Skip to content

Commit

Permalink
Avoid RegexCode/RegexWriter for all engines other than RegexInterpret…
Browse files Browse the repository at this point in the history 
…er (#65986)

* Avoid RegexCode/RegexWriter for all engines other than RegexInterpreter

* Address PR feedback
  • Loading branch information
@stephentoub
stephentoub authored Mar 2, 2022
1 parent 2330b4b commit 093bdc4
Show file tree
Hide file tree
Showing 28 changed files with 433 additions and 480 deletions.
105 changes: 44 additions & 61 deletions src/libraries/System.Text.RegularExpressions/gen/RegexGenerator.Emitter.cs
View file

Large diffs are not rendered by default.

8 changes: 4 additions & 4 deletions src/libraries/System.Text.RegularExpressions/gen/RegexGenerator.Parser.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -178,10 +178,10 @@ private static bool IsSemanticTargetForGeneration(SemanticModel semanticModel, M
}

// Parse the input pattern
RegexCode code;
RegexTree tree;
try
{
code = RegexWriter.Write(RegexParser.Parse(pattern, regexOptions, culture), culture);
tree = RegexParser.Parse(pattern, regexOptions, culture);
}
catch (Exception e)
{
Expand All @@ -199,7 +199,7 @@ private static bool IsSemanticTargetForGeneration(SemanticModel semanticModel, M
pattern,
regexOptions,
matchTimeout ?? Timeout.Infinite,
code);
tree);

var regexType = new RegexType(
regexMethod,
Expand Down Expand Up @@ -233,7 +233,7 @@ static bool IsAllowedKind(SyntaxKind kind) =>
}

/// <summary>A regex method.</summary>
internal sealed record RegexMethod(MethodDeclarationSyntax MethodSyntax, string MethodName, string Modifiers, string Pattern, RegexOptions Options, int MatchTimeout, RegexCode Code);
internal sealed record RegexMethod(MethodDeclarationSyntax MethodSyntax, string MethodName, string Modifiers, string Pattern, RegexOptions Options, int MatchTimeout, RegexTree Tree);

/// <summary>A type holding a regex method.</summary>
internal sealed record RegexType(RegexMethod? Method, string Keyword, string Namespace, string Name)
Expand Down
2 changes: 0 additions & 2 deletions ...raries/System.Text.RegularExpressions/gen/System.Text.RegularExpressions.Generator.csproj
View file
Original file line number Diff line number Diff line change
Expand Up @@ -33,7 +33,6 @@
<Compile Include="..\src\System\Threading\StackHelper.cs" Link="Production\StackHelper.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexCharClass.cs" Link="Production\RegexCharClass.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexCharClass.MappingTable.cs" Link="Production\RegexCharClass.MappingTable.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexCode.cs" Link="Production\RegexCode.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexFindOptimizations.cs" Link="Production\RegexFindOptimizations.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexNode.cs" Link="Production\RegexNode.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexNodeKind.cs" Link="Production\RegexNodeKind.cs" />
Expand All @@ -45,7 +44,6 @@
<Compile Include="..\src\System\Text\RegularExpressions\RegexPrefixAnalyzer.cs" Link="Production\RegexPrefixAnalyzer.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexTree.cs" Link="Production\RegexTree.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexTreeAnalyzer.cs" Link="Production\RegexTreeAnalyzer.cs" />
<Compile Include="..\src\System\Text\RegularExpressions\RegexWriter.cs" Link="Production\RegexWriter.cs" />
<Compile Include="..\src\System\Collections\HashtableExtensions.cs" Link="Production\HashtableExtensions.cs" />
</ItemGroup>

