Introduce canonicalization library for Windows subst drives

MODULE.bazel

@@ -280,7 +280,7 @@ go_sdk.download(version = "1.26.5")
 
 go_deps = use_extension("@gazelle//:extensions.bzl", "go_deps")
 go_deps.from_file(go_mod = "//go/extractor:go.mod")
-use_repo(go_deps, "com_github_stretchr_testify", "org_golang_x_mod", "org_golang_x_tools")
+use_repo(go_deps, "com_github_stretchr_testify", "org_golang_x_mod", "org_golang_x_sys", "org_golang_x_tools")
 
 ripunzip_archive = use_repo_rule("//misc/ripunzip:ripunzip.bzl", "ripunzip_archive")
 

csharp/ql/integration-tests/windows/subst/file.expected

@@ -1,8 +1,8 @@
-| code/Program.cs:0:0:0:0 | code/Program.cs |  |
-| code/obj/Debug/net9.0/.NETCoreApp,Version=v9.0.AssemblyAttributes.cs:0:0:0:0 | code/obj/Debug/net9.0/.NETCoreApp,Version=v9.0.AssemblyAttributes.cs |  |
-| code/obj/Debug/net9.0/dotnet_build.AssemblyInfo.cs:0:0:0:0 | code/obj/Debug/net9.0/dotnet_build.AssemblyInfo.cs |  |
-| code/obj/Debug/net9.0/dotnet_build.GlobalUsings.g.cs:0:0:0:0 | code/obj/Debug/net9.0/dotnet_build.GlobalUsings.g.cs |  |
-| code/obj/Debug/net9.0/dotnet_build.dll:0:0:0:0 | code/obj/Debug/net9.0/dotnet_build.dll |  |
+| code/Program.cs:0:0:0:0 | code/Program.cs | relative |
+| code/dotnet_build.csproj:0:0:0:0 | code/dotnet_build.csproj | relative |
+| code/obj/Debug/net9.0/.NETCoreApp,Version=v9.0.AssemblyAttributes.cs:0:0:0:0 | code/obj/Debug/net9.0/.NETCoreApp,Version=v9.0.AssemblyAttributes.cs | relative |
+| code/obj/Debug/net9.0/dotnet_build.AssemblyInfo.cs:0:0:0:0 | code/obj/Debug/net9.0/dotnet_build.AssemblyInfo.cs | relative |
+| code/obj/Debug/net9.0/dotnet_build.GlobalUsings.g.cs:0:0:0:0 | code/obj/Debug/net9.0/dotnet_build.GlobalUsings.g.cs | relative |
+| code/obj/Debug/net9.0/dotnet_build.dll:0:0:0:0 | code/obj/Debug/net9.0/dotnet_build.dll | relative |
+| code/obj/dotnet_build.csproj.nuget.g.props:0:0:0:0 | code/obj/dotnet_build.csproj.nuget.g.props | relative |
 | file://:0:0:0:0 |  |  |
-| file://Z:/dotnet_build.csproj:0:0:0:0 | Z:/dotnet_build.csproj | relative |
-| file://Z:/obj/dotnet_build.csproj.nuget.g.props:0:0:0:0 | Z:/obj/dotnet_build.csproj.nuget.g.props | relative |

go/extractor/extractor.go

@@ -770,7 +770,7 @@ func normalizedPath(ast *ast.File, fset *token.FileSet) string {
 	if err != nil {
 		return file
 	}
-	return path
+	return util.ResolvePath(path)
 }
 
 // extractFile extracts AST information for the given file

go/extractor/go.mod

@@ -13,7 +13,10 @@ require (
 	golang.org/x/tools v0.48.0
 )
 
-require github.com/stretchr/testify v1.11.1
+require (
+	github.com/stretchr/testify v1.11.1
+	golang.org/x/sys v0.47.0
+)
 
 require (
 	github.com/davecgh/go-spew v1.1.1 // indirect

go/extractor/go.sum

@@ -10,6 +10,8 @@ golang.org/x/mod v0.38.0 h1:MECBjubtXD7yj4HrhIUcywNaGeNVUdfVnxmPajOk4yk=
 golang.org/x/mod v0.38.0/go.mod h1:V6Xz0pq8TQ3dGqVQ1FVHuelZpAL0uNhSkk9ogYP3c40=
 golang.org/x/sync v0.22.0 h1:SZjpbeLmrCk4xhRSZFNZW5gFUeCeFgjekvI/+gfScek=
 golang.org/x/sync v0.22.0/go.mod h1:9xrNwdLfx4jkKbNva9FpL6vEN7evnE43NNNJQ2LF3+0=
+golang.org/x/sys v0.47.0 h1:o7XGOvZQCADBQQ4Y7VNq2dRWQR7JmOUW8Kxx4ZsNgWs=
+golang.org/x/sys v0.47.0/go.mod h1:4GL1E5IUh+htKOUEOaiffhrAeqysfVGipDYzABqnCmw=
 golang.org/x/tools v0.48.0 h1:3+hClM1aLL5mjMKm5ovokw9epgRXPuu2tILgismM6RE=
 golang.org/x/tools v0.48.0/go.mod h1:08xX0orndb/F7jJxGDicx061tyd5pcMto75YMAXr6lk=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405 h1:yhCVgyC4o1eVCa2tZl7eS0r+SDo693bJlVdllGtEeKM=

go/extractor/util/subst_other.go

@@ -0,0 +1,6 @@
+//go:build !windows
+
+package util
+
+// ResolvePath is a no-op on non-Windows platforms.
+func ResolvePath(path string) string { return path }

