C#: Adjust trap location, database ID and archiving of generated sources

csharp/extractor/Semmle.Extraction.CSharp/Entities/File.cs

@@ -34,7 +34,14 @@ namespace Semmle.Extraction.CSharp.Entities
                         lineCounts.Total++;
 
                     trapFile.numlines(this, lineCounts);
-                    Context.TrapWriter.Archive(originalPath, TransformedPath, text.Encoding ?? System.Text.Encoding.Default);
+                    if (BinaryLogExtractionContext.GetAdjustedPath(Context.ExtractionContext, originalPath) is not null)
+                    {
+                        Context.TrapWriter.ArchiveContent(rawText, TransformedPath);
+                    }
+                    else
+                    {
+                        Context.TrapWriter.Archive(originalPath, TransformedPath, text.Encoding ?? System.Text.Encoding.Default);
+                    }
                 }
             }
             else if (IsPossiblyTextFile())

csharp/extractor/Semmle.Extraction.CSharp/Extractor/Analyser.cs

@@ -185,7 +185,8 @@ namespace Semmle.Extraction.CSharp
             {
                 var stopwatch = new Stopwatch();
                 stopwatch.Start();
-                var sourcePath = tree.FilePath;
+                var sourcePath = BinaryLogExtractionContext.GetAdjustedPath(ExtractionContext, tree.FilePath) ?? tree.FilePath;
+
                 var transformedSourcePath = PathTransformer.Transform(sourcePath);
 
                 var trapPath = transformedSourcePath.GetTrapPath(Logger, options.TrapCompression);

csharp/extractor/Semmle.Extraction.CSharp/Extractor/BinaryLogAnalyser.cs

@@ -1,3 +1,4 @@
+using System.Collections.Generic;
 using Microsoft.CodeAnalysis.CSharp;
 using Semmle.Util;
 using Semmle.Util.Logging;
@@ -11,10 +12,15 @@ namespace Semmle.Extraction.CSharp
         {
         }
 
-        public void Initialize(string cwd, string[] args, string outputPath, CSharpCompilation compilationIn, CommonOptions options)
+        public void Initialize(
+            string cwd, string[] args, string outputPath, CSharpCompilation compilation,
+            IEnumerable<Microsoft.CodeAnalysis.SyntaxTree> generatedSyntaxTrees,
+            string compilationIdentifier, CommonOptions options)
         {
-            compilation = compilationIn;
-            ExtractionContext = new ExtractionContext(cwd, args, outputPath, [], Logger, PathTransformer, ExtractorMode.BinaryLog, options.QlTest);
+            base.compilation = compilation;
+            ExtractionContext = new BinaryLogExtractionContext(
+                cwd, args, outputPath, generatedSyntaxTrees, compilationIdentifier,
+                Logger, PathTransformer, options.QlTest);
             this.options = options;
             LogExtractorInfo();
             SetReferencePaths();

csharp/extractor/Semmle.Extraction.CSharp/Extractor/Extractor.cs

@@ -158,13 +158,23 @@ namespace Semmle.Extraction.CSharp
                     var compilerArgs = compilerCall.GetArguments();
                     var args = reader.ReadCommandLineArguments(compilerCall);
 
+                    // Generated syntax trees are always added to the end of the list of syntax trees.
+                    var generatedSyntaxTrees = compilation.SyntaxTrees.Skip(compilationData.Compilation.SyntaxTrees.Count());
+
                     using var analyser = new BinaryLogAnalyser(new LogProgressMonitor(logger), logger, pathTransformer, canonicalPathCache, options.AssemblySensitiveTrap);
 
                     var exit = Analyse(stopwatch, analyser, options,
                         references => [() => compilation.References.ForEach(r => references.Add(r))],
                         (analyser, syntaxTrees) => [() => syntaxTrees.AddRange(compilation.SyntaxTrees)],
                         (syntaxTrees, references) => compilation,
-                        (compilation, options) => analyser.Initialize(compilerCall.ProjectDirectory, compilerArgs?.ToArray() ?? [], TracingAnalyser.GetOutputName(compilation, args), compilation, options),
+                        (compilation, options) => analyser.Initialize(
+                            compilerCall.ProjectDirectory,
+                            compilerArgs?.ToArray() ?? [],
+                            TracingAnalyser.GetOutputName(compilation, args),
+                            compilation,
+                            generatedSyntaxTrees,
+                            diagnosticName,
+                            options),
                         () => { });
 