Expand Down
2 changes: 1 addition & 1 deletion src/libraries/System.Text.RegularExpressions/src/System.Text.RegularExpressions.csproj
View file
Original file line number Diff line number Diff line change
Expand Up @@ -26,11 +26,11 @@
<Compile Include="System\Text\RegularExpressions\Regex.Timeout.cs" />
<Compile Include="System\Text\RegularExpressions\RegexCharClass.cs" />
<Compile Include="System\Text\RegularExpressions\RegexCharClass.MappingTable.cs" />
<Compile Include="System\Text\RegularExpressions\RegexCode.cs" />
<Compile Include="System\Text\RegularExpressions\RegexCompilationInfo.cs" />
<Compile Include="System\Text\RegularExpressions\RegexFindOptimizations.cs" />
<Compile Include="System\Text\RegularExpressions\RegexGeneratorAttribute.cs" />
<Compile Include="System\Text\RegularExpressions\RegexInterpreter.cs" />
<Compile Include="System\Text\RegularExpressions\RegexInterpreterCode.cs" />
<Compile Include="System\Text\RegularExpressions\RegexMatchTimeoutException.cs" />
<Compile Include="System\Text\RegularExpressions\RegexNode.cs" />
<Compile Include="System\Text\RegularExpressions\RegexNodeKind.cs" />
Expand Down
10 changes: 5 additions & 5 deletions ...ibraries/System.Text.RegularExpressions/src/System/Text/RegularExpressions/Regex.Match.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -77,7 +77,7 @@ public bool IsMatch(string input)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Run(quick: true, -1, input, 0, input.Length, UseOptionR() ? input.Length : 0) is null;
return Run(quick: true, -1, input, 0, input.Length, RightToLeft ? input.Length : 0) is null;
}

/// <summary>
Expand All @@ -87,7 +87,7 @@ public bool IsMatch(string input)
/// <returns><see langword="true"/> if the regular expression finds a match; otherwise, <see langword="false"/>.</returns>
/// <exception cref="RegexMatchTimeoutException">A time-out ocurred.</exception>
public bool IsMatch(ReadOnlySpan<char> input) =>
Run(input, UseOptionR() ? input.Length : 0) is null;
Run(input, RightToLeft ? input.Length : 0) is null;

/// <summary>
/// Searches the input string for one or more matches using the previous pattern and options,
Expand Down Expand Up @@ -132,7 +132,7 @@ public Match Match(string input)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Run(quick: false, -1, input, 0, input.Length, UseOptionR() ? input.Length : 0)!;
return Run(quick: false, -1, input, 0, input.Length, RightToLeft ? input.Length : 0)!;
}

/// <summary>
Expand All @@ -159,7 +159,7 @@ public Match Match(string input, int beginning, int length)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Run(quick: false, -1, input, beginning, length, UseOptionR() ? beginning + length : beginning)!;
return Run(quick: false, -1, input, beginning, length, RightToLeft ? beginning + length : beginning)!;
}

/// <summary>
Expand Down Expand Up @@ -187,7 +187,7 @@ public MatchCollection Matches(string input)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return new MatchCollection(this, input, UseOptionR() ? input.Length : 0);
return new MatchCollection(this, input, RightToLeft ? input.Length : 0);
}

/// <summary>
Expand Down
8 changes: 4 additions & 4 deletions ...raries/System.Text.RegularExpressions/src/System/Text/RegularExpressions/Regex.Replace.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,7 @@ public string Replace(string input, string replacement)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Replace(input, replacement, -1, UseOptionR() ? input.Length : 0);
return Replace(input, replacement, -1, RightToLeft ? input.Length : 0);
}

/// <summary>
Expand All @@ -57,7 +57,7 @@ public string Replace(string input, string replacement, int count)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Replace(input, replacement, count, UseOptionR() ? input.Length : 0);
return Replace(input, replacement, count, RightToLeft ? input.Length : 0);
}

/// <summary>
Expand Down Expand Up @@ -111,7 +111,7 @@ public string Replace(string input, MatchEvaluator evaluator)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Replace(evaluator, this, input, -1, UseOptionR() ? input.Length : 0);
return Replace(evaluator, this, input, -1, RightToLeft ? input.Length : 0);
}

/// <summary>
Expand All @@ -125,7 +125,7 @@ public string Replace(string input, MatchEvaluator evaluator, int count)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Replace(evaluator, this, input, count, UseOptionR() ? input.Length : 0);
return Replace(evaluator, this, input, count, RightToLeft ? input.Length : 0);
}

/// <summary>
Expand Down
4 changes: 2 additions & 2 deletions ...ibraries/System.Text.RegularExpressions/src/System/Text/RegularExpressions/Regex.Split.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,7 @@ public string[] Split(string input)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Split(this, input, 0, UseOptionR() ? input.Length : 0);
return Split(this, input, 0, RightToLeft ? input.Length : 0);
}

/// <summary>
Expand All @@ -49,7 +49,7 @@ public string[] Split(string input, int count)
ThrowHelper.ThrowArgumentNullException(ExceptionArgument.input);
}

return Split(this, input, count, UseOptionR() ? input.Length : 0);
return Split(this, input, count, RightToLeft ? input.Length : 0);
}

