[cdac] Add helper methods to Target for reading UTF-8/16 strings

docs/design/datacontracts/RuntimeTypeSystem.md

@@ -125,7 +125,7 @@ partial interface IRuntimeTypeSystem : IContract
     // generated by the runtime, or a MethodDesc that describes a method represented by the System.Reflection.Emit.DynamicMethod class
     // Or something else similar.
     // A no metadata method is also a StoredSigMethodDesc
-    public virtual bool IsNoMetadataMethod(MethodDescHandle methodDesc, out ReadOnlySpan<byte> methodName);
+    public virtual bool IsNoMetadataMethod(MethodDescHandle methodDesc, out string methodName);
 
     // A StoredSigMethodDesc is a MethodDesc for which the signature isn't found in metadata.
     public virtual bool IsStoredSigMethodDesc(MethodDescHandle methodDesc, out ReadOnlySpan<byte> signature);
@@ -708,9 +708,9 @@ And the various apis are implemented with the following algorithms
     public uint GetMethodToken(MethodDescHandle methodDescHandle)
     {
         MethodDesc methodDesc = _methodDescs[methodDescHandle.Address];
-        
+
         TargetPointer methodDescChunk = // Using ChunkIndex from methodDesc, compute the wrapping MethodDescChunk
-        
+
         ushort Flags3AndTokenRemainder = // Read Flags3AndTokenRemainder field from MethodDesc contract using address methodDescHandle.Address
 
         ushort FlagsAndTokenRange = // Read FlagsAndTokenRange field from MethodDescChunk contract using address methodDescChunk

src/native/managed/cdacreader/src/Contracts/Loader_1.cs

@@ -2,8 +2,6 @@
 // The .NET Foundation licenses this file to you under the MIT license.
 
 using System;
-using System.Collections.Generic;
-using System.Runtime.InteropServices;
 
 namespace Microsoft.Diagnostics.DataContractReader.Contracts;
 
@@ -39,26 +37,7 @@ ModuleFlags ILoader.GetFlags(ModuleHandle handle)
     string ILoader.GetPath(ModuleHandle handle)
     {
         Data.Module module = _target.ProcessedData.GetOrAdd<Data.Module>(handle.Address);
-
-        // TODO: [cdac] Add/use APIs on Target for reading strings in target endianness
-        TargetPointer addr = module.Path;
-        while (true)
-        {
-            // Read characters until we find the null terminator
-            char nameChar = _target.Read<char>(addr);
-            if (nameChar == 0)
-                break;
-
-            addr += sizeof(char);
-        }
-
-        int length = (int)(addr.Value - module.Path.Value);
-        if (length == 0)
-            return string.Empty;
-
-        Span<byte> span = stackalloc byte[length];
-        _target.ReadBuffer(module.Path, span);
-        return new string(MemoryMarshal.Cast<byte, char>(span));
+        return _target.ReadUtf16String(module.Path);
     }
 
     TargetPointer ILoader.GetLoaderAllocator(ModuleHandle handle)

src/native/managed/cdacreader/src/Contracts/RuntimeTypeSystem.cs

@@ -2,7 +2,6 @@
 // The .NET Foundation licenses this file to you under the MIT license.
 
 using System;
-using System.Reflection.Metadata.Ecma335;
 
 namespace Microsoft.Diagnostics.DataContractReader.Contracts;
 
@@ -148,11 +147,11 @@ static IContract IContract.Create(Target target, int version)
     // An array method is also a StoredSigMethodDesc
     public virtual bool IsArrayMethod(MethodDescHandle methodDesc, out ArrayFunctionType functionType) => throw new NotImplementedException();
 
