Parsing for inline assembly.

3 files changed, 337 insertions, 0 deletions

diff --git a/libsolidity/inlineasm/AsmData.h b/libsolidity/inlineasm/AsmData.h
new file mode 100644
index 00000000..a38a9d36
--- /dev/null
+++ b/libsolidity/inlineasm/AsmData.h
@@ -0,0 +1,70 @@
+/*
+	This file is part of cpp-ethereum.
+
+	cpp-ethereum is free software: you can redistribute it and/or modify
+	it under the terms of the GNU General Public License as published by
+	the Free Software Foundation, either version 3 of the License, or
+	(at your option) any later version.
+
+	cpp-ethereum is distributed in the hope that it will be useful,
+	but WITHOUT ANY WARRANTY; without even the implied warranty of
+	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+	GNU General Public License for more details.
+
+	You should have received a copy of the GNU General Public License
+	along with cpp-ethereum.  If not, see <http://www.gnu.org/licenses/>.
+*/
+/**
+ * @author Christian <c@ethdev.com>
+ * @date 2016
+ * Parsed inline assembly to be used by the AST
+ */
+
+#pragma once
+
+#include <boost/variant.hpp>
+#include <libevmcore/Instruction.h>
+
+namespace dev
+{
+namespace solidity
+{
+
+class AsmData
+{
+public:
+	/// Direct EVM instruction (except PUSHi and JUMPDEST)
+	struct Instruction { eth::Instruction instruction; };
+	/// Literal number or string (up to 32 bytes)
+	struct Literal { bool isNumber; std::string value; };
+	/// External / internal identifier or label reference
+	struct Identifier { std::string name; };
+	struct FunctionalInstruction;
+	/// Jump label ("name:")
+	struct Label { std::string name; };
+	/// Assignemnt (":= x", moves stack top into x, potentially multiple slots)
+	struct Assignment { Identifier variableName; };
+	struct FunctionalAssignment;
+	struct VariableDeclaration;
+	struct Block;
+	using Statement = boost::variant<Instruction, Literal, Label, Assignment, Identifier, FunctionalAssignment, FunctionalInstruction, VariableDeclaration, Block>;
+	/// Functional assignment ("x := mload(20)", expects push-1-expression on the right hand
+	/// side and requires x to occupy exactly one stack slot.
+	struct FunctionalAssignment { Identifier variableName; std::shared_ptr<Statement> value; };
+	/// Functional instruction, e.g. "mul(mload(20), add(2, x))"
+	struct FunctionalInstruction { Instruction instruction; std::vector<Statement> arguments; };
+	/// Block-scope variable declaration ("let x := mload(20)"), non-hoisted
+	struct VariableDeclaration { std::string name; std::shared_ptr<Statement> value; };
+	/// Block that creates a scope (frees declared stack variables)
+	struct Block { std::vector<Statement> statements; };
+
+	AsmData(Block&& _statements): m_statements(_statements) {}
+
+	Block const& statements() const { return m_statements; }
+
+private:
+	Block m_statements;
+};
+
+}
+}
diff --git a/libsolidity/inlineasm/AsmParser.cpp b/libsolidity/inlineasm/AsmParser.cpp
new file mode 100644
index 00000000..28fd5354
--- /dev/null
+++ b/libsolidity/inlineasm/AsmParser.cpp
@@ -0,0 +1,212 @@
+/*
+	This file is part of cpp-ethereum.
+
+	cpp-ethereum is free software: you can redistribute it and/or modify
+	it under the terms of the GNU General Public License as published by
+	the Free Software Foundation, either version 3 of the License, or
+	(at your option) any later version.
+
+	cpp-ethereum is distributed in the hope that it will be useful,
+	but WITHOUT ANY WARRANTY; without even the implied warranty of
+	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+	GNU General Public License for more details.
+
+	You should have received a copy of the GNU General Public License
+	along with cpp-ethereum.  If not, see <http://www.gnu.org/licenses/>.
+*/
+/**
+ * @author Christian <c@ethdev.com>
+ * @date 2016
+ * Solidity inline assembly parser.
