * No whitespace is eaten if \p keepws is true and no optional
159 * argument exists. This is important if an optional argument is
160 * parsed that would go after a command in ERT: In this case the
161 * whitespace is needed to separate the ERT from the subsequent
162 * word. Without it, the ERT and the next word would be concatenated
163 * during .tex export, thus creating an invalid command.
165 std::string getOpt(bool keepws = false);
167 * the same as getOpt but without the brackets
169 std::string getOptContent();
171 * \returns getFullArg('(', ')') including the parentheses or the
172 * empty string if there is no such argument.
174 std::string getFullParentheseArg();
176 * \returns the contents of the environment \p name.
177 * <tt>\begin{name}</tt> must be parsed already, <tt>\end{name}</tt>
178 * is parsed but not returned.
180 std::string const verbatimEnvironment(std::string const & name);
182 * Returns the character of the current token and increments
183 * the token position.
187 void error(std::string const & msg);
188 /// Parses one token from \p is
191 void push_back(Token const & t);
192 /// The previous token.
193 Token const prev_token() const;
194 /// The current token.
195 Token const curr_token() const;
197 Token const next_token();
198 /// Make the next token current and return that.
199 Token const get_token();
200 /// \return whether the current token starts a new paragraph
202 /// skips spaces (and comments if \p skip_comments is true)
203 /// \return whether whitespace was skipped (not comments)
204 bool skip_spaces(bool skip_comments = false);
205 /// puts back spaces (and comments if \p skip_comments is true)
206 void unskip_spaces(bool skip_comments = false);
208 void lex(std::string const & s);
212 std::string verbatim_item();
214 std::string verbatimOption();
215 /// resets the parser to initial state
218 void setCatCode(char c, CatCode cat);
220 CatCode getCatCode(char c) const;
226 std::vector<Token> tokens_;
230 idocstringstream * iss_;
233 /// latex name of the current encoding
234 std::string encoding_latex_;