-    // Return true if a MethodDesc represents a method without metadata method, either an IL Stub dynamically
+    // Return true if a MethodDesc represents a method without metadata, either an IL Stub dynamically
     // generated by the runtime, or a MethodDesc that describes a method represented by the System.Reflection.Emit.DynamicMethod class
     // Or something else similar.
     // A no metadata method is also a StoredSigMethodDesc
-    public virtual bool IsNoMetadataMethod(MethodDescHandle methodDesc, out ReadOnlySpan<byte> methodName) => throw new NotImplementedException();
+    public virtual bool IsNoMetadataMethod(MethodDescHandle methodDesc, out string methodName) => throw new NotImplementedException();
     // A StoredSigMethodDesc is a MethodDesc for which the signature isn't found in metadata.
     public virtual bool IsStoredSigMethodDesc(MethodDescHandle methodDesc, out ReadOnlySpan<byte> signature) => throw new NotImplementedException();
 

src/native/managed/cdacreader/src/Contracts/RuntimeTypeSystem_1.cs

@@ -3,12 +3,10 @@
 
 using System;
 using System.Collections.Generic;
+using System.Diagnostics;
 using System.Reflection.Metadata.Ecma335;
-using Microsoft.Diagnostics.DataContractReader.Data;
 using Microsoft.Diagnostics.DataContractReader.Contracts.RuntimeTypeSystem_1_NS;
-using System.Diagnostics;
-using System.Text;
-using System.Reflection;
+using Microsoft.Diagnostics.DataContractReader.Data;
 
 namespace Microsoft.Diagnostics.DataContractReader.Contracts;
 
@@ -195,34 +193,16 @@ private class DynamicMethodDesc : IData<DynamicMethodDesc>
         private DynamicMethodDesc(Target target, TargetPointer methodDescPointer)
         {
             _address = methodDescPointer;
-            List<byte> nameBytes = new();
             _desc = target.ProcessedData.GetOrAdd<Data.DynamicMethodDesc>(methodDescPointer);
 
-            if (_desc.MethodName != TargetPointer.Null)
-            {
-                TargetPointer currentNameAddress = _desc.MethodName;
-                do
-                {
-                    byte nameByte = target.Read<byte>(currentNameAddress);
-
-                    if (nameByte == 0)
-                        break;
-
-                    nameBytes.Add(nameByte);
-                    currentNameAddress++;
-                } while (true);
-
-                MethodName = nameBytes.ToArray();
-            }
-            else
-            {
-                MethodName = System.Array.Empty<byte>();
-            }
+            MethodName = _desc.MethodName != TargetPointer.Null
+                ? target.ReadUtf8String(_desc.MethodName)
+                : string.Empty;
 
             _storedSigDesc = target.ProcessedData.GetOrAdd<Data.StoredSigMethodDesc>(methodDescPointer);
         }
 
-        public byte[] MethodName { get; }
+        public string MethodName { get; }
         public DynamicMethodDescExtendedFlags ExtendedFlags => (DynamicMethodDescExtendedFlags)_storedSigDesc.ExtendedFlags;
 
         public bool IsDynamicMethod => ExtendedFlags.HasFlag(DynamicMethodDescExtendedFlags.IsLCGMethod);
@@ -712,13 +692,13 @@ public bool IsArrayMethod(MethodDescHandle methodDescHandle, out ArrayFunctionTy
         return true;
     }
 
-    public bool IsNoMetadataMethod(MethodDescHandle methodDescHandle, out ReadOnlySpan<byte> methodName)
+    public bool IsNoMetadataMethod(MethodDescHandle methodDescHandle, out string methodName)
     {
         MethodDesc methodDesc = _methodDescs[methodDescHandle.Address];
 
         if (methodDesc.Classification != MethodClassification.Dynamic)
         {
-            methodName = default;
+            methodName = string.Empty;
             return false;
         }
 

src/native/managed/cdacreader/src/Legacy/TypeNameBuilder.cs

@@ -65,11 +65,11 @@ public static void AppendMethodImpl(Target target, StringBuilder stringBuilder,
     {
         IRuntimeTypeSystem runtimeTypeSystem = target.Contracts.RuntimeTypeSystem;
         ILoader loader = target.Contracts.Loader;
-        ReadOnlySpan<byte> methodNameSpan;
+        string methodName;
         TypeHandle th = default;
         Contracts.ModuleHandle module = default;
 
