So during my thesis I used xppaut quite extensively. In the end there were some minor problems and I decided to have a look at the code, which was just released at github.
So after 30 years of on and off development by Bard Ermentrout, it kind of showed some rough edges github link to xppaut, which I tried to smooth github link to cleaned version.
So now there were two projects remaining. First automation via the ode files and secondly a new GUI. The first idea required fiddling with the parser which was quite involved. As the GUI work with QT required switching to C++ anyway I decided to write a new parser, that simplifies the code and makes it more manageable.
So here is the parser. The definition of the corresponding ODE file type is given here ODE doc file.
The Idea was to first parse the file and remove comments/empty lines. Then handle all "special" objects and finish with those things defined by "key name = expression". The information would be stored in an opts structure, that contains the name, the expression and possible arguments as well as the line in the original ode file.
I also added some exception handling to give the user a better feedback. The full code in a compilable setting can be found here githublink
xppParserException.h
#ifndef XPPPARSEREXCEPTION_H
#define XPPPARSEREXCEPTION_H
#include <exception>
#include <string>
#include <utility>
#include "xppParserDefines.h"
/* Different parser errors */
enum xppParserError {
DUPLICATED_NAME,
MISSING_ARGUMENT,
MISSING_CLOSING_BRACKET,
MISSING_MARKOV_ASSIGNMENT,
MISSING_OPENING_BRACKET,
EXPECTED_LIST_ARGUMENT,
EXPECTED_NUMBER,
EXPECTED_TABLE_ASSIGNMENT,
RESERVED_FUNCTION,
RESERVED_KEYWORD,
RESERVED_OPTION,
UNKNOWN_ASSIGNMENT,
UNKNOWN_FUNCTION,
UNKNOWN_NAME,
UNKNOWN_OPTION,
UNKNOWN_VARIABLE,
WRONG_ARRAY_ASSIGNMENT,
WRONG_MARKOV_ASSIGNMENT,
WRONG_TABLE_ASSIGNMENT
};
/* Extension of std::exception to accomodate the parsers need */
class xppParserException : public std::exception {
private:
std::string m_msg;
public:
explicit xppParserException(const xppParserError msgType,
const lineNumber line,
const size_t pos)
{
switch (msgType) {
case DUPLICATED_NAME:
m_msg = std::string("Name has already been reserved");
break;
case MISSING_ARGUMENT:
m_msg = std::string("Cannot find function argument");
break;
case MISSING_CLOSING_BRACKET:
m_msg = std::string("Cannot find closing bracket");
break;
case MISSING_MARKOV_ASSIGNMENT:
m_msg = std::string("Missing assignment of markov transition probability");
break;
case MISSING_OPENING_BRACKET:
m_msg = std::string("Cannot find opening bracket");
break;
case EXPECTED_LIST_ARGUMENT:
m_msg = std::string("Missing argument in list");
break;
case EXPECTED_NUMBER:
m_msg = std::string("Cannot parse number");
break;
case EXPECTED_TABLE_ASSIGNMENT:
m_msg = std::string("Missing value for the lookup table");
break;
case RESERVED_FUNCTION:
m_msg = std::string("Given function name is reserved");
break;
case RESERVED_KEYWORD:
m_msg = std::string("Given name is a reserved keyword");
break;
case RESERVED_OPTION:
m_msg = std::string("Given name is a reserved option");
break;
case UNKNOWN_ASSIGNMENT:
m_msg = std::string("Unknown assignment");
break;
case UNKNOWN_FUNCTION:
m_msg = std::string("Unknown function in expression");
break;
case UNKNOWN_NAME:
m_msg = std::string("Unknown name in expression");
break;
case UNKNOWN_OPTION:
m_msg = std::string("Unknown option choosen");
break;
case UNKNOWN_VARIABLE:
m_msg = std::string("Unknown variable for initial condition");
break;
case WRONG_ARRAY_ASSIGNMENT:
m_msg = std::string("Wrong array assignment");
break;
case WRONG_MARKOV_ASSIGNMENT:
m_msg = std::string("Superflous assignment of markov transition probability");
break;
case WRONG_TABLE_ASSIGNMENT:
m_msg = std::string("Wrong assignment of table expression");
break;
}
m_msg += " in line " + std::to_string(line.second) + ":\n";
m_msg += line.first + "\n";
m_msg += std::string(pos, ' ') + "^\n";
}
virtual const char* what() const throw()
{
return m_msg.c_str();
}
};
#endif // XPPPARSEREXCEPTION_H
xppParserDefines.h
#ifndef XPPPARSERDEFINES_H
#define XPPPARSERDEFINES_H
#include <set>
#include <string>
#include <utility>
#include <vector>
/* Basic structure that contains the textual information of an expression*/
struct opts {
unsigned Line;
std::string Name;
std::string Expr;
std::vector<std::string> Args;
explicit opts () {}
explicit opts (const unsigned line) : Line(line) {}
explicit opts (const opts &opt)
: Line(opt.Line), Name(opt.Name), Expr(opt.Expr), Args(opt.Args) {}
};
/* Array of opts structures */
typedef std::vector<opts> optsArray;
/* Array of strings */
typedef std::vector<std::string> stringList;
/* Pair containing a parsed line and the original line number in the ode file.
* This is mainly usefull for debugging if an error is thrown.
