Add support for string mask early exit

Author: Genbox

Src/FastData.Benchmarks/Benchmarks/GPerfAnalyzerBenchmarks.cs

@@ -18,7 +18,7 @@ public GPerfAnalyzerBenchmarks()
         Random rng = new Random(42);
         _data = Enumerable.Range(1, 100).Select(_ => TestHelper.GenerateRandomString(rng, 50)).ToArray();
 
-        StringKeyProperties props = KeyAnalyzer.GetStringProperties(_data, false);
+        StringKeyProperties props = KeyAnalyzer.GetStringProperties(_data, false, false);
         _analyzer = new GPerfAnalyzer(_data.Length, props, new GPerfAnalyzerConfig(), new Simulator(_data.Length, GeneratorEncoding.UTF16), NullLogger<GPerfAnalyzer>.Instance);
     }
 

Src/FastData.Benchmarks/Benchmarks/KeyAnalyzerBenchmarks.cs

@@ -15,5 +15,5 @@ public KeyAnalyzerBenchmarks()
     }
 
     [Benchmark]
-    public object GetStringProperties() => KeyAnalyzer.GetStringProperties(_data, true);
+    public object GetStringProperties() => KeyAnalyzer.GetStringProperties(_data, true, false);
 }

Src/FastData.Benchmarks/Benchmarks/SegmentGeneratorsBenchmarks.cs

@@ -14,7 +14,7 @@ public class SegmentGeneratorsBenchmarks
     private readonly OffsetGenerator _ofGen = new OffsetGenerator();
 
     //We start at 8 and go up to 100 to cover as many cases as possible
-    private readonly StringKeyProperties _props = KeyAnalyzer.GetStringProperties(Enumerable.Range(8, 100).Select(x => TestHelper.GenerateRandomString(Random.Shared, x)).ToArray(), false);
+    private readonly StringKeyProperties _props = KeyAnalyzer.GetStringProperties(Enumerable.Range(8, 100).Select(x => TestHelper.GenerateRandomString(Random.Shared, x)).ToArray(), false, false);
 
     [Benchmark]
     public object BruteForceGenerator() => _bfGen.Generate(_props).ToArray();

Src/FastData.Cli.Tests/CommandOutputs/cpp_-k UInt8_Files_Integers.input.verified.txt

@@ -4,6 +4,7 @@
 // Structure: Auto (Conditional)
 #pragma once
 #include <array>
+#include <cstring>
 #include <cstdint>
 #include <limits>
 #include <string_view>

Src/FastData.Cli.Tests/CommandOutputs/cpp_-s HashTable_Files_Strings.input.verified.txt

@@ -4,6 +4,7 @@
 // Structure: HashTable
 #pragma once
 #include <array>
+#include <cstring>
 #include <cstdint>
 #include <limits>
 #include <string_view>
@@ -43,6 +44,7 @@ public:
              return false;
 
 
+
         const uint64_t hash = get_hash(key);
         const size_t index = hash % 2;
         int8_t i = static_cast<int8_t>(buckets[index] - 1);

Src/FastData.Cli.Tests/CommandOutputs/cpp_Files_Strings.input.verified.txt

@@ -4,6 +4,7 @@
 // Structure: Auto (Conditional)
 #pragma once
 #include <array>
+#include <cstring>
 #include <cstdint>
 #include <limits>
 #include <string_view>
@@ -15,6 +16,7 @@ public:
          if (key.length() != 5u)
              return false;
 
+
         if (key == "test1" || key == "test2")
             return true;
 

Src/FastData.Cli.Tests/CommandOutputs/csharp_-s HashTable_Files_Strings.input.verified.txt

@@ -60,6 +60,10 @@ internal static class MyData
     {
         if (key.Length != 5u)
             return false;
+        ulong first = (ulong)key[0] | ((ulong)key[1] << 16) | ((ulong)key[2] << 32) | ((ulong)key[3] << 48);
+
+        if ((first & 18414092482483257227UL) != 0)
+            return false;
 
 
         ulong hash = Hash(key);

Src/FastData.Cli.Tests/CommandOutputs/csharp_Files_Strings.input.verified.txt

@@ -15,6 +15,10 @@ internal static class MyData
     {
         if (key.Length != 5u)
             return false;
+        ulong first = (ulong)key[0] | ((ulong)key[1] << 16) | ((ulong)key[2] << 32) | ((ulong)key[3] << 48);
+
+        if ((first & 18414092482483257227UL) != 0)
+            return false;
 
 
         switch (key)

Src/FastData.Cli.Tests/CommandOutputs/rust_-s HashTable_Files_Strings.input.verified.txt

@@ -45,6 +45,12 @@ const ENTRIES: [e; 2] = [
         if key.len() != 5 as usize {
             return false;
         }
+        let bytes = key.as_bytes();
+        let first = (bytes[0] as u64) | (bytes[1] as u64) << 8 | (bytes[2] as u64) << 16 | (bytes[3] as u64) << 24 | (bytes[4] as u64) << 32;
+
+        if (first & 878514576011u64) != 0 {
+            return false;
+        }
 
 
         let hash = unsafe { Self::get_hash(key) };

Src/FastData.Cli.Tests/CommandOutputs/rust_Files_Strings.input.verified.txt

@@ -16,6 +16,12 @@ impl MyData {
         if key.len() != 5 as usize {
             return false;
         }
+        let bytes = key.as_bytes();
+        let first = (bytes[0] as u64) | (bytes[1] as u64) << 8 | (bytes[2] as u64) << 16 | (bytes[3] as u64) << 24 | (bytes[4] as u64) << 32;
+
+        if (first & 878514576011u64) != 0 {
+            return false;
+        }
 
 
         if key == "test1" || key == "test2" {