Merge pull request #35 from sveinungf/dev/encoding

New XML encode implementation

.editorconfig

@@ -157,6 +157,8 @@ dotnet_diagnostic.CS9113.severity = none
 [{*Test,*.Benchmark,*.Cmd}/**.cs]
 # CA1305: Specify IFormatProvider
 dotnet_diagnostic.CA1305.severity = none
+# CA1062: Validate arguments of public methods
+dotnet_diagnostic.CA1062.severity = none
 # CA1707: Identifiers should not contain underscores
 dotnet_diagnostic.CA1707.severity = none
 # CA1720: Identifiers should not contain type names
@@ -201,7 +203,7 @@ MA0053.public_class_should_be_sealed = true
 # MA0076: Do not use implicit culture-sensitive ToString in interpolated strings
 dotnet_diagnostic.MA0076.severity = warning
 
-# Disable MA0053 for types that have been publicly exposed in a previous version
+# Disable MA0053 (sealing class) for types that have been publicly exposed in a previous version
 [{AutoFilterOptions.cs,ColumnOptions.cs,RowOptions.cs,SpreadCheetahOptions.cs,WorksheetOptions.cs,WorksheetRowGenerator.cs}]
 dotnet_diagnostic.MA0053.severity = none
 

BannedSymbols.txt

@@ -1,2 +1,3 @@
 M:System.IO.Compression.ZipArchive.CreateEntry(System.String);Pass CompressionLevel parameter
+M:System.Net.WebUtility.HtmlEncode(System.String);Use XmlUtility.XmlEncode
 P:System.Nullable`1.Value;GetValueOrDefault() or the coalesce operator do less work

Directory.Build.props

@@ -9,14 +9,6 @@
     <AccelerateBuildsInVisualStudio>true</AccelerateBuildsInVisualStudio>
   </PropertyGroup>
 
-  <ItemGroup Condition="'$(MSBuildProjectName)' != 'SpreadCheetah.SourceGenerator' And '$(MSBuildProjectName)' != 'SpreadCheetah.SourceGenerator.CSharp8Test'">
-    <PackageReference Include="Polyfill">
-      <PrivateAssets>all</PrivateAssets>
-      <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>
-      <NoWarn>MA0048</NoWarn>
-    </PackageReference>
-  </ItemGroup>
-
   <ItemGroup>
     <AdditionalFiles Include="$(MSBuildThisFileDirectory)/BannedSymbols.txt" Visible="false" />
   </ItemGroup>

SpreadCheetah.Benchmark/SpreadCheetah.Benchmark.csproj

@@ -11,6 +11,10 @@
     <PackageReference Include="DocumentFormat.OpenXml" />
     <PackageReference Include="EPPlusFree" />
     <PackageReference Include="Microsoft.CodeAnalysis.CSharp" PrivateAssets="all" />
+    <PackageReference Include="Polyfill">
+      <PrivateAssets>all</PrivateAssets>
+      <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>
+    </PackageReference>
   </ItemGroup>
 
   <ItemGroup>

SpreadCheetah.SourceGenerator.Test/SpreadCheetah.SourceGenerator.Test.csproj

@@ -11,6 +11,10 @@
       <PrivateAssets>all</PrivateAssets>
     </PackageReference>
     <PackageReference Include="Microsoft.NET.Test.Sdk" />
+    <PackageReference Include="Polyfill">
+      <PrivateAssets>all</PrivateAssets>
+      <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>
+    </PackageReference>
     <PackageReference Include="xunit" />
     <PackageReference Include="xunit.runner.visualstudio">
       <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>

SpreadCheetah.Test/SpreadCheetah.Test.csproj

@@ -30,6 +30,13 @@
     </PackageReference>
   </ItemGroup>
 
+  <ItemGroup Condition="'$(Configuration)' != 'Debug'">
+    <PackageReference Include="Polyfill">
+      <PrivateAssets>all</PrivateAssets>
+      <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>
+    </PackageReference>
+  </ItemGroup>
+
   <ItemGroup Condition="'$(TargetFramework)' == 'net472'">
     <PackageReference Include="System.Collections.Immutable" />
     <PackageReference Include="System.IO.Compression" />

SpreadCheetah.Test/Tests/SpreadsheetRowTests.cs

@@ -111,10 +111,14 @@ public async Task Spreadsheet_AddRow_CellWithoutValue(CellType type, RowCollecti
         "With whitespace",
         "With trailing whitespace ",
         " With leading whitespace",
-        "With-Special-Characters!#¤%&",
+        "With-Special!#¤%Characters",
+        "With-Special<>&'\\Characters",
         "With\"Quotation\"Marks",
         "WithNorwegianCharactersÆØÅ",
-        "WithEmoji\ud83d\udc4d",
+        "With\ud83d\udc4dEmoji",
+        "With🌉Emoji",
+        "With\tValid\nControlCharacters",
+        "WithCharacters\u00a0\u00c9\u00ffBetween160And255",
         "",
         null);
 