*/
typedef std::pair<std::string, unsigned> lineNumber;
/* Vector containing the keywords utilized by xppaut. We cannot use a set here,
* as that would change the ordering of the symbols
*/
static const std::vector<std::string> xppKeywords = {
"!",
"(t+1)",
"'",
"/dt",
"(t)",
"volt",
// "markov", /* Handled separately */
"aux",
"par",
"number",
"(",
//"table", /* This is handeled separately */
//"wiener", /* Handled separately */
//"global",
"init",
"(0)",
"bdry",
"0=",
"solve",
"special",
"set",
"@",
"export"
};
/* Set of the valid mathematical operators */
static const std::set<std::string> xppOperators {
"+",
"-",
"*",
"/",
"^",
"**"
};
/* Set of the reserved function names */
static const std::set<std::string> xppReservedNames {
"sin",
"cos",
"tan",
"atan",
"atan2",
"sinh",
"cosh",
"tanh",
"exp",
"delay",
"ln",
"log",
"log10",
"t",
"pi",
"if",
"then",
"else",
"asin",
"acos",
"heav",
"sign",
"mod",
"flr",
"ran",
"abs",
"del_shft",
"max",
"min",
"normal",
"besselj",
"bessely",
"besseli",
"erf",
"erfc",
"poisson",
"lgamma",
"shift",
"sum",
"int",
"not"
};
/* Set of the options that can be modified in an ode file */
static const std::set<std::string> xppOptionNames {
"MAXSTOR",
"BACK",
"SMALL",
"BIG",
"SMC",
"UMC",
"XNC",
"YNC",
"OUTPUT",
"BUT",
"LT",
"XP",
"YP",
"ZP",
"NPLOT",
"XP2",
"YP2",
"ZP2",
"AXES",
"PHI",
"THETA",
"XLO",
"YLO",
"XHI",
"YHI",
"XMAX",
"XMIN",
"YMAX",
"YMIN",
"ZMAX",
"ZMIN",
"SEED",
"TOTAL",
"DT",
"NJMP",
"NOUT",
"T0",
"TRANS",
"NMESH",
"BANDUP",
"BANDLO",
"METH",
"DTMIN",
"DTMAX",
"VMAXPTS",
"JAC_EPS",
"NEWT_TOL",
"NEWT_ITER",
"ATOLER",
"TOLER",
"BOUND",
"DELAY",
"AUTOEVAL",
"POIMAP",
"POIVAR",
"POIPLN",
"POISGN",
"POISTOP",
"RANGE",
"RANGEOVER",
"RANGESTEP",
"RANGELOW",
"RANGEHIGH",
"RANGERESET",
"RANGEOLDIC",
"TOR_PER",
"FOLD",
"NTST",
"NMAX",
"NPR",
"DSMIN",
"DSMAX",
"DS",
"EPSS",
"EPSL",
"EPSU",
"PARMIN",
"PARMAX",
"NORMMIN",
"NORMMAX",
"AUTOXMIN",
"AUTOXMAX",
"AUTOYMIN",
"AUTOYMAX",
"AUTOVAR",
"BELL",
"COLORMAP"
};
#endif // XPPPARSERDEFINES_H
xppParser.h
#ifndef XPPPARSER_H
#define XPPPARSER_H
#include <iostream>
#include <algorithm>
#include <fstream>
#include <set>
#include <stack>
#include <string>
#include <utility>
#include <vector>
#include "mpParser.h"
#include "mpDefines.h"
#include "keywordTrie.hpp"
#include "xppParserDefines.h"
#include "xppParserException.h"
class xppParser {
public:
xppParser(std::string fn);
xppParser(const xppParser &parser);
private:
void expandArrays (void);
void expandArrayLines (std::vector<lineNumber> &lines,
const std::vector<lineNumber> &expressions,
const int idx);
void extractDefinition (void);
void extractExport (void);
void extractGlobal (void);
void extractMarkov (void);
void extractTable (void);
void extractWiener (void);
void initializeTries (void);
void readFile (void);
void removeComments (void);
void removeWhitespace (void);
void summarizeOde (void);
/* Helper functions */
void checkBrackets (void);
void checkName (const std::string &name, const lineNumber &line, size_t pos);
stringList getList (const std::string &line, unsigned ln,
const std::string &closure, const std::string &delim);
std::string getNextExpr (const lineNumber &line, size_t &pos1,size_t &pos2);
std::string getNextWord (const lineNumber &line, size_t &pos1,size_t &pos2);
bool isNumeric (const std::string &str);
keywordTrie::result keywordSearch(const std::string &key,
const char &character);
/* Filename of the ode file */
const std::string fileName;
/* Vector containing the individual lines from the ode file */
std::vector<lineNumber> lines;
/* Trie of xpp keyword */
keywordTrie::trie keywords;
/* Trie of xpp options */
keywordTrie::trie options;
/* Trie of reserved names */
keywordTrie::trie reservedNames;
/* Trie of the already used names */
keywordTrie::trie usedNames;
/* Options arrays */
optsArray Algebraic;
optsArray Auxiliar;
optsArray Boundaries;
optsArray Constants;
optsArray Equations;
optsArray Exports;
optsArray Functions;
optsArray Globals;
optsArray InitConds;
optsArray Temporaries;
optsArray Markovs;
optsArray Numbers;
optsArray Options;
optsArray Parameters;
optsArray Special;
optsArray Sets;
optsArray Tables;
optsArray Volterra;
opts Wieners;
friend class xppEvaluator;
};
#endif // XPPPARSER_H
xppParser.cpp
#include <xppParser.h>
/**
* @brief xppParser::xppParser Default constructor of the parser object
*
* @param fn string representing the file name of the ode file
*
* This constructs the parser object of a given ode file. First unneeded
* content is discarded, and a basic correctness check is don. Later arrays are
* expanded and special constructs like markov processes and tables are handled.
* Finally, the different keywords are parsed and put into the opts arrays.
