Merge pull request #135 from Washi1337/feature/new-ast-builder

New AST builder.

README.md

@@ -14,19 +14,21 @@ Main Features
 - Data flow analysis
     - Create data flow graphs
     - Inspect stack and variable dependencies of instructions.
+- AST building
+    - Lift control flow graphs to Abstract Syntax Trees (ASTs).
 - Unified generic API.
     - Serialize any kind of graph to the dot file format.
     - Adding a new platform for flow analysis requires minimal effort
 
 
 Supported platforms:
 
-| Architecture | Back-end                                                | Control Flow | Data Flow | Purity Classification | Emulation |
-|--------------|---------------------------------------------------------|--------------|-----------|-----------------------|-----------|
-| CIL          | [AsmResolver](https://github.com/Washi1337/AsmResolver) | ✓            | ✓         | ✓                     | ✓ (WIP)   |
-| CIL          | [dnlib](https://github.com/0xd4d/dnlib)                 | ✓            | ✓         | ✓                     |           |
-| x86 (32-bit) | [Iced](https://github.com/icedland/iced)                | ✓            | ✓         |                       |           |
-| x86 (64-bit) | [Iced](https://github.com/icedland/iced)                | ✓            | ✓         |                       |           |
+| Architecture | Back-end                                                | Control Flow | Data Flow | AST | Purity Classification | Emulation |
+|--------------|---------------------------------------------------------|--------------|-----------|-----|-----------------------|-----------|
+| CIL          | [AsmResolver](https://github.com/Washi1337/AsmResolver) | ✓            | ✓         | ✓   | ✓                     | ✓ (WIP)   |
+| CIL          | [dnlib](https://github.com/0xd4d/dnlib)                 | ✓            | ✓         | ✓   | ✓                     |           |
+| x86 (32-bit) | [Iced](https://github.com/icedland/iced)                | ✓            | ✓         | ✓   |                       |           |
+| x86 (64-bit) | [Iced](https://github.com/icedland/iced)                | ✓            | ✓         | ✓   |                       |           |
 
 
 Compiling
@@ -54,5 +56,4 @@ See [CONTRIBUTING.md](CONTRIBUTING.md).
 
 Found a bug or have questions?
 ------------------------------
-Please use the [issue tracker](https://github.com/Washi1337/Echo/issues). Try to be as descriptive as possible.
-
+Please use the [issue tracker](https://github.com/Washi1337/Echo/issues). Try to be as descriptive as possible.

src/Core/Echo.Ast/Analysis/AstPurityClassifier.cs

@@ -0,0 +1,34 @@
+using Echo.Code;
+
+namespace Echo.Ast.Analysis;
+
+/// <summary>
+/// Provides a wrapper around a <see cref="IPurityClassifier{TInstruction}"/> that is able to classify statements
+/// and expressions with <typeparamref name="TInstruction"/> instructions by purity. 
+/// </summary>
+/// <typeparam name="TInstruction">The type of instructions the statements store.</typeparam>
+public class AstPurityClassifier<TInstruction> : IPurityClassifier<Statement<TInstruction>>
+{
+    /// <summary>
+    /// Creates a new instance of the <see cref="AstPurityClassifier{TInstruction}"/> class.
+    /// </summary>
+    /// <param name="baseClassifier">The base classifier to use for classifying individual instructions in the AST.</param>
+    public AstPurityClassifier(IPurityClassifier<TInstruction> baseClassifier)
+    {
+        BaseClassifier = baseClassifier;
+    }
+
+    /// <summary>
+    /// Gets the base classifier to use for classifying individual instructions in the AST.
+    /// </summary>
+    public IPurityClassifier<TInstruction> BaseClassifier
+    {
+        get;
+    }
+
+    /// <inheritdoc />
+    public Trilean IsPure(in Statement<TInstruction> instruction)
+    {
+        return instruction.Accept(AstPurityVisitor<TInstruction>.Instance, BaseClassifier);
+    }
+}