@@ -144,6 +148,30 @@ public async Task Spreadsheet_AddRow_CellWithStringValue(string? value, CellType
         Assert.Equal(value ?? string.Empty, actualCell.InnerText);
     }
 
+    [Theory]
+    [MemberData(nameof(TestData.CellTypes), MemberType = typeof(TestData))]
+    public async Task Spreadsheet_AddRow_CellWithInvalidControlCharacterStringValue(CellType type, RowCollectionType rowType)
+    {
+        // Arrange
+        const string value = "With\u0000Control\u0010\u001fCharacters";
+        using var stream = new MemoryStream();
+        await using var spreadsheet = await Spreadsheet.CreateNewAsync(stream);
+        await spreadsheet.StartWorksheetAsync("Sheet");
+        var cell = CellFactory.Create(type, value);
+
+        // Act
+        await spreadsheet.AddRowAsync(cell, rowType);
+        await spreadsheet.FinishAsync();
+
+        // Assert
+        SpreadsheetAssert.Valid(stream);
+        using var actual = SpreadsheetDocument.Open(stream, true);
+        var sheetPart = actual.WorkbookPart!.WorksheetParts.Single();
+        var actualCell = sheetPart.Worksheet.Descendants<OpenXmlCell>().Single();
+        Assert.Equal(CellValues.InlineString, actualCell.DataType?.Value);
+        Assert.Equal("WithControlCharacters", actualCell.InnerText);
+    }
+
     public static IEnumerable<object?[]> StringLengths() => TestData.CombineWithCellTypes(
         4095,
         4096,

SpreadCheetah.Test/Tests/XmlUtilityTests.cs

@@ -0,0 +1,60 @@
+#if DEBUG
+using SpreadCheetah.Helpers;
+using System.Text;
+using Xunit;
+
+namespace SpreadCheetah.Test.Tests;
+
+public static class XmlUtilityTests
+
+{
+    [Theory]
+    [InlineData("", "")]
+    [InlineData("Just a regular string without any special characters", "Just a regular string without any special characters")]
+    [InlineData("&", "&")]
+    [InlineData("&<>'\"", "&amp;&lt;&gt;&apos;&quot;")]
+    [InlineData("&OnlyFirstCharacter", "&amp;OnlyFirstCharacter")]
+    [InlineData("&This<string>has'special\"characters in between", "&amp;This&lt;string&gt;has&apos;special&quot;characters in between")]
+    [InlineData("This<string>has'special\"characters in between&", "This&lt;string&gt;has&apos;special&quot;characters in between&amp;")]
+    [InlineData("\tHandling\r\nAllowed\nControl\tCharacters\n", "\tHandling\r\nAllowed\nControl\tCharacters\n")]
+    [InlineData("\u0001Handling\u0002Invalid\u0003Control\u0004Characters\u0005", "HandlingInvalidControlCharacters")]
+    [InlineData("\u0006", "")]
+    [InlineData("\u0007\u0008", "")]
+    public static void XmlUtility_TryXmlEncodeToUtf8_Success(string value, string expected)
+    {
+        // Arrange
+        var buffer = new byte[value.Length * 6];
+
+        // Act
+        var result = XmlUtility.TryXmlEncodeToUtf8(value.AsSpan(), buffer, out var bytesWritten);
+
+        // Assert
+        Assert.True(result);
+
+        var bytes = buffer.AsSpan(0, bytesWritten);
+        var actual = Encoding.UTF8.GetString(bytes.ToArray());
+        Assert.Equal(expected, actual);
+    }
+
+    [Theory]
+    [InlineData("", "")]
+    [InlineData("Just a regular string without any special characters", "Just a regular string without any special characters")]
+    [InlineData("&", "&amp;")]
+    [InlineData("&<>'\"", "&amp;&lt;&gt;&apos;&quot;")]
+    [InlineData("&OnlyFirstCharacter", "&amp;OnlyFirstCharacter")]
+    [InlineData("&This<string>has'special\"characters in between", "&amp;This&lt;string&gt;has&apos;special&quot;characters in between")]
+    [InlineData("This<string>has'special\"characters in between&", "This&lt;string&gt;has&apos;special&quot;characters in between&amp;")]
+    [InlineData("\tHandling\r\nAllowed\nControl\tCharacters\n", "\tHandling\r\nAllowed\nControl\tCharacters\n")]
+    [InlineData("\u0001Handling\u0002Invalid\u0003Control\u0004Characters\u0005", "HandlingInvalidControlCharacters")]
+    [InlineData("\u0006", "")]
+    [InlineData("\u0007\u0008", "")]
+    public static void XmlUtility_XmlEncode_Success(string value, string expected)
+    {
+        // Act
+        var result = XmlUtility.XmlEncode(value);
+
+        // Assert
+        Assert.Equal(expected, result);
+    }
+}
+#endif