*/
xppParser::xppParser(std::string fn)
: fileName(fn)
{
try {
/* Initially read in the ode file */
readFile();
/* Initialize the keyword tries for command parsing */
initializeTries();
/* Check for incorrect brackets */
checkBrackets();
/* Remove all comments */
removeComments();
/* Remove unnecessary whitespaces */
removeWhitespace();
/* Expand array descriptions */
expandArrays();
/* Extract exports to dlls */
extractExport();
/* Extract markov processes */
extractMarkov();
/* Extract table definitions */
extractTable();
/* Extract wiener processes */
extractWiener();
/* Extract globals */
extractGlobal();
/* Extract all other definitions */
extractDefinition();
/* Catch errors */
} catch (xppParserException& e) {
std::cerr << e.what();
} catch (std::runtime_error& e) {
std::cerr << e.what();
} catch (std::exception& e) {
std::cerr << e.what();
} catch (...) {
std::cerr << "Unexpected error\n";
throw;
}
}
/**
* @brief xppParser::xppParser Copy constructor of the parser object
* @param parser The parser object that should be copied
*/
xppParser::xppParser(const xppParser &parser)
: /* Keyword tries */
keywords(parser.keywords),
options(parser.options),
reservedNames(parser.reservedNames),
usedNames(parser.usedNames),
/* Opts arrays */
Algebraic(parser.Algebraic),
Auxiliar(parser.Auxiliar),
Boundaries(parser.Boundaries),
Constants(parser.Constants),
Equations(parser.Equations),
Exports(parser.Exports),
Functions(parser.Functions),
Globals(parser.Globals),
InitConds(parser.InitConds),
Temporaries(parser.Temporaries),
Markovs(parser.Markovs),
Numbers(parser.Numbers),
Options(parser.Options),
Parameters(parser.Parameters),
Special(parser.Special),
Sets(parser.Sets),
Tables(parser.Tables),
Volterra(parser.Volterra),
Wieners(parser.Wieners)
{}
/**
* @brief Checks whether brackets are closed properly
*/
void xppParser::checkBrackets() {
using charPos = std::pair<char, size_t>;
std::stack<charPos> brackets;
for (lineNumber line : lines) {
auto start = line.first.begin();
for (auto it = start; it != line.first.end(); ++it) {
switch (*it) {
case '(':
brackets.push(charPos(')', std::distance(start, it)));
break;
case '{':
brackets.push(charPos('}', std::distance(start, it)));
break;
case '[':
brackets.push(charPos(']', std::distance(start, it)));
break;
case ')':
if (brackets.empty()) {
throw xppParserException(MISSING_OPENING_BRACKET, line,
std::distance(start, it));
} else if (brackets.top().first != ')') {
throw xppParserException(MISSING_CLOSING_BRACKET, line,
brackets.top().second);
}
brackets.pop();
break;
case '}':
if (brackets.empty()) {
throw xppParserException(MISSING_OPENING_BRACKET, line,
std::distance(start, it));
} else if (brackets.top().first != '}') {
throw xppParserException(MISSING_CLOSING_BRACKET, line,
brackets.top().second);
}
brackets.pop();
break;
case ']':
if (brackets.empty()) {
throw xppParserException(MISSING_OPENING_BRACKET, line,
std::distance(start, it));
} else if (brackets.top().first != ']') {
throw xppParserException(MISSING_CLOSING_BRACKET, line,
brackets.top().second);
}
brackets.pop();
break;
default:
continue;
}
}
if (!brackets.empty()) {
throw xppParserException(MISSING_CLOSING_BRACKET, line,
brackets.top().second);
}
}
}
/**
* @brief Checks whether a given name is already taken or reserved
*
* @par name: The name of the new definition
* @par line: The original line that was parsed from the ode file
* @par pos: The position of the name in line
*/
void xppParser::checkName(const std::string &name, const lineNumber &line, size_t pos) {
keywords.setWholeWords(true);
if (!usedNames.parseText(name).empty()) {
throw xppParserException(DUPLICATED_NAME, line, pos);
} else if (!reservedNames.parseText(name).empty()) {
throw xppParserException(RESERVED_FUNCTION, line, pos);
} else if (!keywords.parseText(name).empty()) {
throw xppParserException(RESERVED_KEYWORD, line, pos);
} else if (!options.parseText(name).empty()) {
throw xppParserException(RESERVED_OPTION, line, pos);
}
keywords.setWholeWords(false);
usedNames.addString(name);
}
/**
* @brief Expand array expressions in ode files
*
* This function searches for the array expressions found in ode files, that are
* marked by "x[start .. end]" for single line or "%[start .. end] ... %" for
* multi line statements. Every array line is duplicated end-start+1 times,
* while preserving the original order of lines.
*/
void xppParser::expandArrays() {
auto line = lines.begin();
while (line != lines.end()) {
std::size_t pos1, pos2, pos3;
/* Search for opening brackets */
pos1 = line->first.find("[");
if (pos1 != std::string::npos) {
/* Search for the closing bracket and the dots between the start and
* the end indices. The dots are only necessary for single line
* statements.
*/
pos2 = line->first.find("..");
pos3 = line->first.find("]");
if (pos3 == std::string::npos) {
throw xppParserException(MISSING_CLOSING_BRACKET, *line, pos1);
} else if (pos2 != std::string::npos && pos2 > pos3) {
throw xppParserException(WRONG_ARRAY_ASSIGNMENT, *line, pos2);
}
/* Determine the range of the indices of the array */
int start, end;
try {
start = std::stoi(line->first.substr(pos1+1, pos2-pos1-1));
} catch (std::invalid_argument) {
throw xppParserException(EXPECTED_NUMBER, *line, pos1+1);
}
try {
end = std::stoi(line->first.substr(pos2+2, pos3-pos2-2));
} catch (std::invalid_argument) {
throw xppParserException(EXPECTED_NUMBER, *line, pos2+2);
}
/* Copy the lines of the assignment into a separate vector */
std::vector<lineNumber> arrayExpressions;
auto line2 = std::next(line);
if (line->first.substr(pos1-1, 1) == "%") {
/* Multiline statements end with a line with a single "%" */
while (true) {
size_t endArray = line2->first.find("%");
if (endArray != std::string::npos) {
++line2;
break;
}
arrayExpressions.push_back(*line2);
++line2;
}
} else {
arrayExpressions.push_back(*line);
/* Change the first bracket to [j] to unify expression handling
* with the multiline case.
*/
arrayExpressions[0].first.replace(pos1, pos3, "[j]");
}
lines.erase(line, line2);
/* Expand the array expressions and insert it*/
std::vector<lineNumber> arrayLines;
arrayLines.reserve(end-start+1);
for (int j = start; j <= end; j++) {
expandArrayLines(arrayLines, arrayExpressions, j);
}
lines.insert(line, arrayLines.begin(), arrayLines.end());
} else {
++line;
}
}
}
/**
* @brief Inserts a given array expression into the ode list for given j
*
* @param lines A vector that is to be filled with the resulting lines.
* @param expressions A vector containing the individual lines of array
* expressions.
* @param idx The index that should be used for evaluation of the expressions.
*
* This function takes the array expressions and replaces every assignment of
* type [...] with the evaluated expression for j=idx. Afterwards the resulting
* line is inserted into the provided vector.
