chore: modify regex settings (#10539)

* chore: modify regex settings

* chore: fix ci error of plantuml test

Author: filzrev

src/Docfx.Build/HtmlTemplate.cs

@@ -8,8 +8,11 @@
 
 namespace Docfx.Build;
 
-struct HtmlTemplate
+partial struct HtmlTemplate
 {
+    [GeneratedRegex("(\\s+[a-zA-Z0-9_-]+)=([\"']){(\\d)}[\"']")]
+    private static partial Regex AttributePlaceholderRegex();
+
     private string? _html;
 
     public override string ToString() => _html ?? "";
@@ -18,7 +21,7 @@ struct HtmlTemplate
 
     public static HtmlTemplate Html(FormattableString template)
     {
-        var format = Regex.Replace(template.Format, "(\\s+[a-zA-Z0-9_-]+)=([\"']){(\\d)}[\"']", RenderAttribute);
+        var format = AttributePlaceholderRegex().Replace(template.Format, RenderAttribute);
         var html = string.Format(format, Array.ConvertAll(template.GetArguments(), Render));
         return new() { _html = html };
 

src/Docfx.Build/ResourceFileReaders/ResourceFileReader.cs

@@ -36,8 +36,7 @@ bool filter(string s)
         {
             if (selector != null)
             {
-                var regex = new Regex(selector, RegexOptions.IgnoreCase);
-                return regex.IsMatch(s);
+                return Regex.IsMatch(s, selector, RegexOptions.IgnoreCase);
             }
             else
             {

src/Docfx.Build/TemplateProcessors/ViewRenderers/MustacheTemplateRenderer.cs

@@ -11,13 +11,18 @@ internal partial class MustacheTemplateRenderer : ITemplateRenderer
 {
     public const string Extension = ".tmpl";
 
-    [GeneratedRegex(@"{{\s*!\s*include\s*\(:?(:?['""]?)\s*(?<file>(.+?))\1\s*\)\s*}}", RegexOptions.IgnoreCase | RegexOptions.Compiled, "en-AU")]
-    private static partial Regex IncludeRegex();
 
-    [GeneratedRegex(@"{{\s*!\s*master\s*\(:?(:?['""]?)\s*(?<file>(.+?))\1\s*\)\s*}}\s*\n?", RegexOptions.IgnoreCase | RegexOptions.Compiled, "en-AU")]
-    private static partial Regex MasterPageRegex();
-
-    [GeneratedRegex(@"{{\s*!\s*body\s*}}\s*\n?", RegexOptions.IgnoreCase | RegexOptions.Compiled, "en-AU")]
+    // Following regex are not supported by GeneratedRegexAttribute. (Because it contains case-insensitive backreferences)
+    // When using GeneratedRegexAttribute. following message are shown as information leve.
+    //  SYSLIB1044: The regex generator couldn't generate a complete source implementation for the specified regular expression due to an internal limitation. See the explanation in the generated source for more details.
+    // And Regex instance is created with following remarks comments.
+    //   A custom Regex-derived type could not be generated because the expression contains case-insensitive backreferences which are not supported by the source generator.
+#pragma warning disable SYSLIB1045
+    private static readonly Regex IncludeRegex = new(@"{{\s*!\s*include\s*\(:?(:?['""]?)\s*(?<file>(.+?))\1\s*\)\s*}}", RegexOptions.Compiled | RegexOptions.IgnoreCase);
+    private static readonly Regex MasterPageRegex = new(@"{{\s*!\s*master\s*\(:?(:?['""]?)\s*(?<file>(.+?))\1\s*\)\s*}}\s*\n?", RegexOptions.Compiled | RegexOptions.IgnoreCase);
+#pragma warning restore SYSLIB1045
+
+    [GeneratedRegex(@"{{\s*!\s*body\s*}}\s*\n?", RegexOptions.IgnoreCase)]
     private static partial Regex MasterPageBodyRegex();
 
     private readonly ResourceFileReader _reader;
@@ -42,7 +47,7 @@ public MustacheTemplateRenderer(ResourceFileReader reader, ResourceInfo info, st
             })
             .Build();
 
-        var processedTemplate = ParseTemplateHelper.ExpandMasterPage(reader, info, MasterPageRegex(), MasterPageBodyRegex());
+        var processedTemplate = ParseTemplateHelper.ExpandMasterPage(reader, info, MasterPageRegex, MasterPageBodyRegex());
 
         _template = processedTemplate;
 
@@ -68,7 +73,7 @@ public string Render(object model)
     /// <param name="template"></param>
     private IEnumerable<string> ExtractDependencyResourceNames(string template)
     {
-        foreach (Match match in IncludeRegex().Matches(template))
+        foreach (Match match in IncludeRegex.Matches(template))
         {
             var filePath = match.Groups["file"].Value;
             foreach (var name in ParseTemplateHelper.GetResourceName(filePath, Path, _reader))

src/Docfx.Build/TemplateProcessors/ViewRenderers/ParseTemplateHelper.cs

@@ -78,10 +78,9 @@ public static IEnumerable<string> GetResourceName(string file, string templateNa
         {
             file = regexPatternMatch.Groups[1].Value;
             var resourceKey = GetRelativeResourceKey(templateName, file);
-            var regex = new Regex(resourceKey, RegexOptions.IgnoreCase);
             foreach (var name in reader.Names)
             {
-                if (regex.IsMatch(name))
+                if (Regex.IsMatch(name, resourceKey, RegexOptions.IgnoreCase))
                 {
                     yield return name;
                 }

