From 7ed6b1f33cee13a84f6cb77f7840bae48337adb5 Mon Sep 17 00:00:00 2001
From: IlyaShurupov <ilusha.main@gmail.com>
Date: Tue, 31 Oct 2023 22:43:37 +0300
Subject: [PATCH] tmp

---
 Language/private/SimpleParser.cpp |  11 +-
 Language/public/Grammar.hpp       | 162 +++++++++++++++++++++++++++++-
 2 files changed, 167 insertions(+), 6 deletions(-)

diff --git a/Language/private/SimpleParser.cpp b/Language/private/SimpleParser.cpp
index 213b789..1c8b16f 100644
--- a/Language/private/SimpleParser.cpp
+++ b/Language/private/SimpleParser.cpp
@@ -1,4 +1,6 @@
 
+#include "NewPlacement.hpp"
+
 #include "SimpleParser.hpp"
 
 using namespace tp;
@@ -22,14 +24,13 @@ SimpleParser::SimpleParser() {
 	}
 
 	// Define Regular grammar
-	RegularGrammar regularGrammar;
+	RegularGrammar<uint1, int1> rg;
 	{
 		// this is basically ast from existing tokenizer
-		regularGrammar.addRule(regularGrammar.alternate(regularGrammar.repeat(regularGrammar.symbols("asd")), regularGrammar.symbol("a")));
-		regularGrammar.addRule(regularGrammar.alternate(regularGrammar.repeat(regularGrammar.symbols("asd")), regularGrammar.symbol("a")));
+		rg.addRule(rg.seq({ rg.val('a'), rg.val('b') }), 0);
 	}
 
-	mUnifiedGrammarParser.compileTables(contextFreeGrammar, regularGrammar);
+	mUnifiedGrammarParser.compileTables(contextFreeGrammar, rg);
 }
 
 void SimpleParser::compileTables(const Sentence& grammar) {
@@ -38,7 +39,7 @@ void SimpleParser::compileTables(const Sentence& grammar) {
 
 	// compile each ast into RegularGrammar and ContextFree Grammar api instructions
 	ContextFreeGrammar userContextFreeGrammar;
-	RegularGrammar userRegularGrammar;
+	RegularGrammar<uint1, int1> userRegularGrammar;
 
 	// ...
 	// split ast into RE and CF part
diff --git a/Language/public/Grammar.hpp b/Language/public/Grammar.hpp
index e8bd32f..5f7984f 100644
--- a/Language/public/Grammar.hpp
+++ b/Language/public/Grammar.hpp
@@ -1,9 +1,169 @@
 
 #pragma once
 
+#include "Buffer.hpp"
 #include "LanguageCommon.hpp"
+#include "Utils.hpp"
 
 namespace tp {
+
 	class ContextFreeGrammar {};
-	class RegularGrammar {};
+
+	template <typename tAlphabetType, typename tTokType>
+	class RegularGrammar {
+
+		struct Node {
+			enum Type {
+				NONE,
+				ANY,
+				OR,
+				IF,
+				CLASS,
+				COMPOUND,
+				REPEAT,
+				VAL,
+			} mType = NONE;
+
+			explicit Node(Type type) :
+				mType(type) {}
+
+			virtual ~Node() = default;
+		};
+
+		class ValueNode : public Node {
+		public:
+			explicit ValueNode(tAlphabetType val) :
+				mVal(val),
+				Node(Node::VAL) {}
+
+			~ValueNode() override = default;
+
+		private:
+			tAlphabetType mVal;
+		};
+
+		class CompoundNode : public Node {
+		public:
+			CompoundNode() :
+				Node(Node::COMPOUND) {}
+
+			CompoundNode(const InitialierList<const Node*>& nodes) :
+				Node(Node::COMPOUND) {
+				mSequence = nodes;
+			}
+
+			~CompoundNode() override {
+				for (auto iter : mSequence) {
+					delete iter.data();
+				}
+				mSequence.clear();
+			}
+
+		private:
+			Buffer<const Node*> mSequence;
+		};
+
+		class AlternationNode : public Node {
+		public:
+			AlternationNode() :
+				Node(Node::OR) {}
+
+			AlternationNode(const Node* a, const Node* b) :
+				Node(Node::OR) {
+				mFirst = a;
+				mSecond = b;
+			}
+
+			~AlternationNode() override {
+				delete mFirst;
+				delete mSecond;
+			}
+
+		private:
+			const Node* mFirst = nullptr;
+			const Node* mSecond = nullptr;
+		};
+
+		class IfNode : public Node {
+		public:
+			IfNode() :
+				Node(Node::IF) {}
+
+			explicit IfNode(const Node* a) :
+				Node(Node::IF) {
+				mNode = a;
+			}
+
+			~IfNode() override { delete mNode; }
+
+		private:
+			const Node* mNode = nullptr;
+		};
+
+		class AnyNode : public Node {
+		public:
+			AnyNode() :
+				Node(Node::ANY) {}
+
+			~AnyNode() override = default;
+		};
+
+		class RepetitionNode : public Node {
+		public:
+			RepetitionNode() :
+				Node(Node::REPEAT) {}
+
+			explicit RepetitionNode(const Node* rep, bool plus = false) :
+				Node(Node::REPEAT) {
+				mNode = rep;
+				mPlus = plus;
+			}
+
+			~RepetitionNode() override { delete mNode; }
+
+		private:
+			Node* mNode = nullptr;
+			bool mPlus = false;
+		};
+
+		class ClassNode : public Node {
+		public:
+			ClassNode() :
+				Node(Node::CLASS) {}
+
+			explicit ClassNode(const Buffer<Range<tAlphabetType>>& ranges, bool exclude = false) :
+				Node(Node::CLASS) {
+				mExclude = exclude;
+			}
+
+			~ClassNode() override { mRanges.removeAll(); }
+
+		private:
+			Buffer<Range<tAlphabetType>> mRanges;
+			bool mExclude = false;
+		};
+
+	public:
+		RegularGrammar() = default;
+
+		~RegularGrammar() {
+			for (auto rule : mRules) {
+				delete rule->t1;
+			}
+		}
+
+		void addRule(const Node* node, tTokType id) { mRules.append({ node, id }); }
+
+	public:
+		const Node* seq(const InitialierList<const Node*>& nodes) { return new CompoundNode(nodes); }
+		const Node* val(tAlphabetType in) { return new ValueNode(in); }
+		const Node* alt(const Node* a, const Node* b) { return new AlternationNode(a, b); }
+		const Node* may(const Node* a) { return new IfNode(a); }
+		const Node* any() { return new AnyNode(); }
+		const Node* rep(const Node* rep, bool plus = false) { return new RepetitionNode(rep, plus); }
+		const Node* ranges(const Buffer<Range<tAlphabetType>>& ranges, bool exclude = false) { return new ClassNode(ranges, exclude); }
+
+	private:
+		Buffer<Pair<const Node*, tTokType>> mRules;
+	};
 }