                     switch (exit)

csharp/extractor/Semmle.Extraction/Entities/File.cs

@@ -8,7 +8,8 @@ namespace Semmle.Extraction.Entities
             : base(cx, path)
         {
             originalPath = path;
-            transformedPathLazy = new Lazy<PathTransformer.ITransformedPath>(() => Context.ExtractionContext.PathTransformer.Transform(originalPath));
+            var adjustedPath = BinaryLogExtractionContext.GetAdjustedPath(Context.ExtractionContext, originalPath) ?? path;
+            transformedPathLazy = new Lazy<PathTransformer.ITransformedPath>(() => Context.ExtractionContext.PathTransformer.Transform(adjustedPath));
         }
 
         protected readonly string originalPath;

csharp/extractor/Semmle.Extraction/Extractor/BinaryLogExtractionContext.cs

@@ -0,0 +1,60 @@
+using System.Collections.Generic;
+using System.IO;
+using System.Linq;
+using Microsoft.CodeAnalysis;
+using Semmle.Util.Logging;
+
+namespace Semmle.Extraction
+{
+    public class BinaryLogExtractionContext : ExtractionContext
+    {
+        private readonly IEnumerable<SyntaxTree> generatedSyntaxTrees;
+        private readonly string compilationIdentifier;
+        private readonly string generatedFolderName;
+
+        public BinaryLogExtractionContext(string cwd, string[] args, string outputPath,
+            IEnumerable<SyntaxTree> generatedSyntaxTrees, string compilationIdentifier,
+            ILogger logger, PathTransformer pathTransformer, bool isQlTest)
+            : base(cwd, args, outputPath, [], logger, pathTransformer, ExtractorMode.BinaryLog, isQlTest)
+        {
+            this.generatedSyntaxTrees = generatedSyntaxTrees;
+            this.compilationIdentifier = compilationIdentifier;
+
+            // Compute a unique folder name for the generated files:
+            generatedFolderName = "generated";
+
+            if (Directory.Exists(generatedFolderName))
+            {
+                var counter = 0;
+                do
+                {
+                    generatedFolderName = $"generated{counter++}";
+                }
+                while (Directory.Exists(generatedFolderName));
+            }
+        }
+
+        private string? GetAdjustedPath(string path)
+        {
+            var syntaxTree = generatedSyntaxTrees.FirstOrDefault(t => t.FilePath == path);
+            if (syntaxTree is null)
+            {
+                return null;
+            }
+
+            return Path.Join(generatedFolderName, compilationIdentifier, path);
+        }
+
+        public static string? GetAdjustedPath(ExtractionContext extractionContext, string sourcePath)
+        {
+            if (extractionContext.Mode.HasFlag(ExtractorMode.BinaryLog)
+                && extractionContext is BinaryLogExtractionContext binaryLogExtractionContext
+                && binaryLogExtractionContext.GetAdjustedPath(sourcePath) is string adjustedPath)
+            {
+                return adjustedPath;
+            }
+
+            return null;
+        }
+    }
+}

csharp/extractor/Semmle.Extraction/TrapWriter.cs

@@ -105,14 +105,42 @@ namespace Semmle.Extraction
         /// <param name="transformedPath">The transformed path to the input file.</param>
         /// <param name="inputEncoding">The encoding used by the input file.</param>
         public void Archive(string originalPath, PathTransformer.ITransformedPath transformedPath, Encoding inputEncoding)
+        {
+            Archive(() =>
+            {
+                var fullInputPath = Path.GetFullPath(originalPath);
+                return File.ReadAllText(fullInputPath, inputEncoding);
+            }, transformedPath);
+        }
+
+        public void ArchiveContent(string contents, PathTransformer.ITransformedPath transformedPath)
+        {
+            Archive(() => contents, transformedPath);
+        }
+
+        private void Archive(Func<string> getContent, PathTransformer.ITransformedPath transformedPath)
         {
             if (string.IsNullOrEmpty(archive))
+            {
                 return;
+            }
 