*/
void xppParser::expandArrayLines(std::vector<lineNumber>& lines,
const std::vector<lineNumber>& expressions,
const int idx) {
/* Initialize the parser with the variable "j" */
mup::ParserX parser;
mup::Value j((mup::int_type)idx);
mup::Value result;
parser.DefineVar("j", mup::Variable(&j));
for (const lineNumber &expr : expressions) {
lineNumber temp = expr;
size_t pos1 = temp.first.find("[");
size_t pos2 = temp.first.find("]");
while (pos1 != std::string::npos) {
/* Evaluate the expression within the brackets and replace the
* brackets with the result.
*/
parser.SetExpr(temp.first.substr(pos1+1, pos2-pos1-1));
result = parser.Eval();
/* Check whether the result is an integer */
if (result.GetType() != 'i') {
throw xppParserException(WRONG_ARRAY_ASSIGNMENT, expr, pos1+1);
}
temp.first.replace(pos1, pos2-pos1+1, result.ToString());
pos1 = temp.first.find("[", pos2);
pos2 = temp.first.find("]", pos1);
}
lines.push_back(temp);
}
}
/**
* @brief Extract definitions from the ode file
*
* This extracts definitions that are given in the keyword list and are marked
* by an equal sign.
*/
void xppParser::extractDefinition(void) {
auto line = lines.begin();
while (line != lines.end()) {
/* Search for the first keyword. In most cases it should be the first
* consecutive string that precedes a whitespace or equal sign.
*/
std::size_t pos1 = 0, pos2 = 0;
std::string key = getNextWord(*line, pos1, pos2);
if (pos2 == std::string::npos) {
throw xppParserException(UNKNOWN_ASSIGNMENT, *line, pos1+1);
}
/* Search for keywords */
auto res = keywordSearch(key, line->first.at(pos2));
while (pos2 != std::string::npos) {
opts opt(line->second);
switch (res.id) {
case 0: /* !Name */
opt.Name = line->first.substr(pos1+1, pos2-pos1-1);
break;
case 1: /* Name(t+1) */
case 2: /* Name' */
case 4: /* Name(t) */
case 11: /* Name(0) */
opt.Name = line->first.substr(pos1, res.start);
break;
case 3: /* dName/dt */
opt.Name = line->first.substr(pos1+1, res.start-1);
break;
case 9: {/* Name(args...) */
size_t pos3 = line->first.find("(", pos1);
opt.Name = line->first.substr(pos1, pos3-pos1);
opt.Args = getList(line->first.substr(pos3, pos2-pos3),
opt.Line, ")", ",");
break;
}
case 12:
break;
case 13: /* 0=Expression */
opt.Name = "Initial Condition";
break;
case 19: /* Name=Expression */
opt.Name = line->first.substr(pos1, pos2-pos1);
break;
default: /* keyword Name */
opt.Name = getNextWord(*line, pos1, pos2);
break;
}
/* Check whether the name is already taken/reserved, except for
* initial conditions, where we check for existence.
*/
if (res.id != 10 &&
res.id != 17) {
checkName(opt.Name, *line, pos1);
} else if (res.id == 10) {
if (usedNames.parseText(opt.Name).empty()) {
throw xppParserException(UNKNOWN_VARIABLE, *line, pos1);
}
} else if (res.id == 17) {
if (options.parseText(opt.Name).empty()) {
throw xppParserException(UNKNOWN_OPTION, *line, pos1);
}
}
/* Get the expression */
opt.Expr = getNextExpr(*line, pos1, pos2);
/* Check numbers are indeed numeric expressions */
if (res.id == 8) {
if (!isNumeric(opt.Expr)) {
throw xppParserException(EXPECTED_NUMBER, *line, pos1);
}
/* Check if all function arguments are used */
} else if (res.id == 9) {
size_t pos3 = opt.Name.length()+1;
for (std::string &str : opt.Args) {
if (opt.Expr.find(str) == std::string::npos) {
throw xppParserException(MISSING_ARGUMENT, *line, pos3);
}
pos3 += str.length()+1;
}
}
/* Find the type of the keyword */
switch(res.id) {
case 0:
Constants.push_back(opt);
break;
case 1:
case 2:
case 3:
Equations.push_back(opt);
break;
case 4:
Volterra.push_back(opt);
break;
case 5:
Constants.push_back(opt);
break;
case 6:
Auxiliar.push_back(opt);
break;
case 7:
Parameters.push_back(opt);
break;
case 8:
Numbers.push_back(opt);
break;
case 9:
Functions.push_back(opt);
break;
case 10:
InitConds.push_back(opt);
break;
case 11:
Volterra.push_back(opt);
break;
case 12:
/* Boundary expressions do not have a name */
opt.Expr = opt.Name;
opt.Name = "";
Boundaries.push_back(opt);
break;
case 13:
Volterra.push_back(opt);
break;
case 14:
Algebraic.push_back(opt);
break;
case 15:
/* Extract the argument list */
pos1 = opt.Expr.find("(");
opt.Args = getList(opt.Expr.substr(pos1), opt.Line, ")", ",");
opt.Expr.resize(pos1);
Special.push_back(opt);
break;
case 16:
/* Sets are a comma separated list */
opt.Args = getList(opt.Expr, opt.Line, "}", ",");
opt.Expr = "";
Sets.push_back(opt);
break;
case 17:
Options.push_back(opt);
break;
case 18:
Exports.push_back(opt);
break;
case 19:
Temporaries.push_back(opt);
break;
default:
throw xppParserException(UNKNOWN_ASSIGNMENT, *line, pos1+1);
break;
}
}
lines.erase(line);
}
}
/**
* @brief Extract exports to link to procompiled C routines inside xppaut
*
* This function extracts the in and out arguments of the respective function.
* To keep with the other structures the Args vector contains both the input and
* the output. Therefore the Expr string contains the number of input arguments.