+ */
+
+#include <libsolidity/inlineasm/AsmParser.h>
+#include <ctype.h>
+#include <algorithm>
+#include <libevmcore/Instruction.h>
+#include <libsolidity/parsing/Scanner.h>
+
+using namespace std;
+using namespace dev;
+using namespace dev::solidity;
+
+shared_ptr<AsmData> InlineAssemblyParser::parse(std::shared_ptr<Scanner> const& _scanner)
+{
+	try
+	{
+		m_scanner = _scanner;
+		return make_shared<AsmData>(parseBlock());
+	}
+	catch (FatalError const&)
+	{
+		if (m_errors.empty())
+			throw; // Something is weird here, rather throw again.
+	}
+	return nullptr;
+}
+
+AsmData::Block InlineAssemblyParser::parseBlock()
+{
+	expectToken(Token::LBrace);
+	AsmData::Block block;
+	while (m_scanner->currentToken() != Token::RBrace)
+		block.statements.emplace_back(parseStatement());
+	m_scanner->next();
+	return block;
+}
+
+AsmData::Statement InlineAssemblyParser::parseStatement()
+{
+	switch (m_scanner->currentToken())
+	{
+	case Token::Let:
+		return parseVariableDeclaration();
+	case Token::LBrace:
+		return parseBlock();
+	case Token::Assign:
+	{
+		m_scanner->next();
+		expectToken(Token::Colon);
+		string name = m_scanner->currentLiteral();
+		expectToken(Token::Identifier);
+		return AsmData::Assignment{AsmData::Identifier{name}};
+	}
+	default:
+		break;
+	}
+	// Options left:
+	// Simple instruction (might turn into functional),
+	// literal,
+	// identifier (might turn into label or functional assignment)
+	AsmData::Statement statement(parseElementaryOperation());
+	switch (m_scanner->currentToken())
+	{
+	case Token::LParen:
+		return parseFunctionalInstruction(statement);
+	case Token::Colon:
+	{
+		if (statement.type() != typeid(AsmData::Identifier))
+			fatalParserError("Label name / variable name must precede \":\".");
+		string const& name = boost::get<AsmData::Identifier>(statement).name;
+		m_scanner->next();
+		if (m_scanner->currentToken() == Token::Assign)
+		{
+			// functional assignment
+			m_scanner->next();
+			unique_ptr<AsmData::Statement> value;
+			value.reset(new AsmData::Statement(parseExpression()));
+			return AsmData::FunctionalAssignment{{move(name)}, move(value)};
+		}
+		else
+			// label
+			return AsmData::Label{name};
+	}
+	default:
+		break;
+	}
+	return statement;
+}
+
+AsmData::Statement InlineAssemblyParser::parseExpression()
+{
+	AsmData::Statement operation = parseElementaryOperation(true);
+	if (m_scanner->currentToken() == Token::LParen)
+		return parseFunctionalInstruction(operation);
+	else
+		return operation;
+}
+
+AsmData::Statement InlineAssemblyParser::parseElementaryOperation(bool _onlySinglePusher)
+{
+	// Allowed instructions, lowercase names.
+	static map<string, eth::Instruction> s_instructions;
+	if (s_instructions.empty())
+		for (auto const& instruction: eth::c_instructions)
+		{
+			if (
+				instruction.second == eth::Instruction::JUMPDEST ||
+				(eth::Instruction::PUSH1 <= instruction.second && instruction.second <= eth::Instruction::PUSH32)
+			)
+				continue;
+			string name = instruction.first;
+			transform(name.begin(), name.end(), name.begin(), [](unsigned char _c) { return tolower(_c); });
+			s_instructions[name] = instruction.second;
+		}
+
+	//@TODO track location
+
+	switch (m_scanner->currentToken())
+	{
+	case Token::Identifier:
+	{
+		string literal = m_scanner->currentLiteral();
+		// first search the set of instructions.
