Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Support writing base64 JSON segments #111041

Merged
merged 3 commits into from
Jan 14, 2025
Merged
Show file tree
Hide file tree
Changes from 2 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
129 changes: 88 additions & 41 deletions src/libraries/Fuzzing/DotnetFuzzing/Fuzzers/Utf8JsonWriterFuzzer.cs
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@

using System;
using System.Buffers;
using System.Buffers.Text;
using System.Collections;
using System.Collections.Generic;
using System.Diagnostics;
Expand Down Expand Up @@ -33,8 +34,11 @@ internal sealed class Utf8JsonWriterFuzzer : IFuzzer
private const byte NewLineFlag = 1 << 3;
private const byte SkipValidationFlag = 1 << 4;

// Options for choosing between UTF-8 and UTF-16 encoding
private const byte EncodingFlag = 1 << 5;
// Options for choosing between base64, UTF-8 and UTF-16 encoding
private const byte EncodingMask = 0b11 << 5;
private const byte Utf8EncodingFlag = 0b00 << 5;
private const byte Utf16EncodingFlag = 0b01 << 5;
private const byte Base64EncodingFlag = 0b10 << 5;

public void FuzzTarget(ReadOnlySpan<byte> bytes)
{
Expand All @@ -53,8 +57,13 @@ public void FuzzTarget(ReadOnlySpan<byte> bytes)
ReadOnlySpan<char> chars = MemoryMarshal.Cast<byte, char>(bytes);

// Validate that the indices are within bounds of the input
bool utf8 = (optionsByte & EncodingFlag) == 0;
if (!(0 <= slice1 && slice1 <= slice2 && slice2 <= (utf8 ? bytes.Length : chars.Length)))
int encoding = optionsByte & EncodingMask;
if (encoding is not Utf8EncodingFlag and not Utf16EncodingFlag and not Base64EncodingFlag)
{
return;
}

if (!(0 <= slice1 && slice1 <= slice2 && slice2 <= (encoding is Utf16EncodingFlag ? chars.Length : bytes.Length)))
{
return;
}
Expand All @@ -63,7 +72,7 @@ public void FuzzTarget(ReadOnlySpan<byte> bytes)
bool indented = (optionsByte & IndentFlag) == 0;
JsonWriterOptions options = new()
{
Encoder = (optionsByte & EncodingFlag) == 0 ? JavaScriptEncoder.Default : JavaScriptEncoder.UnsafeRelaxedJsonEscaping,
Encoder = (optionsByte & EncoderFlag) == 0 ? JavaScriptEncoder.Default : JavaScriptEncoder.UnsafeRelaxedJsonEscaping,
Indented = indented,
MaxDepth = (optionsByte & MaxDepthFlag) == 0 ? 1 : 0,
NewLine = (optionsByte & NewLineFlag) == 0 ? "\n" : "\r\n",
Expand All @@ -74,9 +83,9 @@ public void FuzzTarget(ReadOnlySpan<byte> bytes)
int maxExpandedSizeBytes = 6 * bytes.Length + 2;
byte[] expectedBuffer = ArrayPool<byte>.Shared.Rent(maxExpandedSizeBytes);
Span<byte> expected =
expectedBuffer.AsSpan(0, utf8
? EncodeToUtf8(bytes, expectedBuffer, options.Encoder)
: EncodeToUtf8(chars, expectedBuffer, options.Encoder));
expectedBuffer.AsSpan(0, encoding == Utf16EncodingFlag
? EncodeToUtf8(chars, expectedBuffer, options.Encoder)
: EncodeToUtf8(bytes, expectedBuffer, options.Encoder, encoding == Base64EncodingFlag));

// Compute the actual result by using Utf8JsonWriter. Each iteration is a different slice of the input, but the result should be the same.
byte[] actualBuffer = new byte[expected.Length];
Expand All @@ -89,14 +98,14 @@ public void FuzzTarget(ReadOnlySpan<byte> bytes)
{
using MemoryStream stream = new(actualBuffer);
using Utf8JsonWriter writer = new(stream, options);
if (utf8)

if (encoding == Utf16EncodingFlag)
{
WriteStringValueSegments(writer, bytes, ranges);
WriteStringValueSegments(writer, chars, ranges);
}
else
{
WriteStringValueSegments(writer, chars, ranges);
WriteStringValueSegments(writer, bytes, ranges, encoding == Base64EncodingFlag);
}

writer.Flush();
Expand All @@ -110,7 +119,7 @@ public void FuzzTarget(ReadOnlySpan<byte> bytes)
}