*/
void xppParser::extractExport(void) {
auto line = lines.begin();
while (line != lines.end()) {
std::size_t pos1 = line->first.find("export");
std::size_t pos2 = line->first.find(" ", pos1);
if (pos1 != std::string::npos) {
opts opt(line->second);
opt.Args = getList(getNextWord(*line, pos1, pos2), opt.Line, "}", ",");;
opt.Expr = std::to_string(opt.Args.size());
stringList temp = getList(getNextWord(*line, pos1, pos2), opt.Line, "}", ",");
opt.Args.insert(opt.Args.end(), temp.begin(), temp.end());
Exports.push_back(opt);
lines.erase(line);
} else {
++line;
}
}
}
/**
* @brief Extract a global flag implementing a delta function
*
* This function extracts the condition, the sign of the flag and the resets
*/
void xppParser::extractGlobal(void) {
auto line = lines.begin();
while (line != lines.end()) {
std::size_t pos1 = line->first.find("global");
std::size_t pos2 = line->first.find(" ", pos1);
if (pos1 != std::string::npos) {
opts opt(line->second);
/* Parse the sign flag. For simplicity store it in the name slot */
opt.Name = getNextWord(*line, pos1, pos2);
/* Parse flip condition. */
pos1 = line->first.find_first_not_of(" ", pos2);
pos2 = line->first.find("{", pos1);
opt.Expr = line->first.substr(pos1, pos2-pos1-1);
/* The individual resets are separated by a semicolon */
opt.Args = getList(line->first.substr(pos2), opt.Line, "}", ";");
Globals.push_back(opt);
lines.erase(line);
} else {
++line;
}
}
}
/**
* @brief Extract a markov process from the parsed lines
*
* This function extracts the definition of markov processes as they constitute
* one of the few multiline statements in an ode file.
*/
void xppParser::extractMarkov(void) {
auto line = lines.begin();
while (line != lines.end()) {
std::size_t pos1 = line->first.find("markov");
std::size_t pos2 = line->first.find(" ", pos1);
if (pos1 != std::string::npos) {
opts opt(line->second);
opt.Name = getNextWord(*line, pos1, pos2);
checkName(opt.Name, *line, pos1);
/* Parse the number of states */
int nstates;
try {
nstates = std::stoi(getNextWord(*line, pos1, pos2));
} catch (std::invalid_argument) {
throw xppParserException(EXPECTED_NUMBER, *line, pos1);
}
opt.Expr = nstates;
/* Parse the transition probabilities */
opt.Args.reserve(nstates*nstates);
for (int i=0; i < nstates; i++) {
auto line2 = std::next(line);
pos2 = 0;
for (int j=0; j < nstates; j++) {
pos1 = line2->first.find("{", pos2);
pos2 = line2->first.find("}", pos1);
if (pos1 == std::string::npos) {
throw xppParserException(MISSING_MARKOV_ASSIGNMENT,
*line2, pos2);
}
opt.Args.push_back(line2->first.substr(pos1+1, pos2-pos1-1));
}
if (pos2 != line2->first.size()-1) {
throw xppParserException(WRONG_MARKOV_ASSIGNMENT,
*line2, line2->first.size());
}
lines.erase(line2);
}
Markovs.push_back(opt);
lines.erase(line);
} else {
++line;
}
}
}
/**
* @brief Extract a table
*
* This function extracts a precomputed table either from a file or calculates
* it from the definition.
*/
void xppParser::extractTable(void) {
auto line = lines.begin();
while (line != lines.end()) {
std::size_t pos1 = line->first.find("table");
std::size_t pos2 = line->first.find(" ", pos1);
if (pos1 != std::string::npos) {
opts opt(line->second);
unsigned npoints;
double xLow, xHigh;
/* Initialize the parser */
mup::ParserX parser;
/* Parse the name */
opt.Name = getNextWord(*line, pos1, pos2);
checkName(opt.Name, *line, pos1);
/* If the table has to be calculated there is a % sign instead of a
* filename
*/
std::string fn = getNextWord(*line, pos1, pos2);
if (fn == "%") {
/* Get the number of points */
try {
npoints = std::stoi(getNextWord(*line, pos1, pos2));
} catch (std::invalid_argument) {
throw xppParserException(EXPECTED_NUMBER, *line, pos1);
}
/* Get the bounds */
parser.SetExpr(getNextWord(*line, pos1, pos2));
try {
xLow = parser.Eval().GetFloat();
} catch (mup::ParserError) {
throw xppParserException(EXPECTED_NUMBER, *line, pos1);
}
parser.SetExpr(getNextWord(*line, pos1, pos2));
try {
xHigh = parser.Eval().GetFloat();
} catch (mup::ParserError) {
throw xppParserException(EXPECTED_NUMBER, *line, pos1);
}
/* Parse the defining function */
mup::Value t;
parser.DefineVar("t", mup::Variable(&t));
parser.SetExpr(getNextWord(*line, pos1, pos2));
/* Evaluate the table expression */
double dx = (xHigh - xLow)/(npoints-1);
try {
opt.Args.reserve(npoints);
for(unsigned j = 0; j < npoints; j++) {
t = (mup::float_type)(xLow + j * dx);
opt.Args.push_back(parser.Eval().ToString());
}
} catch (mup::ParserError) {
throw xppParserException(WRONG_TABLE_ASSIGNMENT, *line, pos1);
}
} else {
/* Open file for parsing */
std::ifstream fileStream(fn.c_str(), std::ios::in);
if (fileStream.fail()) {
throw std::runtime_error("Cannot open table file " + fn + "\n");
}
lineNumber temp;
temp.second = 1;
/* Get the number of points */
try {
getline(fileStream, temp.first);
++temp.second;
npoints = std::stoi(temp.first);
} catch (std::invalid_argument) {
throw xppParserException(EXPECTED_NUMBER, temp, pos1);
}
/* Get the bounds */
getline(fileStream, temp.first);
++temp.second;
parser.SetExpr(temp.first);
try {
xLow = parser.Eval().GetFloat();
} catch (mup::ParserError) {
throw xppParserException(EXPECTED_NUMBER, temp, pos1);
}
getline(fileStream, temp.first);
++temp.second;
parser.SetExpr(temp.first);
try {
xHigh = parser.Eval().GetFloat();
} catch (mup::ParserError) {
throw xppParserException(EXPECTED_NUMBER, temp, pos1);
}
/* Parse the table values */
try {
opt.Args.reserve(npoints);
for(unsigned j = 0; j < npoints; j++) {
if (getline(fileStream, temp.first)) {
++temp.second;
parser.SetExpr(temp.first);
opt.Args.push_back(parser.Eval().ToString());
} else {
throw xppParserException(EXPECTED_TABLE_ASSIGNMENT,
temp, pos1);
}
}
} catch (mup::ParserError) {
throw xppParserException(WRONG_TABLE_ASSIGNMENT, temp, pos1);
}
fileStream.close();
}
Tables.push_back(opt);
lines.erase(line);
} else {
++line;
}
}
}
/**
* @brief Extracts wiener processes
*
* This function extracts the declaration of wiener processes as their
* definition does not contain an equal sign but only the names of the
* individual proceses.