SpreadCheetah/CellWriters/BaseCellWriter.cs

@@ -277,9 +277,10 @@ protected bool FinishWritingFormulaCellValue(in Cell cell, string formulaText, r
         // Write the "</f><v>" part
         if (cellValueIndex < cachedValueStartIndex)
         {
-            var separator = FormulaCellHelper.EndFormulaBeginCachedValue;
-            if (separator.Length > Buffer.FreeCapacity) return false;
-            Buffer.Advance(SpanHelper.GetBytes(separator, Buffer.GetSpan()));
+            if (!FormulaCellHelper.EndFormulaBeginCachedValue.TryCopyTo(Buffer.GetSpan()))
+                return false;
+
+            Buffer.Advance(FormulaCellHelper.EndFormulaBeginCachedValue.Length);
             cellValueIndex = cachedValueStartIndex;
         }
 

SpreadCheetah/DataCell.cs

@@ -1,5 +1,4 @@
 using SpreadCheetah.CellValueWriters;
-using System.Net;
 
 namespace SpreadCheetah;
 
@@ -18,7 +17,7 @@ public readonly record struct DataCell
     /// </summary>
     public DataCell(string? value)
     {
-        StringValue = value != null ? WebUtility.HtmlEncode(value) : string.Empty;
+        StringValue = value;
         Type = value != null ? CellWriterType.String : CellWriterType.Null;
     }
 

SpreadCheetah/Formula.cs

@@ -1,5 +1,3 @@
-using System.Net;
-
 namespace SpreadCheetah;
 
 /// <summary>
@@ -15,6 +13,6 @@ public readonly record struct Formula
     /// </summary>
     public Formula(string? formulaText)
     {
-        FormulaText = formulaText != null ? WebUtility.HtmlEncode(formulaText) : string.Empty;
+        FormulaText = formulaText ?? "";
     }
 }

SpreadCheetah/Helpers/SpanHelper.cs

@@ -7,12 +7,6 @@ namespace SpreadCheetah.Helpers;
 
 internal static class SpanHelper
 {
-    public static int GetBytes(ReadOnlySpan<byte> source, Span<byte> destination)
-    {
-        source.CopyTo(destination);
-        return source.Length;
-    }
-
     [MethodImpl(MethodImplOptions.AggressiveInlining)]
     public static bool TryCopyTo(this ReadOnlySpan<byte> source, Span<byte> bytes, ref int bytesWritten)
     {
@@ -81,7 +75,7 @@ public static bool TryWrite(ReadOnlySpan<char> value, Span<byte> bytes, ref int
     }
 
 #if NETSTANDARD2_0
-    public static bool TryWriteLongString(string value, ref int valueIndex, Span<byte> bytes, ref int bytesWritten)
+    public static bool TryWriteLongString(string? value, ref int valueIndex, Span<byte> bytes, ref int bytesWritten)
         => TryWriteLongString(value.AsSpan(), ref valueIndex, bytes, ref bytesWritten);
 #endif
 

SpreadCheetah/Helpers/Utf8Helper.cs

@@ -1,3 +1,4 @@
+using System.Runtime.CompilerServices;
 using System.Text;
 
 namespace SpreadCheetah.Helpers;
@@ -6,13 +7,26 @@ internal static class Utf8Helper
 {
     public const int MaxBytePerChar = 6;
 
-    private static readonly UTF8Encoding Utf8NoBom = new(false);
+    public static readonly UTF8Encoding Utf8NoBom = new(false);
 
     public static int GetBytes(ReadOnlySpan<char> chars, Span<byte> bytes) => Utf8NoBom.GetBytes(chars, bytes);
 
-#if NETSTANDARD2_0
-    public static bool TryGetBytes(string? chars, Span<byte> bytes, out int bytesWritten) => TryGetBytes(chars.AsSpan(), bytes, out bytesWritten);
-#endif
+    public static bool TryGetBytes(ReadOnlySpan<char> chars, Span<byte> bytes, out int bytesWritten)
+    {
+        // Try first with an approximate value length, then try with a more accurate value length
+        if (DestinationCanFitTranscodedString(chars, bytes))
+        {
+            bytesWritten = Utf8NoBom.GetBytes(chars, bytes);
+            return true;
+        }
 
-    public static bool TryGetBytes(ReadOnlySpan<char> chars, Span<byte> bytes, out int bytesWritten) => Utf8NoBom.TryGetBytesInternal(chars, bytes, out bytesWritten);
+        bytesWritten = 0;
+        return false;
+    }
+
+    [MethodImpl(MethodImplOptions.AggressiveInlining)]
+    public static bool DestinationCanFitTranscodedString(ReadOnlySpan<char> chars, Span<byte> bytes)
+    {
+        return bytes.Length >= chars.Length * MaxBytePerChar || bytes.Length >= Utf8NoBom.GetByteCount(chars);
+    }
 }