// Additional test for mixing UTF-8 and UTF-16 encoding. The alignment math is easier in UTF-16 mode so just run it for that.
if (!utf8)
if (encoding == Utf16EncodingFlag)
{
Array.Clear(expectedBuffer);

Expand All @@ -124,9 +133,16 @@ public void FuzzTarget(ReadOnlySpan<byte> bytes)
using MemoryStream stream = new(actualBuffer);
using Utf8JsonWriter writer = new(stream, options);

// UTF-16 + UTF-8
writer.WriteStringValueSegment(firstSegment, false);

Assert.Throws<InvalidOperationException, ReadOnlySpan<byte>>(state => writer.WriteStringValueSegment(state, true), secondSegment);

stream.Position = 0;
writer.Reset();

// UTF-16 + Base64
writer.WriteStringValueSegment(firstSegment, false);
Assert.Throws<InvalidOperationException, ReadOnlySpan<byte>>(state => writer.WriteBase64StringSegment(state, true), secondSegment);
}

Array.Clear(expectedBuffer);
Expand All @@ -135,25 +151,67 @@ public void FuzzTarget(ReadOnlySpan<byte> bytes)
ReadOnlySpan<byte> firstSegment = bytes[0..(2 * slice1)];
ReadOnlySpan<char> secondSegment = chars[slice1..];

expected = expectedBuffer.AsSpan(0, EncodeToUtf8(firstSegment, secondSegment, expectedBuffer, options.Encoder));
expected = expectedBuffer.AsSpan(0, EncodeToUtf8(firstSegment, expectedBuffer, options.Encoder, base64Encode: false));

actualBuffer = new byte[expected.Length];
using MemoryStream stream = new(actualBuffer);
using Utf8JsonWriter writer = new(stream, options);

// UTF-8 + UTF-16
writer.WriteStringValueSegment(firstSegment, false);
Assert.Throws<InvalidOperationException, ReadOnlySpan<char>>(state => writer.WriteStringValueSegment(state, true), secondSegment);

stream.Position = 0;
writer.Reset();

// UTF-8 + Base64
writer.WriteStringValueSegment(firstSegment, false);
Assert.Throws<InvalidOperationException, ReadOnlySpan<byte>>(state => writer.WriteBase64StringSegment(state, true), MemoryMarshal.AsBytes(secondSegment));
}

Array.Clear(expectedBuffer);

{
ReadOnlySpan<byte> firstSegment = bytes[0..(2 * slice1)];
ReadOnlySpan<char> secondSegment = chars[slice1..];

expected = expectedBuffer.AsSpan(0, EncodeToUtf8(firstSegment, expectedBuffer, options.Encoder, base64Encode: true));

actualBuffer = new byte[expected.Length];
using MemoryStream stream = new(actualBuffer);
using Utf8JsonWriter writer = new(stream, options);

// Base64 + UTF-16
writer.WriteBase64StringSegment(firstSegment, false);
Assert.Throws<InvalidOperationException, ReadOnlySpan<char>>(state => writer.WriteStringValueSegment(state, true), secondSegment);

stream.Position = 0;
writer.Reset();

// Base64 + UTF-8
writer.WriteBase64StringSegment(firstSegment, false);
Assert.Throws<InvalidOperationException, ReadOnlySpan<byte>>(state => writer.WriteStringValueSegment(state, true), MemoryMarshal.AsBytes(secondSegment));
}
}

ArrayPool<byte>.Shared.Return(expectedBuffer);
}

private static void WriteStringValueSegments(Utf8JsonWriter writer, ReadOnlySpan<byte> bytes, ReadOnlySpan<Range> ranges)
private static void WriteStringValueSegments(Utf8JsonWriter writer, ReadOnlySpan<byte> bytes, ReadOnlySpan<Range> ranges, bool base64Encode)
{
for (int i = 0; i < ranges.Length; i++)
if (base64Encode)
{
for (int i = 0; i < ranges.Length; i++)
{
writer.WriteBase64StringSegment(bytes[ranges[i]], i == ranges.Length - 1);
}
}
else
{
writer.WriteStringValueSegment(bytes[ranges[i]], i == ranges.Length - 1);
for (int i = 0; i < ranges.Length; i++)
{
writer.WriteStringValueSegment(bytes[ranges[i]], i == ranges.Length - 1);
}
}
}