/// <summary>
Expand Down
99 changes: 51 additions & 48 deletions src/libraries/System.Text.RegularExpressions/src/System/Text/RegularExpressions/Regex.cs
View file
Original file line number Diff line number Diff line change
Expand Up @@ -20,8 +20,6 @@ namespace System.Text.RegularExpressions
/// </summary>
public partial class Regex : ISerializable
{
internal const int MaxOptionShift = 11;

[StringSyntax(StringSyntaxAttribute.Regex)]
protected internal string? pattern; // The string pattern provided
protected internal RegexOptions roptions; // the top-level options from the options string
Expand All @@ -33,7 +31,6 @@ public partial class Regex : ISerializable

private WeakReference<RegexReplacement?>? _replref; // cached parsed replacement pattern
private volatile RegexRunner? _runner; // cached runner
private RegexCode? _code; // if interpreted, this is the code for RegexInterpreter

protected Regex()
{
Expand Down Expand Up @@ -63,64 +60,69 @@ public Regex([StringSyntax(StringSyntaxAttribute.Regex, "options")] string patte

internal Regex(string pattern, CultureInfo? culture)
{
// Call Init directly rather than delegating to a Regex ctor that takes
// options to enable linking / tree shaking to remove the Regex compiler
// and NonBacktracking implementation if it's not used.
Init(pattern, RegexOptions.None, s_defaultMatchTimeout, culture ?? CultureInfo.CurrentCulture);
// Validate arguments.
ValidatePattern(pattern);

// Parse and store the argument information.
RegexTree tree = Init(pattern, RegexOptions.None, s_defaultMatchTimeout, ref culture);

// Create the interpreter factory.
factory = new RegexInterpreterFactory(tree, culture);

// NOTE: This overload _does not_ delegate to the one that takes options, in order
// to avoid unnecessarily rooting the support for RegexOptions.NonBacktracking/Compiler
// if no options are ever used.
}

internal Regex(string pattern, RegexOptions options, TimeSpan matchTimeout, CultureInfo? culture)
{
culture ??= RegexParser.GetTargetCulture(options);
Init(pattern, options, matchTimeout, culture);
// Validate arguments.
ValidatePattern(pattern);
ValidateOptions(options);
ValidateMatchTimeout(matchTimeout);

// Parse and store the argument information.
RegexTree tree = Init(pattern, options, matchTimeout, ref culture);

// Create the appropriate factory.
if ((options & RegexOptions.NonBacktracking) != 0)
{
// If we're in non-backtracking mode, create the appropriate factory.
factory = new SymbolicRegexRunnerFactory(_code, options, matchTimeout, culture);
_code = null;
factory = new SymbolicRegexRunnerFactory(tree, options, matchTimeout, culture);
}
else if (RuntimeFeature.IsDynamicCodeCompiled && UseOptionC())
else
{
// If the compile option is set and compilation is supported, then compile the code.
// If the compiler can't compile this regex, it'll return null, and we'll fall back
// to the interpreter.
factory = Compile(pattern, _code, options, matchTimeout != InfiniteMatchTimeout);
if (factory is not null)
if (RuntimeFeature.IsDynamicCodeCompiled && (options & RegexOptions.Compiled) != 0)
{
_code = null;
// If the compile option is set and compilation is supported, then compile the code.
// If the compiler can't compile this regex, it'll return null, and we'll fall back
// to the interpreter.
factory = Compile(pattern, tree, options, matchTimeout != InfiniteMatchTimeout);
}

// If no factory was created, fall back to creating one for the interpreter.
factory ??= new RegexInterpreterFactory(tree, culture);
}
}

/// <summary>Initializes the instance.</summary>
/// <remarks>
/// This is separated out of the constructor so that an app only using 'new Regex(pattern)'
/// rather than 'new Regex(pattern, options)' can avoid statically referencing the Regex
/// compiler, such that a tree shaker / linker can trim it away if it's not otherwise used.
/// </remarks>
[MemberNotNull(nameof(_code))]
private void Init(string pattern, RegexOptions options, TimeSpan matchTimeout, CultureInfo culture)
/// <summary>Stores the supplied arguments and capture information, returning the parsed expression.</summary>
private RegexTree Init(string pattern, RegexOptions options, TimeSpan matchTimeout, [NotNull] ref CultureInfo? culture)
{
ValidatePattern(pattern);
ValidateOptions(options);
ValidateMatchTimeout(matchTimeout);

this.pattern = pattern;
internalMatchTimeout = matchTimeout;
roptions = options;
internalMatchTimeout = matchTimeout;
culture ??= RegexParser.GetTargetCulture(options);