+		if (s_instructions.count(literal))
+		{
+			eth::Instruction const& instr = s_instructions[literal];
+			if (_onlySinglePusher)
+			{
+				eth::InstructionInfo info = eth::instructionInfo(instr);
+				if (info.ret != 1)
+					fatalParserError("Instruction " + info.name + " not allowed in this context.");
+			}
+			m_scanner->next();
+			return AsmData::Instruction{instr};
+		}
+		else
+			m_scanner->next();
+			return AsmData::Identifier{literal};
+		break;
+	}
+	case Token::StringLiteral:
+	case Token::Number:
+	{
+		AsmData::Literal literal{
+			m_scanner->currentToken() == Token::Number,
+			m_scanner->currentLiteral()
+		};
+		m_scanner->next();
+		return literal;
+	}
+	default:
+		break;
+	}
+	fatalParserError("Expected elementary inline assembly operation.");
+	return {};
+}
+
+AsmData::VariableDeclaration InlineAssemblyParser::parseVariableDeclaration()
+{
+	expectToken(Token::Let);
+	string name = m_scanner->currentLiteral();
+	expectToken(Token::Identifier);
+	expectToken(Token::Colon);
+	expectToken(Token::Assign);
+	unique_ptr<AsmData::Statement> value;
+	value.reset(new AsmData::Statement(parseExpression()));
+	return AsmData::VariableDeclaration{name, move(value)};
+}
+
+AsmData::FunctionalInstruction InlineAssemblyParser::parseFunctionalInstruction(AsmData::Statement const& _instruction)
+{
+	if (_instruction.type() != typeid(AsmData::Instruction))
+		fatalParserError("Assembly instruction required in front of \"(\")");
+	eth::Instruction instr = boost::get<AsmData::Instruction>(_instruction).instruction;
+	eth::InstructionInfo instrInfo = eth::instructionInfo(instr);
+	if (eth::Instruction::DUP1 <= instr && instr <= eth::Instruction::DUP16)
+		fatalParserError("DUPi instructions not allowed for functional notation");
+	if (eth::Instruction::SWAP1 <= instr && instr <= eth::Instruction::SWAP16)
+		fatalParserError("SWAPi instructions not allowed for functional notation");
+
+	expectToken(Token::LParen);
+	vector<AsmData::Statement> arguments;
+	unsigned args = unsigned(instrInfo.args);
+	for (unsigned i = 0; i < args; ++i)
+	{
+		arguments.push_back(parseExpression());
+		if (i != args - 1)
+			expectToken(Token::Comma);
+	}
+	expectToken(Token::RParen);
+	return AsmData::FunctionalInstruction{{instr}, move(arguments)};
+}
diff --git a/libsolidity/inlineasm/AsmParser.h b/libsolidity/inlineasm/AsmParser.h
new file mode 100644
index 00000000..fe84470d
--- /dev/null
+++ b/libsolidity/inlineasm/AsmParser.h
@@ -0,0 +1,55 @@
+/*
+	This file is part of cpp-ethereum.
+
+	cpp-ethereum is free software: you can redistribute it and/or modify
+	it under the terms of the GNU General Public License as published by
+	the Free Software Foundation, either version 3 of the License, or
+	(at your option) any later version.
+
+	cpp-ethereum is distributed in the hope that it will be useful,
+	but WITHOUT ANY WARRANTY; without even the implied warranty of
+	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+	GNU General Public License for more details.
+
+	You should have received a copy of the GNU General Public License
+	along with cpp-ethereum.  If not, see <http://www.gnu.org/licenses/>.
+*/
+/**
+ * @author Christian <c@ethdev.com>
+ * @date 2016
+ * Solidity inline assembly parser.
+ */
+
+#pragma once
+
+#include <memory>
+#include <vector>
+#include <libsolidity/inlineasm/AsmData.h>
+#include <libsolidity/parsing/ParserBase.h>
+
+namespace dev
+{
+namespace solidity
+{
+
+class InlineAssemblyParser: public ParserBase
+{
+public:
+	InlineAssemblyParser(ErrorList& _errors): ParserBase(_errors) {}
+
+	/// Parses an inline assembly block starting with `{` and ending with `}`.
+	/// @returns an empty shared pointer on error.
+	std::shared_ptr<AsmData> parse(std::shared_ptr<Scanner> const& _scanner);
+
+protected:
+	AsmData::Block parseBlock();
+	AsmData::Statement parseStatement();
+	/// Parses a functional expression that has to push exactly one stack element
+	AsmData::Statement parseExpression();
+	AsmData::Statement parseElementaryOperation(bool _onlySinglePusher = false);
+	AsmData::VariableDeclaration parseVariableDeclaration();
+	AsmData::FunctionalInstruction parseFunctionalInstruction(AsmData::Statement const& _instruction);
+};
+
+}
+}