*/
void xppParser::extractWiener(void) {
auto line = lines.begin();
while (line != lines.end()) {
std::size_t pos1 = line->first.find("wiener");
std::size_t pos2 = line->first.find(" ", pos1);
if (pos1 != std::string::npos) {
while (pos2 != std::string::npos) {
Wieners.Args.push_back(getNextWord(*line, pos1, pos2));
checkName(Wieners.Args.back(), *line, pos1);
}
lines.erase(line);
} else {
++line;
}
}
}
/**
* @brief Extracts elements from a brace enclosed list
*
* @par line: String we are searching in
* @par encloure: The bracket type { or (
* @par delim: The delimiter type , or ;
*
* @return Args: Vector of strings of the list elements
*/
stringList xppParser::getList(const std::string& line,
unsigned ln,
const std::string &closure,
const std::string &delim) {
size_t pos1 = 1;
size_t pos2 = line.find_first_of(delim+closure, pos1);
stringList temp;
while (pos2 != std::string::npos) {
temp.push_back(line.substr(pos1, pos2-pos1));
if (temp.back().empty()) {
throw xppParserException(EXPECTED_LIST_ARGUMENT,
std::make_pair(line, ln), pos1);
}
pos1 = pos2+1;
pos2 = line.find_first_of(delim+closure, pos1);
}
return temp;
}
/**
* @brief Gets the next expression between pos2 and a commata outside braces
*
* @par line: String we are searching in
* @par pos1: Old position. Will be updated to the position after the commata
* or std::string::npos.
* @par pos2: Position of the equal sign before the expression
*
* @return string: String between pos2 and the next commata outside of braces
*/
std::string xppParser::getNextExpr(const lineNumber &line,
size_t &pos1,
size_t &pos2) {
pos1 = pos2 + 1;
/* Search for the next commata outside of brackets */
std::stack<char> brackets;
auto it = line.first.begin();
for (std::advance(it, pos2); it != line.first.end(); ++it) {
switch (*it) {
case '(':
brackets.push(')');
break;
case '{':
brackets.push('}');
break;
case '[':
brackets.push(']');
break;
case ')':
brackets.pop();
break;
case '}':
brackets.pop();
break;
case ']':
brackets.pop();
break;
case ',':
if (brackets.empty()) {
pos2 = std::distance(line.first.begin(), it)+1;
it = std::prev(line.first.end());
} else {
continue;
}
break;
default:
continue;
break;
}
}
/* If none was found copy the whole string */
if (pos1 == pos2 +1) {
pos2 = std::string::npos;
}
std::string expr = line.first.substr(pos1, pos2-pos1-1);
/* Remove whitespaces */
expr.erase(remove_if(expr.begin(), expr.end(), isspace), expr.end());
return expr;
}
/**
* @brief Gets the next whitespace separated word after pos2
*
* @par line: String we are searching in
* @par pos1: Old position which will be udpated
* @par pos2: Old position at which the search startes
*
* @return pos1: First position of the new word
* @return pos2: Position of the first whitespace character after the word
* @return string: String between [pos1, pos2-1]
*/
std::string xppParser::getNextWord(const lineNumber &line,
size_t &pos1,
size_t &pos2) {
pos1 = line.first.find_first_not_of(" ", pos2);
pos2 = line.first.find_first_of(" =", pos1);
return line.first.substr(pos1, pos2-pos1);
}
/**
* @brief Initializes the keyword tree from the keyword list
*/
void xppParser::initializeTries (void) {
keywords.addString(xppKeywords);
options.setCaseSensitivity(false);
options.setWholeWords(true);
options.addString(xppOptionNames);
reservedNames.setWholeWords(true);
reservedNames.addString(xppReservedNames);
usedNames.setWholeWords(true);
}
/**
* @brief Checks whether a provided string is a numeric expression
*
* @par str: The std::string containing the expression
*
* @return true if the string is a numeric expression, false otherwise
*/
bool xppParser::isNumeric(const std::string &str) {
char* p;
std::strtod(str.c_str(), &p);
return (*p == 0.0);
}
/**
* @brief Reads in the ode file and stores the lines in a vector
*
* This function reads in the ode file given by fileName, ignoring empty lines
* as well lines containing only whitespace. Furthermore it trims and truncates
* all whitespaces.
*/
void xppParser::readFile(void) {
std::ifstream fileStream(fileName.c_str(), std::ios::in);
if (fileStream.fail()) {
throw std::runtime_error("Cannot open ode file " + fileName + "\n");
}
/* Parse in nonempty lines */
std::string temp;
int lineCount = 1;
while(getline(fileStream, temp)) {
size_t pos1 = temp.find_first_not_of(" \t\f\v\r\n");
if (temp == "done") {
break;
} else if (!temp.empty() && pos1 != std::string::npos) {
/* Remove trailing and superflous whitespaces */
temp.erase(0, pos1);
temp.resize(temp.find_last_not_of(" \t\f\v\r\n")+1);
auto last = std::unique(temp.begin(), temp.end(),
[](char l, char r){return std::isspace(l) &&
std::isspace(r);});
temp.erase(last, temp.end());
lines.push_back(std::make_pair(temp, lineCount));
}
++lineCount;
}
if (lines.size() == 0) {
throw std::runtime_error("Empty ode file " + fileName + "\n");
}
fileStream.close();
}
/**
* @brief Removes comments from ode files
*
* This function purges all standard comments from a ode file. If a line would
* be comment only it is deleted.
*/
void xppParser::removeComments() {
auto line = lines.begin();
while (line != lines.end()) {
std::size_t pos1 = line->first.find("#");
if (pos1 == 0) {
lines.erase(line);
} else if (pos1 != std::string::npos) {
/* Check if this is the definition of a convolutional integral
* indicated by curly braces sourrunding the # sign
*/
std::size_t subpos1 = line->first.find_last_of("{", pos1);
std::size_t subpos2 = line->first.find("}", subpos1);
if (subpos1 == std::string::npos || subpos2 < pos1) {
line->first.resize(pos1);
}
++line;
} else {
++line;
}
}
}
/**
* @brief Removes unneeded whitespaces
*
* This function searches for whitespaces within parentheses or before/after an
* equal sign and removes those.