// Parse the input
RegexTree tree = RegexParser.Parse(pattern, roptions, culture);
// Parse the pattern.
RegexTree tree = RegexParser.Parse(pattern, options, culture);

// Generate the RegexCode from the node tree. This is required for interpreting,
// and is used as input into RegexOptions.Compiled and RegexOptions.NonBacktracking.
_code = RegexWriter.Write(tree, culture);
// Store the relevant information, constructing the appropriate factory.
capnames = tree.CaptureNameToNumberMapping;
capslist = tree.CaptureNames;
caps = tree.CaptureNumberSparseMapping;
capsize = tree.CaptureCount;

capnames = tree.CapNames;
capslist = tree.CapsList;
caps = _code.Caps;
capsize = _code.CapSize;
return tree;
}

internal static void ValidatePattern(string pattern)
Expand All @@ -133,9 +135,9 @@ internal static void ValidatePattern(string pattern)

internal static void ValidateOptions(RegexOptions options)
{
const int MaxOptionShift = 11;
if (((((uint)options) >> MaxOptionShift) != 0) ||
((options & RegexOptions.ECMAScript) != 0 &&
(options & ~(RegexOptions.ECMAScript | RegexOptions.IgnoreCase | RegexOptions.Multiline | RegexOptions.Compiled | RegexOptions.NonBacktracking | RegexOptions.CultureInvariant)) != 0))
((options & RegexOptions.ECMAScript) != 0 && (options & ~(RegexOptions.ECMAScript | RegexOptions.IgnoreCase | RegexOptions.Multiline | RegexOptions.Compiled | RegexOptions.NonBacktracking | RegexOptions.CultureInvariant)) != 0))
{
ThrowHelper.ThrowArgumentOutOfRangeException(ExceptionArgument.options);
}
Expand Down Expand Up @@ -199,8 +201,8 @@ protected IDictionary? CapNames
/// instantiating a non-compiled regex.
/// </summary>
[MethodImpl(MethodImplOptions.NoInlining)]
private static RegexRunnerFactory? Compile(string pattern, RegexCode code, RegexOptions options, bool hasTimeout) =>
RegexCompiler.Compile(pattern, code, options, hasTimeout);
private static RegexRunnerFactory? Compile(string pattern, RegexTree regexTree, RegexOptions options, bool hasTimeout) =>
RegexCompiler.Compile(pattern, regexTree, options, hasTimeout);

[Obsolete(Obsoletions.RegexCompileToAssemblyMessage, DiagnosticId = Obsoletions.RegexCompileToAssemblyDiagId, UrlFormat = Obsoletions.SharedUrlFormat)]
public static void CompileToAssembly(RegexCompilationInfo[] regexinfos, AssemblyName assemblyname) =>
Expand Down Expand Up @@ -254,7 +256,7 @@ public static string Unescape(string str)
/// <summary>
/// Indicates whether the regular expression matches from right to left.
/// </summary>
public bool RightToLeft => UseOptionR();
public bool RightToLeft => (roptions & RegexOptions.RightToLeft) != 0;

/// <summary>
/// Returns the regular expression pattern passed into the constructor
Expand Down Expand Up @@ -554,13 +556,14 @@ internal void Run<TState>(string input, int startat, ref TState state, MatchCall

/// <summary>Creates a new runner instance.</summary>
private RegexRunner CreateRunner() =>
factory?.CreateInstance() ??
new RegexInterpreter(_code!, RegexParser.GetTargetCulture(roptions));
// The factory needs to be set by the ctor. `factory` is a protected field, so it's possible a derived
// type nulls out the factory after we've set it, but that's the nature of the design.
factory!.CreateInstance();

/// <summary>True if the <see cref="RegexOptions.Compiled"/> option was set.</summary>
protected bool UseOptionC() => (roptions & RegexOptions.Compiled) != 0;

/// <summary>True if the <see cref="RegexOptions.RightToLeft"/> option was set.</summary>
protected internal bool UseOptionR() => (roptions & RegexOptions.RightToLeft) != 0;
protected internal bool UseOptionR() => RightToLeft;
}
}
Loading

0 comments on commit 093bdc4

Please sign in to comment.