-        bool isNoMetadataMethod = runtimeTypeSystem.IsNoMetadataMethod(method, out methodNameSpan);
+        bool isNoMetadataMethod = runtimeTypeSystem.IsNoMetadataMethod(method, out methodName);
         if (isNoMetadataMethod)
         {
             if (runtimeTypeSystem.IsDynamicMethod(method))
@@ -91,7 +91,7 @@ public static void AppendMethodImpl(Target target, StringBuilder stringBuilder,
 
         if (isNoMetadataMethod)
         {
-            stringBuilder.Append(Encoding.UTF8.GetString(methodNameSpan));
+            stringBuilder.Append(methodName);
         }
         else if (runtimeTypeSystem.IsArrayMethod(method, out ArrayFunctionType functionType))
         {

src/native/managed/cdacreader/src/Target.cs

@@ -5,9 +5,9 @@
 using System.Collections.Generic;
 using System.Diagnostics;
 using System.Diagnostics.CodeAnalysis;
-using System.Net;
 using System.Numerics;
 using System.Runtime.CompilerServices;
+using System.Text;
 using Microsoft.Diagnostics.DataContractReader.Data;
 
 namespace Microsoft.Diagnostics.DataContractReader;
@@ -265,6 +265,12 @@ private static DataType GetDataType(string type)
 
     public int PointerSize => _config.PointerSize;
 
+    /// <summary>
+    /// Read a value from the target in target endianness
+    /// </summary>
+    /// <typeparam name="T">Type of value to read</typeparam>
+    /// <param name="address">Address to start reading from</param>
+    /// <returns>Value read from the target</returns>
     public T Read<T>(ulong address) where T : unmanaged, IBinaryInteger<T>, IMinMaxValue<T>
     {
         if (!TryRead(address, _config.IsLittleEndian, _reader, out T value))
@@ -319,6 +325,11 @@ private static bool IsSigned<T>() where T : struct, INumberBase<T>, IMinMaxValue
         return T.IsNegative(T.MinValue);
     }
 
+    /// <summary>
+    /// Read a pointer from the target in target endianness
+    /// </summary>
+    /// <param name="address">Address to start reading from</param>
+    /// <returns>Pointer read from the target</returns>}
     public TargetPointer ReadPointer(ulong address)
     {
         if (!TryReadPointer(address, _config, _reader, out TargetPointer pointer))
@@ -352,6 +363,64 @@ public void ReadPointers(ulong address, Span<TargetPointer> buffer)
         }
     }
 
+    /// <summary>
+    /// Read a null-terminated UTF-8 string from the target
+    /// </summary>
+    /// <param name="address">Address to start reading from</param>
+    /// <returns>String read from the target</returns>}
+    public string ReadUtf8String(ulong address)
+    {
+        // Read characters until we find the null terminator
+        ulong end = address;
+        while (Read<byte>(end) != 0)
+        {
+            end += sizeof(byte);
+        }
+
+        int length = (int)(end - address);
+        if (length == 0)
+            return string.Empty;
+
+        Span<byte> span = length <= StackAllocByteThreshold
+            ? stackalloc byte[length]
+            : new byte[length];
+        ReadBuffer(address, span);
+        return Encoding.UTF8.GetString(span);
+    }
+
+    /// <summary>
+    /// Read a null-terminated UTF-16 string from the target in target endianness
+    /// </summary>
+    /// <param name="address">Address to start reading from</param>
+    /// <returns>String read from the target</returns>}
+    public string ReadUtf16String(ulong address)
+    {
+        // Read characters until we find the null terminator
+        ulong end = address;
+        while (Read<char>(end) != 0)
+        {
+            end += sizeof(char);
+        }
+
+        int length = (int)(end - address);
+        if (length == 0)
+            return string.Empty;
+
+        Span<byte> span = length <= StackAllocByteThreshold
+            ? stackalloc byte[length]
+            : new byte[length];
+        ReadBuffer(address, span);
+        string result = _config.IsLittleEndian
+            ? Encoding.Unicode.GetString(span)
+            : Encoding.BigEndianUnicode.GetString(span);
+        return result;
+    }
+
+    /// <summary>
+    /// Read a native unsigned integer from the target in target endianness
+    /// </summary>
+    /// <param name="address">Address to start reading from</param>
+    /// <returns>Value read from the target</returns>
     public TargetNUInt ReadNUInt(ulong address)
     {
         if (!TryReadNUInt(address, _config, _reader, out ulong value))