-            // Calling GetFullPath makes this use the canonical capitalisation, if the file exists.
-            var fullInputPath = Path.GetFullPath(originalPath);
+            var dest = FileUtils.NestPaths(logger, archive, transformedPath.Value);
+            try
+            {
+                var tmpSrcFile = Path.GetTempFileName();
+                File.WriteAllText(tmpSrcFile, getContent(), utf8);
 
-            ArchivePath(fullInputPath, transformedPath, inputEncoding);
+                FileUtils.MoveOrReplace(tmpSrcFile, dest);
+            }
+            catch (Exception ex)
+            {
+                // If this happened, it was probably because
+                // - the same file was compiled multiple times, or
+                // - the file doesn't exist (due to wrong #line directive or because it's an in-memory source generated AST).
+                // In any case, this is not a fatal error.
+                logger.LogWarning("Problem archiving " + dest + ": " + ex);
+            }
         }
 
         /// <summary>
@@ -185,37 +213,6 @@ namespace Semmle.Extraction
             emitter.EmitTrap(Writer);
         }
 
-        /// <summary>
-        /// Attempts to archive the specified input file to the normal area of the source archive.
-        /// The file's path must be sufficiently short so as to render the path of its copy in the
-        /// source archive less than the system path limit of 260 characters.
-        /// </summary>
-        /// <param name="fullInputPath">The full path to the input file.</param>
-        /// <param name="transformedPath">The transformed path to the input file.</param>
-        /// <param name="inputEncoding">The encoding used by the input file.</param>
-        /// <exception cref="PathTooLongException">If the output path in the source archive would
-        /// exceed the system path limit of 260 characters.</exception>
-        private void ArchivePath(string fullInputPath, PathTransformer.ITransformedPath transformedPath, Encoding inputEncoding)
-        {
-            var dest = FileUtils.NestPaths(logger, archive, transformedPath.Value);
-            try
-            {
-                var contents = File.ReadAllText(fullInputPath, inputEncoding);
-                var tmpSrcFile = Path.GetTempFileName();
-                File.WriteAllText(tmpSrcFile, contents, utf8);
-
-                FileUtils.MoveOrReplace(tmpSrcFile, dest);
-            }
-            catch (Exception ex)
-            {
-                // If this happened, it was probably because
-                // - the same file was compiled multiple times, or
-                // - the file doesn't exist (due to wrong #line directive or because it's an in-memory source generated AST).
-                // In any case, this is not a fatal error.
-                logger.LogWarning("Problem archiving " + dest + ": " + ex);
-            }
-        }
-
         private static string TrapExtension(CompressionMode compression)
         {
             switch (compression)

csharp/ql/integration-tests/all-platforms/binlog/Files.expected

@@ -1,5 +1,5 @@
 | Program.cs:0:0:0:0 | Program.cs |
-| System.Text.RegularExpressions.Generator/System.Text.RegularExpressions.Generator.RegexGenerator/RegexGenerator.g.cs:0:0:0:0 | System.Text.RegularExpressions.Generator/System.Text.RegularExpressions.Generator.RegexGenerator/RegexGenerator.g.cs |
+| generated/test.csproj (net8.0)/System.Text.RegularExpressions.Generator/System.Text.RegularExpressions.Generator.RegexGenerator/RegexGenerator.g.cs:0:0:0:0 | generated/test.csproj (net8.0)/System.Text.RegularExpressions.Generator/System.Text.RegularExpressions.Generator.RegexGenerator/RegexGenerator.g.cs |
 | obj/Debug/net8.0/.NETCoreApp,Version=v8.0.AssemblyAttributes.cs:0:0:0:0 | obj/Debug/net8.0/.NETCoreApp,Version=v8.0.AssemblyAttributes.cs |
 | obj/Debug/net8.0/test.AssemblyInfo.cs:0:0:0:0 | obj/Debug/net8.0/test.AssemblyInfo.cs |
 | obj/Debug/net8.0/test.GlobalUsings.g.cs:0:0:0:0 | obj/Debug/net8.0/test.GlobalUsings.g.cs |