From ad369989078488cee72b61d548d1a4adad7a24c5 Mon Sep 17 00:00:00 2001 From: Jan Travnicek <Jan.Travnicek@fit.cvut.cz> Date: Sat, 12 Jul 2014 13:16:29 +0200 Subject: [PATCH] Noncontracting grammar --- alib2/src/grammar/ContextSensitive/CSG.cpp | 3 +- .../NonContractingGrammar.cpp | 132 ++++++++++++++++++ .../ContextSensitive/NonContractingGrammar.h | 51 +++++++ alib2/src/grammar/GrammarToXMLComposer.h | 4 +- 4 files changed, 186 insertions(+), 4 deletions(-) create mode 100644 alib2/src/grammar/ContextSensitive/NonContractingGrammar.cpp create mode 100644 alib2/src/grammar/ContextSensitive/NonContractingGrammar.h diff --git a/alib2/src/grammar/ContextSensitive/CSG.cpp b/alib2/src/grammar/ContextSensitive/CSG.cpp index 672e2b7de1..a27da90794 100644 --- a/alib2/src/grammar/ContextSensitive/CSG.cpp +++ b/alib2/src/grammar/ContextSensitive/CSG.cpp @@ -102,9 +102,8 @@ bool CSG::addRule(const std::vector<alphabet::Symbol>& leftHandSide, const std:: for(const alphabet::Symbol& symbol : rightHandSide) { if(terminalAlphabet.find(symbol) == terminalAlphabet.end() && nonterminalAlphabet.find(symbol) == nonterminalAlphabet.end()) throw GrammarException("Symbol \"" + (std::string) symbol.getSymbol() + "\" is not neither terminal nor nonterminal symbol"); - if(generatesEpsilon && symbol == initialSymbol) { + if(generatesEpsilon && symbol == initialSymbol) throw GrammarException("Initial symbol is already allowed to be rewritten to epsilon"); - } } return rules[leftHandSide].insert(rightHandSide).second; diff --git a/alib2/src/grammar/ContextSensitive/NonContractingGrammar.cpp b/alib2/src/grammar/ContextSensitive/NonContractingGrammar.cpp new file mode 100644 index 0000000000..a1d73a8a5c --- /dev/null +++ b/alib2/src/grammar/ContextSensitive/NonContractingGrammar.cpp @@ -0,0 +1,132 @@ +/* + * NonContractingGrammar.cpp + * + * Created on: Nov 17, 2013 + * Author: Jan Travnicek + */ + +#include "NonContractingGrammar.h" +#include "../../std/map.hpp" +#include "../GrammarException.h" +#include <algorithm> + +#include "../../alphabet/Symbol.h" + +namespace grammar { + +NonContractingGrammar::NonContractingGrammar(const alphabet::Symbol& initialSymbol) : TerminalNonterminalAlphabetInitialSymbol(initialSymbol), generatesEpsilon(false) { + +} + +NonContractingGrammar::NonContractingGrammar(const std::set<alphabet::Symbol>& nonterminalAlphabet, const std::set<alphabet::Symbol>& terminalAlphabet, const alphabet::Symbol& initialSymbol) : TerminalNonterminalAlphabetInitialSymbol(initialSymbol), generatesEpsilon(false) { + setNonterminalAlphabet(nonterminalAlphabet); + setTerminalAlphabet(terminalAlphabet); +} + +GrammarBase* NonContractingGrammar::clone() const { + return new NonContractingGrammar(*this); +} + +GrammarBase* NonContractingGrammar::plunder() && { + return new NonContractingGrammar(std::move(*this)); +} + +bool NonContractingGrammar::removeTerminalSymbol(const alphabet::Symbol& symbol) { + for(const std::pair<std::vector<alphabet::Symbol>, std::set<std::vector<alphabet::Symbol>>>& rule : rules) { + if(std::find(rule.first.begin(), rule.first.end(), symbol) != rule.first.end()) + throw GrammarException("Symbol \"" + (std::string) symbol.getSymbol() + "\" is used in rule."); + + for(const std::vector<alphabet::Symbol>& rhs : rule.second) + if(std::find(rhs.begin(), rhs.end(), symbol) != rhs.end()) + throw