Expand All @@ -165,10 +223,20 @@ private static void WriteStringValueSegments(Utf8JsonWriter writer, ReadOnlySpan
}
}

private static int EncodeToUtf8(ReadOnlySpan<byte> bytes, Span<byte> destBuffer, JavaScriptEncoder encoder)
private static int EncodeToUtf8(ReadOnlySpan<byte> bytes, Span<byte> destBuffer, JavaScriptEncoder encoder, bool base64Encode)
{
destBuffer[0] = (byte)'"';
encoder.EncodeUtf8(bytes, destBuffer[1..], out _, out int written, isFinalBlock: true);

int written;
if (base64Encode)
{
Base64.EncodeToUtf8(bytes, destBuffer[1..], out _, out written, isFinalBlock: true);
}
else
{
encoder.EncodeUtf8(bytes, destBuffer[1..], out _, out written, isFinalBlock: true);
}

destBuffer[++written] = (byte)'"';
return written + 1;
}
Expand All @@ -181,27 +249,6 @@ private static int EncodeToUtf8(ReadOnlySpan<char> chars, Span<byte> destBuffer,
return written + 1;
}

private static int EncodeToUtf8(ReadOnlySpan<byte> bytes, ReadOnlySpan<char> chars, Span<byte> destBuffer, JavaScriptEncoder encoder)
{
int written = 1;
destBuffer[0] = (byte)'"';
encoder.EncodeUtf8(bytes, destBuffer[1..], out _, out int writtenTemp, isFinalBlock: true);
written += writtenTemp;
destBuffer[written += EncodeTranscode(chars, destBuffer[written..], encoder, isFinalBlock: true)] = (byte)'"';
return written + 1;
}

private static int EncodeToUtf8(ReadOnlySpan<char> chars, ReadOnlySpan<byte> bytes, Span<byte> destBuffer, JavaScriptEncoder encoder)
{
int written = 1;
destBuffer[0] = (byte)'"';
written += EncodeTranscode(chars, destBuffer[1..], encoder, isFinalBlock: true);
encoder.EncodeUtf8(bytes, destBuffer[written..], out _, out int writtenTemp, isFinalBlock: true);
written += writtenTemp;
destBuffer[written] = (byte)'"';
return written + 1;
}

private static int EncodeTranscode(ReadOnlySpan<char> chars, Span<byte> destBuffer, JavaScriptEncoder encoder, bool isFinalBlock = true)
{
var utf16buffer = ArrayPool<char>.Shared.Rent(6 * chars.Length);
Expand Down
1 change: 1 addition & 0 deletions src/libraries/System.Text.Json/ref/System.Text.Json.cs
Original file line number Diff line number Diff line change
Expand Up @@ -681,6 +681,7 @@ public void WriteStringValue(string? value) { }
public void WriteStringValue(System.Text.Json.JsonEncodedText value) { }
public void WriteStringValueSegment(System.ReadOnlySpan<byte> value, bool isFinalSegment) { }
public void WriteStringValueSegment(System.ReadOnlySpan<char> value, bool isFinalSegment) { }
public void WriteBase64StringSegment(ReadOnlySpan<byte> value, bool isFinalSegment) { }
}
}
namespace System.Text.Json.Nodes
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -12,9 +12,9 @@ namespace System.Text.Json
{
public sealed partial class Utf8JsonWriter
{
private bool HasPartialCodePoint => PartialCodePointLength != 0;
private bool HasPartialStringData => PartialStringDataLength != 0;

private void ClearPartialCodePoint() => PartialCodePointLength = 0;
private void ClearPartialStringData() => PartialStringDataLength = 0;

private void ValidateEncodingDidNotChange(SegmentEncoding currentSegmentEncoding)
{
Expand All @@ -32,7 +32,7 @@ private void ValidateNotWithinUnfinalizedString()
}

Debug.Assert(PreviousSegmentEncoding == SegmentEncoding.None);
Debug.Assert(!HasPartialCodePoint);
Debug.Assert(!HasPartialStringData);
}

private void ValidateWritingValue()
Expand Down
Loading
Loading