We have extra and unncessary checks.

Author: lemire

src/UTF8.cs

@@ -124,7 +124,8 @@ private static (int utfAdjust, int scalarAdjust) GetFinalScalarUtfAdjustments(by
                     } // Too short
                     // range check
                     codePoint = (uint)(firstByte & 0b00011111) << 6 | (uint)(buf[pos + 1] & 0b00111111);
-                    if ((codePoint < 0x80) || (0x7ff < codePoint))
+                    // codePoint is necessarily <= 0x7ff
+                    if (codePoint < 0x80)
                     {
                         return buf + pos;
                     } // Overlong
@@ -141,7 +142,8 @@ private static (int utfAdjust, int scalarAdjust) GetFinalScalarUtfAdjustments(by
                                  (uint)(buf[pos + 1] & 0b00111111) << 6 |
                                  (uint)(buf[pos + 2] & 0b00111111);
                     // Either overlong or too large:
-                    if ((codePoint < 0x800) || (0xffff < codePoint) ||
+                    // codePoint is necessarily <= 0xffff
+                    if ((codePoint < 0x800) ||
                         (0xd7ff < codePoint && codePoint < 0xe000))
                     {
                         return buf + pos;
@@ -238,7 +240,8 @@ private static (int utfAdjust, int scalarAdjust) GetFinalScalarUtfAdjustments(by
                     } // Too short
                     // range check
                     codePoint = (uint)(firstByte & 0b00011111) << 6 | (uint)(pInputBuffer[pos + 1] & 0b00111111);
-                    if ((codePoint < 0x80) || (0x7ff < codePoint))
+                    // codePoint is necessarily <= 0x7ff
+                    if (codePoint < 0x80)
                     {
                         utf16CodeUnitCountAdjustment = TempUtf16CodeUnitCountAdjustment;
                         scalarCountAdjustment = TempScalarCountAdjustment;
@@ -261,7 +264,8 @@ private static (int utfAdjust, int scalarAdjust) GetFinalScalarUtfAdjustments(by
                                  (uint)(pInputBuffer[pos + 1] & 0b00111111) << 6 |
                                  (uint)(pInputBuffer[pos + 2] & 0b00111111);
                     // Either overlong or too large:
-                    if ((codePoint < 0x800) || (0xffff < codePoint) ||
+                    // codePoint is necessarily <= 0xffff
+                    if ((codePoint < 0x800) ||
                         (0xd7ff < codePoint && codePoint < 0xe000))
                     {
                         utf16CodeUnitCountAdjustment = TempUtf16CodeUnitCountAdjustment;

test/AsciiTest.cs

@@ -1,4 +1,5 @@
-namespace tests;
+namespace tests;
+
 using System.Text;
 using SimdUnicode;
 

test/UTF8ValidationTests.cs

@@ -1,4 +1,5 @@
 namespace tests;
+
 using System.Text;
 using SimdUnicode;
 using System.Diagnostics;
@@ -1245,7 +1246,8 @@ public static bool ValidateUtf8Fuschia(byte[] data)
                 if ((data[pos + 1] & 0b11000000) != 0b10000000) return false;
 
                 codePoint = (uint)((byte1 & 0b00011111) << 6 | (data[pos + 1] & 0b00111111));
-                if (codePoint < 0x80 || 0x7ff < codePoint) return false;
+                // codePoint is necessarily <= 0x7ff
+                if (codePoint < 0x80) return false;
                 pos += 2;
             }
             else if ((byte1 & 0b11110000) == 0b11100000)
@@ -1255,7 +1257,7 @@ public static bool ValidateUtf8Fuschia(byte[] data)
                 if ((data[pos + 2] & 0b11000000) != 0b10000000) return false;
 
                 codePoint = (uint)((byte1 & 0b00001111) << 12 | (data[pos + 1] & 0b00111111) << 6 | (data[pos + 2] & 0b00111111));
-                if (codePoint < 0x800 || 0xffff < codePoint || (0xd7ff < codePoint && codePoint < 0xe000)) return false;
+                if (codePoint < 0x800 || (0xd7ff < codePoint && codePoint < 0xe000)) return false;
                 pos += 3;
             }
             else if ((byte1 & 0b11111000) == 0b11110000)

test/helpers/randomutf8.cs

@@ -1,4 +1,5 @@
 namespace tests;
+
 using System;
 using System.Collections.Generic;
 using System.Linq;