*/
void xppParser::removeWhitespace() {
auto line = lines.begin();
while (line != lines.end()) {
std::size_t pos1, pos2, pos3;
/* Remove all whitespaces between brackets */
pos1 = line->first.find("(");
if (pos1 != std::string::npos) {
pos2 = line->first.find(")");
pos3 = line->first.find_first_of(" \t\f\v\r", pos1);
while (pos3 < pos2) {
line->first.erase(pos3, 1);
pos3 = line->first.find_first_of(" \t\f\v\r", pos1);
}
}
/* Search around the equal signs */
pos1 = line->first.find("=");
if (pos1 != std::string::npos) {
pos2 = line->first.find_first_not_of(" \t\f\v\r", pos1+1);
if (pos2 != pos1+1) {
line->first.erase(pos1+1, pos2-pos1-1);
}
pos2 = line->first.find_last_not_of(" \t\f\v\r", pos1-1);
if (pos2 != pos1-1) {
line->first.erase(pos2+1, pos1-pos2-1);
}
}
/* Remove trailing commata etc */
pos1 = line->first.find_last_not_of(",;");
if (pos1 != line->first.size()-1) {
line->first.resize(pos1+1);
}
++line;
}
}
/**
* @brief Checks a keyword search in case of multiple matches
*
* @par expr: The std::string to be searched for the keyword
* @par character: The character after expr.
*
* This sanitizes the keyword search in case we have multiple matches. If
* multiple keywords were found, the name of the expression contains a keyword.
* This is only possible if it is either an temporary expression or a keyword
* that is merged with the name. Consequently remove all other findings. If
* there were no valid keywords left/found this must be a temporary expression.
* So create a fake result with and index equal to the size of xppKeywords.
*/
keywordTrie::result xppParser::keywordSearch(const std::string &expr,
const char &character) {
keywordTrie::resultCollection results = keywords.parseText(expr);
if(character == '=') {
auto it = results.begin();
for (keywordTrie::result &res : results) {
if (res.id == 0 || /* !Name */
res.id == 2 || /* Name' */
res.id == 3 || /* dName/dt */
res.id == 9) { /* Name(Args) */
++it;
} else if (res.id == 1 || /* Name(t) */
res.id == 4 || /* Name(t+1) */
res.id == 11) { /* Name(0) */
results.erase(--it); /* Remove Name(Args) */
} else {
results.erase(it);
}
}
/* No valid keyword left, this must be an expression */
if (results.size() == 0) {
results.push_back(keywordTrie::result("", xppKeywords.size()));
}
}
return results.front();
}
/**
* @brief Returns the different parsed objects and their definition
*/
void xppParser::summarizeOde() {
for (opts &opt : Constants) {
std::cout << "Added constant " << opt.Name
<< "=" << opt.Expr << std::endl;
}
for (opts &opt : Numbers) {
std::cout << "Added number " << opt.Name
<< "=" << opt.Expr << std::endl;
}
for (opts &opt : Temporaries) {
std::cout << "Added temporary " << opt.Name
<< "=" << opt.Expr << std::endl;
}
for (opts &opt : Parameters) {
std::cout << "Added parameter " << opt.Name
<< "=" << opt.Expr << std::endl;
}
for (opts &opt : Algebraic) {
std::cout << "Added algebraic expression " << opt.Name
<< "=" << opt.Expr << std::endl;
}
for (opts &opt : Auxiliar) {
std::cout << "Added auxiliar variable " << opt.Name
<< "=" << opt.Expr << std::endl;
}
for (opts &opt : Boundaries) {
std::cout << "Added boundary condition: " << opt.Name << std::endl;
}
for (opts &opt : Globals) {
std::cout << "Added delta function " << opt.Name
<< "=" << opt.Expr << std::endl;
}
for (opts &opt : Functions) {
std::cout << "Added function " << opt.Name << "(";
std::string temp;
for (std::string &arg : opt.Args) {
temp += arg + ",";
}
temp.pop_back();
std::cout << temp << ")=" << opt.Expr << std::endl;
}
for (opts &opt : Equations) {
std::cout << "Added ODE " << opt.Name
<< "'=" << opt.Expr << std::endl;
}
for (opts &opt : Volterra) {
std::cout << "Added differential ODE " << opt.Name
<< "=" << opt.Expr << std::endl;
}
if (Wieners.Args.size() > 0) {
std::cout << "Added wiener processes ";
for (std::string &str : Wieners.Args) {
std::cout << str << " ";
}
std::cout << std::endl;
}
for (opts &opt : Markovs) {
std::cout << "Added markov processes " << opt.Name << std::endl;
}
for (opts &opt : InitConds) {
std::cout << "Set initial conditions " << opt.Name
<< "=" << opt.Expr << std::endl;
}
}
keywordTrie.hpp
/*
* Copyright (C) 2016 Michael Schellenberger Costa.
*
* This code is based on a C-implementation of the keyword trie construction,
* preprocessing and text search by Bernhard Haubold.
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in all
* copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*/
#ifndef KEYWORDTRIE_HPP
#define KEYWORDTRIE_HPP
#include <queue>
#include <set>
#include <stdexcept>
#include <string>
#include <vector>
namespace keywordTrie {
/**
* @brief The node struct containing the information of a trie node.
*/
struct node {
int id = -1; /**< Keyword index */
const unsigned depth= 0; /**< Depth in the trie*/
const char c = '\0'; /**< Character labelling the incoming edge */
const node *parent = nullptr; /**< Parent node */
node *failure = nullptr; /**< Failure link */
node *output = nullptr; /**< Output link */
std::vector<node*> children; /**< Child nodes */
explicit node () {}
explicit node (const unsigned d, const char &character, const node *par, node *root)
: depth(d), c(character), parent(par), failure(root), output(root) {}
};
/**
* @brief The result struct containing the information about matches during a
* search.