src/Core/Echo.Ast/Analysis/AstPurityVisitor.cs

@@ -0,0 +1,89 @@
+using Echo.Code;
+
+namespace Echo.Ast.Analysis;
+
+/// <summary>
+/// Provides a mechanism for traversing an AST and determining its purity.  
+/// </summary>
+/// <typeparam name="TInstruction">The type of instructions to store in each expression.</typeparam>
+public class AstPurityVisitor<TInstruction> : IAstNodeVisitor<TInstruction, IPurityClassifier<TInstruction>, Trilean>
+{
+    /// <summary>
+    /// Gets the singleton instance of the <see cref="AstPurityVisitor{TInstruction}"/> class.
+    /// </summary>
+    public static AstPurityVisitor<TInstruction> Instance
+    {
+        get;
+    } = new();
+
+    /// <inheritdoc />
+    public Trilean Visit(CompilationUnit<TInstruction> unit, IPurityClassifier<TInstruction> state)
+    {
+        return unit.Root.Accept(this, state);
+    }
+
+    /// <inheritdoc />
+    public Trilean Visit(AssignmentStatement<TInstruction> statement, IPurityClassifier<TInstruction> state) => false;
+
+    /// <inheritdoc />
+    public Trilean Visit(ExpressionStatement<TInstruction> statement, IPurityClassifier<TInstruction> state)
+    {
+        return statement.Expression.Accept(this, state);
+    }
+
+    /// <inheritdoc />
+    public Trilean Visit(PhiStatement<TInstruction> statement, IPurityClassifier<TInstruction> state) => false;
+
+    /// <inheritdoc />
+    public Trilean Visit(BlockStatement<TInstruction> statement, IPurityClassifier<TInstruction> state)
+    {
+        var result = Trilean.True;
+
+        for (int i = 0; i < statement.Statements.Count && result != Trilean.False; i++)
+            result &= statement.Statements[i].Accept(this, state);
+
+        return result;
+    }
+
+    /// <inheritdoc />
+    public Trilean Visit(ExceptionHandlerStatement<TInstruction> statement, IPurityClassifier<TInstruction> state)
+    {
+        var result = statement.ProtectedBlock.Accept(this, state);
+
+        for (int i = 0; i < statement.Handlers.Count && result != Trilean.False; i++)
+            result &= statement.Handlers[i].Accept(this, state);
+
+        return result;
+    }
+
+    /// <inheritdoc />
+    public Trilean Visit(HandlerClause<TInstruction> clause, IPurityClassifier<TInstruction> state)
+    {
+        var result = Trilean.True;
+
+        if (clause.Prologue is not null)
+            result &= clause.Prologue.Accept(this, state);
+
+        if (result.ToBooleanOrFalse())
+            result &= clause.Contents.Accept(this, state);
+
+        if (clause.Epilogue is not null && result.ToBooleanOrFalse())
+            result &= clause.Epilogue.Accept(this, state);
+
+        return result;
+    }
+
+    /// <inheritdoc />
+    public Trilean Visit(InstructionExpression<TInstruction> expression, IPurityClassifier<TInstruction> state)
+    {
+        var result = state.IsPure(expression.Instruction);
+
+        for (int i = 0; i < expression.Arguments.Count && result != Trilean.False; i++)
+            result &= expression.Arguments[i].Accept(this, state);
+
+        return result;
+    }
+
+    /// <inheritdoc />
+    public Trilean Visit(VariableExpression<TInstruction> expression, IPurityClassifier<TInstruction> state) => true;
+}

src/Core/Echo.Ast/Analysis/FlowControlDeterminer.cs

@@ -1,28 +1,49 @@
-using Echo.Code;
+using System;
+using Echo.Code;
 
 namespace Echo.Ast.Analysis
 {
     internal sealed class FlowControlDeterminer<TInstruction>
-        : IAstNodeVisitor<TInstruction, object, InstructionFlowControl>
+        : IAstNodeVisitor<TInstruction, object?, InstructionFlowControl>
     {
         private readonly IArchitecture<TInstruction> _isa;
 
         internal FlowControlDeterminer(IArchitecture<TInstruction> isa) =>
             _isa = isa;
 
-        public InstructionFlowControl Visit(AssignmentStatement<TInstruction> assignmentStatement, object state) =>
-            assignmentStatement.Expression.Accept(this, state);
+        public InstructionFlowControl Visit(CompilationUnit<TInstruction> unit, object? state) =>
+            throw new NotSupportedException();
 
-        public InstructionFlowControl Visit(ExpressionStatement<TInstruction> expressionStatement, object state) =>
-            expressionStatement.Expression.Accept(this, state);
+        public InstructionFlowControl Visit(AssignmentStatement<TInstruction> statement, object? state) =>
+            statement.Expression.Accept(this, state);
 
-        public InstructionFlowControl Visit(PhiStatement<TInstruction> phiStatement, object state) =>
+        public InstructionFlowControl Visit(ExpressionStatement<TInstruction> statement, object? state) =>
+            statement.Expression.Accept(this, state);
+
+        public InstructionFlowControl Visit(PhiStatement<TInstruction> statement, object? state) =>
             InstructionFlowControl.Fallthrough;
 