GrammarException("Symbol \"" + (std::string) symbol.getSymbol() + "\" is used in rule."); + } + + return terminalAlphabet.erase(symbol); +} + +bool NonContractingGrammar::removeNonterminalSymbol(const alphabet::Symbol& symbol) { + for(const std::pair<std::vector<alphabet::Symbol>, std::set<std::vector<alphabet::Symbol>>>& rule : rules) { + if(std::find(rule.first.begin(), rule.first.end(), symbol) != rule.first.end()) + throw GrammarException("Symbol \"" + (std::string) symbol.getSymbol() + "\" is used in rule."); + + for(const std::vector<alphabet::Symbol>& rhs : rule.second) + if(std::find(rhs.begin(), rhs.end(), symbol) != rhs.end()) + throw GrammarException("Symbol \"" + (std::string) symbol.getSymbol() + "\" is used in rule."); + } + + if(initialSymbol == symbol) + throw GrammarException("Symbol \"" + (std::string) symbol.getSymbol() + "\" is initial symbol."); + + + return nonterminalAlphabet.erase(symbol); +} + +bool NonContractingGrammar::addRule(const std::vector<alphabet::Symbol>& leftHandSide, const std::vector<alphabet::Symbol>& rightHandSide) { + int lSize = leftHandSide.size(); + int rSize = rightHandSide.size(); + + if(lSize == 1 && leftHandSide[0] == initialSymbol && rSize == 0) { + for(const auto& rule : rules) { + for(const auto& ruleRHS : rule.second) { + if(any_of(ruleRHS.begin(), ruleRHS.end(), [&](const alphabet::Symbol& symbol) { return initialSymbol == symbol; })) { + throw GrammarException("Initial symbol " + (std::string) initialSymbol + "used on right hand side of already existing rule"); + } + } + } + + generatesEpsilon = true; + return rules[leftHandSide].insert(rightHandSide).second; + } else { + if(lSize > rSize) + throw GrammarException("Invalid size of right hand side of a rule"); + + if(std::all_of(leftHandSide.begin(), leftHandSide.end(), [&](const alphabet::Symbol symbol) {return !nonterminalAlphabet.count(symbol);})) + throw GrammarException("Rule must rewrite nonterminal symbol"); + + for(const alphabet::Symbol& symbol : leftHandSide) + if(terminalAlphabet.find(symbol) == terminalAlphabet.end() && nonterminalAlphabet.find(symbol) == nonterminalAlphabet.end()) + throw GrammarException("Symbol \"" + (std::string) symbol.getSymbol() + "\" is not neither terminal nor nonterminal symbol"); + + for(const alphabet::Symbol& symbol : rightHandSide) { + if(terminalAlphabet.find(symbol) == terminalAlphabet.end() && nonterminalAlphabet.find(symbol) == nonterminalAlphabet.end()) + throw GrammarException("Symbol \"" + (std::string) symbol.getSymbol() + "\" is not neither terminal nor nonterminal symbol"); + if(generatesEpsilon && symbol == initialSymbol) + throw GrammarException("Initial symbol is already allowed to be rewritten to epsilon"); + } + + return rules[leftHandSide].insert(rightHandSide).second; + } +} + +const std::map<std::vector<alphabet::Symbol>, std::set<std::vector<alphabet::Symbol>>> NonContractingGrammar::getRules() const { + return rules; +} + +bool NonContractingGrammar::removeRule(const std::vector<alphabet::Symbol>& leftHandSide, const std::vector<alphabet::Symbol>& rightHandSide) { + int lSize = leftHandSide.size(); + int rSize = rightHandSide.size(); + + if(lSize == 1 && leftHandSide[0] == initialSymbol && rSize == 0) { + generatesEpsilon = false; + } + return