small optimization

Author: Nick-Nuon

src/SimdUnicode.csproj

@@ -2,7 +2,7 @@
 
   <PropertyGroup>
     <OutputType>Library</OutputType>
-    <TargetFramework>net7.0</TargetFramework>
+    <TargetFramework>net8.0</TargetFramework>
     <Nullable>enable</Nullable>
     <!-- This is required for SIMD, sse c# - How to run unsafe code in "visual studio code"? - Stack Overflow https://stackoverflow.com/questions/50636693/how-to-run-unsafe-code-in-visual-studio-code -->
     <AllowUnsafeBlocks>true</AllowUnsafeBlocks> 

src/UTF8_validation.cs

@@ -8,6 +8,12 @@
 // Vector256 https://learn.microsoft.com/en-us/dotnet/api/system.runtime.intrinsics.vector256-1?view=net-7.0
 //  I extend it as needed
 
+
+// |                      Method |    N |       Mean |     Error |    StdDev |   Gen0 | Allocated |
+// |---------------------------- |----- |-----------:|----------:|----------:|-------:|----------:|
+// | SIMDUtf8ValidationValidUtf8 |  100 |   165.8 us |   2.87 us |   2.55 us | 0.4883 |  54.69 KB |
+// | SIMDUtf8ValidationValidUtf8 | 8000 | 8,733.5 us | 167.05 us | 211.27 us |      - |  33.21 KB |
+
 public static class Vector256Extensions
 {
     // Gets the second lane of the current vector and the first lane of the previous vector and returns, then shift it right by an appropriate number of bytes (less than 16, or less than 128 bits)
@@ -177,7 +183,7 @@ public static unsafe class Utf8Utility
     }
 
 // C# docs suggests that classes are allocated on the heap:
-// it doesnt seem to do much in this case but I tthought the suggestion to be sensible. 
+// it doesnt seem to do much in this case but I thought the suggestion to be sensible. 
     public struct utf8_validation
     {
         public struct utf8_checker
@@ -359,12 +365,14 @@ private Vector256<byte> must_be_2_3_continuation(Vector256<byte> prev2, Vector25
                 Vector256<byte> is_fourth_byte = Avx2.SubtractSaturate(prev3, Vector256.Create((byte)(0b11110000u - 1)));
 
                 Vector256<byte> combined = Avx2.Or(is_third_byte, is_fourth_byte);
+                return combined;
+
+                // Vector256<sbyte> signedCombined = combined.AsSByte();
 
-                Vector256<sbyte> signedCombined = combined.AsSByte();
-                Vector256<sbyte> zero = Vector256<sbyte>.Zero;
-                Vector256<sbyte> comparisonResult = Avx2.CompareGreaterThan(signedCombined, zero);
+                // Vector256<sbyte> zero = Vector256<sbyte>.Zero;
+                // Vector256<sbyte> comparisonResult = Avx2.CompareGreaterThan(signedCombined, zero);
 
-                return comparisonResult.AsByte();
+                // return comparisonResult.AsByte();
             }
 
 
@@ -377,6 +385,13 @@ private Vector256<byte> must_be_2_3_continuation(Vector256<byte> prev2, Vector25
             };
             Vector256<byte> maxValue = Vector256.Create(MaxArray);
 
+    //         private static readonly Vector256<byte> maxValue = Vector256.Create(
+    // 255, 255, 255, 255, 255, 255, 255, 255,
+    // 255, 255, 255, 255, 255, 255, 255, 255,
+    // 255, 255, 255, 255, 255, 255, 255, 255,
+    // 255, 255, 255, 255, 255, 0b11110000 - 1, 0b11100000 - 1, 0b11000000 - 1);
+
+
             [MethodImpl(MethodImplOptions.AggressiveInlining)]
 
             private Vector256<byte> is_incomplete(Vector256<byte> input)

src/helpers.cs

@@ -28,5 +28,30 @@ public static Vector256<byte> CompareGreaterThan(Vector256<byte> left, Vector256
 
             return result;
         }
+
+        private static readonly int[] previousCollectionCounts = new int[GC.MaxGeneration + 1];
+
+        public static void CheckForGCCollections(string sectionName)
+        {
+            bool collectionOccurred = false;
+
+            for (int i = 0; i <= GC.MaxGeneration; i++)
+            {
+                int currentCount = GC.CollectionCount(i);
+                if (currentCount != previousCollectionCounts[i])
+                {
+                    Console.WriteLine($"GC occurred in generation {i} during '{sectionName}'. Collections: {currentCount - previousCollectionCounts[i]}");
+                    previousCollectionCounts[i] = currentCount;
+                    collectionOccurred = true;
+                }
+            }
+
+            if (!collectionOccurred)
+            {
+                Console.WriteLine($"No GC occurred during '{sectionName}'.");
+            }
+        }
     }
+
+    
 }

test/AsciiTest.cs

@@ -2,8 +2,6 @@ namespace tests;
 using System.Text;
 using SimdUnicode;
 
-//TODO (Nick Nuon): Test UTF8 Generator works correctly
-
 public class AsciiTest
 {
     [Fact]
@@ -140,7 +138,7 @@ public void TestNoErrorGetIndexOfFirstNonAsciiByte()
             {
                 fixed (byte* pAscii = ascii)
                 {
-                    nuint result = Ascii.GetIndexOfFirstNonAsciiByte(pAscii, (nuint)ascii.Length);
+                    nuint result = SimdUnicode.Ascii.GetIndexOfFirstNonAsciiByte(pAscii, (nuint)ascii.Length);
                     if (result != (nuint)ascii.Length)
                     {
                         throw new Exception($"Unexpected non-ASCII character found at index {result}");
@@ -170,7 +168,7 @@ public void TestErrorGetIndexOfFirstNonAsciiByte()
                 {
                     fixed (byte* pAscii = ascii)
                     {
-                        nuint result = Ascii.GetIndexOfFirstNonAsciiByte(pAscii, (nuint)ascii.Length);
+                        nuint result = SimdUnicode.Ascii.GetIndexOfFirstNonAsciiByte(pAscii, (nuint)ascii.Length);
                         if (result != (nuint)i)
                         {
                             throw new Exception($"Expected non-ASCII character at index {i}, but found at index {result}");