*/
struct result {
std::string keyword; /**< The found keyword */
unsigned id; /**< The index of the keyword in the keyword list*/
unsigned start; /**< The starting position of the match */
unsigned end; /**< The end position of the match */
explicit result (const std::string &key, const unsigned id)
: keyword(key), id(id) {}
explicit result (const result &res, const unsigned endPos)
: keyword(res.keyword), id(res.id), start(endPos-res.keyword.size()+1),
end(endPos) {}
};
typedef std::vector<keywordTrie::result> resultCollection;
typedef std::vector<resultCollection> resultTable;
/**
* @brief The trie class representing the keyword trie.
*/
class trie
{
private:
node *root = nullptr; /**< The root node */
std::vector<node*> trieNodes; /**< Container of the node pointers */
resultCollection keywords; /**< Container of the result stubs */
bool caseSensitive = true; /**< Flag for case sensitivity */
bool wholeWords = false; /**< Flag for result validity */
public:
/**
* @brief trie Initializes the trie structure with its root node.
*/
trie() {
root = new node();
root->parent = root;
root->failure = root;
root->output = root;
trieNodes.push_back(root);
}
/**
* @brief trie Copy an existing keyword trie.
*/
trie(const trie &Trie)
: trie() {
for (const result &res : Trie.keywords) {
addString(res.keyword, false);
}
addFailureLinks();
}
/**
* @brief ~trie Destructor of the trie that frees nodes from heap.
*/
~trie() {for (node* N : trieNodes) delete N;}
/**
* @brief addString Insert a new keyword into the keyword trie.
* @param key The new keyword to be inserted.
* @param addFailure Flag to signal whether the failure links should
* immediately be updated.
*/
void addString (const std::string &key, bool addFailure) {
if (key.empty()) {
return;
}
node *current = root;
for (const char &character : key) {
const char c = caseSensitive ? character : std::tolower(character);
current = addChild(current, c);
}
if (current->id != -1) {
throw std::runtime_error(
"Attempted to add two identical strings to the keyword tree.");
}
current->id = keywords.size();
keywords.push_back(result(key, keywords.size()));
if (addFailure) {
addFailureLinks();
}
}
/**
* @brief addString Wrapper around addString(std::string, bool).
* @param key The new keyword to be inserted.
*/
void addString (const std::string &key) {addString(key, true);}
/**
* @brief addStrings Wrapper around addString(std::string, bool) to add a
* set of strings.
* @param keyList The set containing the keys.
*/
void addString(const std::set<std::string> &keyList) {
for (const std::string &key : keyList) {
addString(key, false);
}
addFailureLinks();
}
/**
* @brief addStrings Wrapper around addString(std::string, bool) to add a
* vector of strings.
* @param keyList The vector containing the keys.
*/
void addString(const std::vector<std::string> &keyList) {
for (const std::string &key : keyList) {
addString(key, false);
}
addFailureLinks();
}
/**
* @brief parseText Parses a text with the trie.
* @param text The text to be parsed.
* @return Returns a vector with all matches.
*/
resultCollection parseText (const std::string &text) const {
resultCollection results;
if (text.empty()) {
return results;
}
node *current= root;
for (unsigned i=0; i < text.size(); i++) {
const char c = caseSensitive ? text.at(i) : std::tolower(text.at(i));
current = findChild(current, c);
if (current->id != -1) {
if (!wholeWords || current->depth == text.size()) {
results.push_back(result(keywords.at(current->id), i));
}
}
/* Process the output links for possible additional matches */
if (!wholeWords) {
node *temp = current->output;
while (temp != root) {
results.push_back(result(keywords.at(temp->id), i));
temp = temp->output;
}
}
}
return results;
}
/**
* @brief setCaseSensitivity Set the case sensitivity flag.
* @param flag The new flag.
*/
void setCaseSensitivity (bool flag) {
caseSensitive = flag;
if (!keywords.empty()) {
throw std::runtime_error("Switching case sensitivity with existing "
"trie might lead to invalid results");
}
}
/**
* @brief setWholeWords Defines whether partial matches a valid.
* @param flag The new flag.
*/
void setWholeWords (bool flag) {
wholeWords = flag;
}
private:
/**
* @brief addChild Add a child node to the trie.
* @param parrent The pointer to the parrent node of the new one.
* @param character The character on the edge to the new node.
* @return The pointer to the newly created node.
*/
node* addChild (node *current, const char &character) {
for (node *child : current->children) {
if (child->c == character) {
return child;
}
}
trieNodes.push_back(new node(current->depth+1, character, current, root));
current->children.push_back(trieNodes.back());
return trieNodes.back();
}
/**
* @brief findChild Searches for a child node with given character or adds one.
* @param current The pointer to the current node.
* @param character The character that is searched.
* @param addWord Flag sign to decide whether a new node should be added.
* @return The pointer to the matching node (possibly after failure links),
* root or the newly created one.
*/
node* findChild (const node *current, const char &character) const {
for (node *child : current->children) {
if (child->c == character) {
return child;
}
}
return traverseFail(current, character);
}
/**
* @brief traverseFail Traverse the failure links during a search.
* @param current The original node.
* @param character The character that is beeing searched.
* @return The pointer to the matching node after a failure link or root
*/
node* traverseFail (const node *current, const char &character) const {
node *temp = current->failure;
while (temp != root) {
for (node *failchild : temp->children) {
if (failchild->c == character) {
return failchild;
}
}
temp = temp->failure;
}
for (node *rootchild : root->children) {
if (rootchild->c == character) {
return rootchild;
}
}
return root;
}
/**
* @brief addFailureLinks Utilize a breadth first search to generate the
* failure links.
*/
void addFailureLinks() {
std::queue<node*> q;
q.push(root);
while (!q.empty()) {
node *temp = q.front();
for (node *child : temp->children) {
q.push(child);
}
/* A failure link with just one less charater is the optimum and will
* never change.
*/
if (temp->failure->depth < temp->depth - 1) {
for (node *failchild : temp->parent->failure->children) {
if (failchild->c == temp->c) {
temp->failure = failchild;
}
}
}
/* Process the failure links for possible additional matches */
node *out = temp->failure;
while (out != root) {
if (out->id != -1) {
break;
}
out = out->failure;
}
temp->output = out;
q.pop();
}
}
};
} // namespace keywordTrie
#endif // KEYWORDTRIE_HPP