-        public InstructionFlowControl Visit(InstructionExpression<TInstruction> instructionExpression, object state) =>
-            _isa.GetFlowControl(instructionExpression.Instruction);
+        public InstructionFlowControl Visit(BlockStatement<TInstruction> statement, object? state)
+        {
+            return statement.Statements.Count > 0
+                ? statement.Statements[statement.Statements.Count - 1].Accept(this, state)
+                : InstructionFlowControl.Fallthrough;
+        }
+
+        public InstructionFlowControl Visit(ExceptionHandlerStatement<TInstruction> statement, object? state)
+        {
+            return statement.ProtectedBlock.Accept(this, state);
+        }
+
+        public InstructionFlowControl Visit(HandlerClause<TInstruction> clause, object? state)
+        {
+            throw new NotSupportedException();
+        }
+
+        public InstructionFlowControl Visit(InstructionExpression<TInstruction> expression, object? state) =>
+            _isa.GetFlowControl(expression.Instruction);
 
-        public InstructionFlowControl Visit(VariableExpression<TInstruction> variableExpression, object state) =>
+        public InstructionFlowControl Visit(VariableExpression<TInstruction> expression, object? state) =>
             InstructionFlowControl.Fallthrough;
     }
 }

src/Core/Echo.Ast/Analysis/ReadVariableFinder.cs

@@ -0,0 +1,41 @@
+using System.Buffers;
+using System.Collections.Generic;
+using Echo.Code;
+
+namespace Echo.Ast.Analysis
+{
+    internal sealed class ReadVariableFinder<TInstruction> : AstNodeListener<TInstruction>
+    {
+        private readonly IArchitecture<TInstruction> _architecture;
+
+        public ReadVariableFinder(IArchitecture<TInstruction> architecture)
+        {
+            _architecture = architecture;
+        }
+
+        internal HashSet<IVariable> Variables { get; } = new();
+
+        public override void ExitVariableExpression(VariableExpression<TInstruction> expression)
+        {
+            base.ExitVariableExpression(expression);
+            Variables.Add(expression.Variable);
+        }
+
+        public override void ExitInstructionExpression(InstructionExpression<TInstruction> expression)
+        {
+            int count = _architecture.GetReadVariablesCount(expression.Instruction);
+            if (count == 0)
+                return;
+
+            var variables = ArrayPool<IVariable>.Shared.Rent(count);
+
+            int actualCount = _architecture.GetReadVariables(expression.Instruction, variables);
+            for (int i = 0; i < actualCount; i++)
+                Variables.Add(variables[i]);
+
+            ArrayPool<IVariable>.Shared.Return(variables);
+
+            base.ExitInstructionExpression(expression);
+        }
+    }
+}

src/Core/Echo.Ast/Analysis/WrittenVariableFinder.cs

@@ -0,0 +1,48 @@
+using System.Buffers;
+using System.Collections.Generic;
+using Echo.Code;
+
+namespace Echo.Ast.Analysis
+{
+    internal sealed class WrittenVariableFinder<TInstruction> : AstNodeListener<TInstruction>
+    {
+        private readonly IArchitecture<TInstruction> _architecture;
+
+        public WrittenVariableFinder(IArchitecture<TInstruction> architecture)
+        {
+            _architecture = architecture;
+        }
+
+        internal HashSet<IVariable> Variables { get; } = new();
+
+        public override void ExitAssignmentStatement(AssignmentStatement<TInstruction> statement)
+        {
+            base.ExitAssignmentStatement(statement);
+            for (int i = 0; i < statement.Variables.Count; i++)
+                Variables.Add(statement.Variables[i]);
+        }
+
+        public override void ExitPhiStatement(PhiStatement<TInstruction> phiStatement)
+        {
+            base.ExitPhiStatement(phiStatement);
+            Variables.Add(phiStatement.Representative);
+        }
+
+        public override void ExitInstructionExpression(InstructionExpression<TInstruction> expression)
+        {
+            int count = _architecture.GetWrittenVariablesCount(expression.Instruction);
+            if (count == 0)
+                return;
+
+            var variables = ArrayPool<IVariable>.Shared.Rent(count);
+
+            int actualCount = _architecture.GetWrittenVariables(expression.Instruction, variables);
+            for (int i = 0; i < actualCount; i++)
+                Variables.Add(variables[i]);
+
+            ArrayPool<IVariable>.Shared.Return(variables);
+
+            base.ExitInstructionExpression(expression);
+        }
+    }
+}