src/Docfx.Dotnet/DotnetApiCatalog.ApiPage.cs

@@ -18,6 +18,10 @@ namespace Docfx.Dotnet;
 
 partial class DotnetApiCatalog
 {
+    // Regex to match any character other than a word character(alphabet/numeric/underscore)
+    [GeneratedRegex(@"\W")]
+    private static partial Regex NonWordCharRegex();
+
     private static void CreatePages(Action<string, string, ApiPage> output, List<(IAssemblySymbol symbol, Compilation compilation)> assemblies, ExtractMetadataConfig config, DotnetApiOptions options)
     {
         Directory.CreateDirectory(config.OutputFolder);
@@ -122,7 +126,7 @@ void Heading(int level, string title, string? id = null)
             void Api(int level, string title, ISymbol symbol, Compilation compilation)
             {
                 var uid = VisitorHelper.GetId(symbol);
-                var id = Regex.Replace(uid, @"\W", "_");
+                var id = NonWordCharRegex().Replace(uid, "_");
                 var commentId = VisitorHelper.GetCommentId(symbol);
                 var source = config.DisableGitFeatures ? null : VisitorHelper.GetSourceDetail(symbol, compilation);
                 var git = source?.Remote is null ? null

src/Docfx.Dotnet/Filters/ConfigFilterRuleItem.cs

@@ -20,7 +20,7 @@ public string UidRegex
         }
         set
         {
-            _uidRegex = new Regex(value);
+            _uidRegex = new Regex(value, RegexOptions.Compiled);
         }
     }
 

src/Docfx.Dotnet/Parsers/XmlComment.cs

@@ -38,6 +38,9 @@ internal partial class XmlComment
     [GeneratedRegex(@"^\s*<!--\s*</(.*)>\s*-->$")]
     private static partial Regex XmlEndRegionRegex();
 
+    [GeneratedRegex(@"^(\s*)&gt;", RegexOptions.Multiline)]
+    private static partial Regex BlockQuoteRegex();
+
     private readonly XmlCommentParserContext _context;
 
     public string Summary { get; private set; }
@@ -603,7 +606,7 @@ static string HandleBlockQuote(string xml)
         {
             // > is encoded to &gt; in XML. When interpreted as markdown, > is as blockquote
             // Decode standalone &gt; to > to enable the block quote markdown syntax
-            return Regex.Replace(xml, @"^(\s*)&gt;", "$1>", RegexOptions.Multiline);
+            return BlockQuoteRegex().Replace(xml, "$1>");
         }
 
         static void MarkdownXmlDecode(MarkdownObject node)

src/Docfx.Dotnet/SymbolUrlResolver.cs

@@ -16,6 +16,10 @@ enum SymbolUrlKind
 
 internal static partial class SymbolUrlResolver
 {
+    // Regex to match any character other than a word character(alphabet/numeric/underscore)
+    [GeneratedRegex(@"\W")]
+    private static partial Regex NonWordCharRegex();
+
     public static string? GetSymbolUrl(ISymbol symbol, Compilation compilation, MemberLayout memberLayout, SymbolUrlKind urlKind, HashSet<IAssemblySymbol> allAssemblies, SymbolFilter filter)
     {
         // Reduce symbol into generic definitions
@@ -53,8 +57,8 @@ internal static partial class SymbolUrlResolver
             "!" => null,
             "N" or "T" => $"{VisitorHelper.PathFriendlyId(uid)}{ext}",
             "M" or "F" or "P" or "E" => memberLayout is MemberLayout.SeparatePages && !symbol.IsEnumMember()
-                ? $"{VisitorHelper.PathFriendlyId(VisitorHelper.GetOverloadId(symbol))}{ext}#{Regex.Replace(uid, @"\W", "_")}"
-                : $"{VisitorHelper.PathFriendlyId(VisitorHelper.GetId(symbol.ContainingType))}{ext}#{Regex.Replace(uid, @"\W", "_")}",
+                ? $"{VisitorHelper.PathFriendlyId(VisitorHelper.GetOverloadId(symbol))}{ext}#{NonWordCharRegex().Replace(uid, "_")}"
+                : $"{VisitorHelper.PathFriendlyId(VisitorHelper.GetId(symbol.ContainingType))}{ext}#{NonWordCharRegex().Replace(uid, "_")}",
             _ => throw new NotSupportedException($"Unknown comment ID format '{type}'"),
         };
     }

src/Docfx.MarkdigEngine.Extensions/CodeSnippet/CodeSnippetExtractor.cs

@@ -9,7 +9,7 @@ namespace Docfx.MarkdigEngine.Extensions;
 
 public partial class CodeSnippetExtractor
 {
-    [GeneratedRegex(@"^[\w\.-]+$", RegexOptions.IgnoreCase, "en-AU")]
+    [GeneratedRegex(@"^[\w\.-]+$", RegexOptions.IgnoreCase)]
     private static partial Regex TagnameFormat();
 
     private readonly string StartLineTemplate;

test/.editorconfig

@@ -4,3 +4,5 @@ root = false
 [*.cs]
 csharp_style_unused_value_expression_statement_preference = discard_variable:silent # IDE0058: Remove unnecessary expression value
 csharp_style_unused_value_assignment_preference           = discard_variable:silent # IDE0059: Remove unnecessary value assignment
+
+dotnet_diagnostic.SYSLIB1045.severity = silent  # SYSLIB1045: Convert to 'GeneratedRegexAttribute'.