rules[leftHandSide].erase(rightHandSide); +} + +bool NonContractingGrammar::operator==(const GrammarBase& other) const { + return other == *this; +} + +bool NonContractingGrammar::operator==(const NonContractingGrammar& other) const { + return this->nonterminalAlphabet == other.nonterminalAlphabet && this->terminalAlphabet == other.terminalAlphabet && this->initialSymbol == other.initialSymbol && this->rules == other.rules; +} + +void NonContractingGrammar::operator>>(std::ostream& out) const { + out << "(NonContractingGrammar" + << "nonterminalAlphabet = " << nonterminalAlphabet + << "terminalAlphabet = " << terminalAlphabet + << "initialSymbol = " << initialSymbol + << "rules = " << rules + << ")"; +} + +} /* namespace grammar */ diff --git a/alib2/src/grammar/ContextSensitive/NonContractingGrammar.h b/alib2/src/grammar/ContextSensitive/NonContractingGrammar.h new file mode 100644 index 0000000000..7cf45360ae --- /dev/null +++ b/alib2/src/grammar/ContextSensitive/NonContractingGrammar.h @@ -0,0 +1,51 @@ +/* + * NonContractingGrammar.h + * + * Created on: Nov 17, 2013 + * Author: Jan Travnicek + */ + +#ifndef NON_CONTRACTING_GRAMMAR_H_ +#define NON_CONTRACTING_GRAMMAR_H_ + +#include "../GrammarBase.h" +#include <map> +#include "../common/TerminalNonterminalAlphabetInitialSymbol.h" + +namespace grammar { + +/** + * Unrestricted grammar. Type 0 in Chomsky hierarchy. Produces recursively enumerable language. + */ +class NonContractingGrammar : public std::element<NonContractingGrammar, GrammarBase>, public TerminalNonterminalAlphabetInitialSymbol { + std::map<std::vector<alphabet::Symbol>, std::set<std::vector<alphabet::Symbol>>> rules; + bool generatesEpsilon; +public: + NonContractingGrammar(const alphabet::Symbol& initialSymbol); + + NonContractingGrammar(const std::set<alphabet::Symbol>& nonTerminalSymbols, const std::set<alphabet::Symbol>& terminalSymbols, const alphabet::Symbol& initialSymbol); + + virtual GrammarBase* clone() const; + + virtual GrammarBase* plunder() &&; + + bool addRule(const std::vector<alphabet::Symbol>& leftHandSide, const std::vector<alphabet::Symbol>& rightHandSide); + + const std::map<std::vector<alphabet::Symbol>, std::set<std::vector<alphabet::Symbol>>> getRules() const; + + bool removeRule(const std::vector<alphabet::Symbol>& leftHandSide, const std::vector<alphabet::Symbol>& rightHandSide); + + bool removeTerminalSymbol(const alphabet::Symbol& symbol); + + bool removeNonterminalSymbol(const alphabet::Symbol& symbol); + + virtual bool operator==(const GrammarBase& other) const; + + virtual bool operator==(const NonContractingGrammar& other) const; + + virtual void operator>>(std::ostream& os) const; +}; + +} /* namespace grammar */ + +#endif /* NON_CONTRACTING_GRAMMAR_H_ */ diff --git a/alib2/src/grammar/GrammarToXMLComposer.h b/alib2/src/grammar/GrammarToXMLComposer.h index c0433201cc..3c03188448 100644 --- a/alib2/src/grammar/GrammarToXMLComposer.h +++ b/alib2/src/grammar/GrammarToXMLComposer.h @@ -21,8 +21,8 @@ class CFG; class EpsilonFreeCFG; class CNF; class GNF; -class CSG; -class NonContractingGrammar; +#include "ContextSensitive/CSG.h" +#include "ContextSensitive/NonContractingGrammar.h" #include "Unrestricted/ContextPreservingUnrestrictedGrammar.h" #include "Unrestricted/UnrestrictedGrammar.h" #include "../sax/Token.h" -- GitLab