go/extractor/util/subst_windows.go

@@ -0,0 +1,91 @@
+//go:build windows
+
+package util
+
+import (
+	"os"
+	"path/filepath"
+	"syscall"
+	"unsafe"
+
+	"golang.org/x/sys/windows"
+)
+
+var (
+	dll         *syscall.DLL
+	procResolve *syscall.Proc
+	procFree    *syscall.Proc
+	available   bool
+)
+
+func init() {
+	dist := os.Getenv("CODEQL_DIST")
+	if dist == "" {
+		return
+	}
+	dllPath := filepath.Join(dist, "tools", "win64", "canonicalize.dll")
+	d, err := syscall.LoadDLL(dllPath)
+	if err != nil {
+		return
+	}
+	p, err := d.FindProc("resolve_subst")
+	if err != nil {
+		d.Release()
+		return
+	}
+	f, err := d.FindProc("resolve_subst_free")
+	if err != nil {
+		d.Release()
+		return
+	}
+	dll = d
+	procResolve = p
+	procFree = f
+	available = true
+}
+
+// If "path" is an absolute path starting with a "subst"ed drive letter, return an
+// equivalent path with the drive letter replaced by its target. Otherwise return
+// "path" unchanged.
+func ResolvePath(path string) string {
+	if len(path) < 3 {
+		return path
+	}
+	if path[1] != ':' {
+		return path
+	}
+	if path[2] != '\\' && path[2] != '/' {
+		return path
+	}
+	c := path[0]
+	if !((c >= 'A' && c <= 'Z') || (c >= 'a' && c <= 'z')) {
+		return path
+	}
+
+	resolved, ok := resolveDrive(path[:3])
+	if !ok {
+		return path
+	}
+	return resolved + path[2:]
+}
+
+// Given a drive root like "X:\" (or "X:/"), returns the path that drive is
+// "subst"ed to. Returns false if the drive is not "subst"ed or an error occurred.
+func resolveDrive(driveRoot string) (string, bool) {
+	if !available {
+		return "", false
+	}
+	driveBytes, err := windows.ByteSliceFromString(driveRoot)
+	if err != nil {
+		return "", false
+	}
+	ret, _, _ := procResolve.Call(uintptr(unsafe.Pointer(&driveBytes[0])))
+	if ret == 0 {
+		return "", false
+	}
+	result := windows.BytePtrToString((*byte)(unsafe.Pointer(ret)))
+	if procFree != nil {
+		procFree.Call(ret)
+	}
+	return result, true
+}

go/ql/integration-tests/subst/file.expected

@@ -1 +1 @@
-| file://Z:/main.go:0:0:0:0 | Z:/main.go | relative |
+| code/main.go:0:0:0:0 | code/main.go | relative |

java/kotlin-extractor/src/main/java/com/semmle/util/files/FileUtil.java

@@ -1237,13 +1237,14 @@ public static String relativePathLink (File f, File base)
 
 	/**
 	 * Try to convert a file into a canonical file. Handles the possible IO exception by just making
-	 * the path absolute.
+	 * the path absolute. Also resolves subst drives on Windows.
 	 */
 	public static File tryMakeCanonical (File f)
 	{
 		try {
-			return f.getCanonicalFile();
-		}
+			// getCanonicalFile does not canonicalize subst drives on Windows, so do this separately. This
+			// is a no-op on non-Windows platforms.
+			return SubstResolver.resolve(f.getCanonicalFile());		}
 		catch (IOException ignored) {
 			Exceptions.ignore(ignored, "Can't log error: Could be too verbose.");
 			return new File(simplifyPath(f));

java/kotlin-extractor/src/main/java/com/semmle/util/files/SubstResolver.java

@@ -0,0 +1,82 @@
+package com.semmle.util.files;
+
+import java.io.File;
+import java.net.URISyntaxException;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * Resolves Windows {@code subst}ed drive letters to their underlying paths. On non-Windows
+ * platforms, or when the native library failed to load, resolving will be a no-op.
+ */
+public class SubstResolver {
+  private static final boolean available;
+
+  static {
+    boolean loaded = false;
+    if (File.separatorChar == '\\') {
+      String dist = System.getenv("CODEQL_DIST");
+      if (dist != null && !dist.isEmpty()) {
+        try {
+          Path library = Paths.get(dist).resolve("tools")
+                  .resolve("win64").resolve("canonicalize.dll").toAbsolutePath();
+          System.load(library.toString());
+          loaded = true;
+        } catch (RuntimeException | UnsatisfiedLinkError ignored) {
+        }
+      }
+    }
+    available = loaded;
+  }
+
+  private SubstResolver() {}
+
+  /**
+   * Given a drive root like {@code "X:\\"} (or {@code "X:/"}), returns the path that drive is
+   * {@code subst}ed to, or {@code null} if the drive root was not subst drive or if an internal
+   * error occurred.
+   */
+  private static native String nativeResolveSubst(String driveRoot);
+
+  /**
+   * If {@code f} is an absolute path starting with a {@code subst}ed drive letter, return an
+   * equivalent path with the drive letter replaced by its target. Otherwise return {@code f}
+   * unchanged.
+   */
+  public static File resolve(File f) {
+    if (!available) {
+      return f;
+    }
+    String path = f.getPath();
+    if (path.length() < 3 || path.charAt(1) != ':') {
+      return f;
+    }
+    char sep = path.charAt(2);
+    if (sep != '\\' && sep != '/') {
+      return f;
+    }
+    if (!isDriveLetter(path.charAt(0))) {
+      return f;
+    }
+
+    String resolved;
+    try {
+      resolved = nativeResolveSubst(path.substring(0, 3));
+    } catch (UnsatisfiedLinkError ignored) {
+      return f;
+    }
+    if (resolved == null) {
+      return f;
+    }
+
+    // Append the remainder of the original path. The native side strips away
+    // any trailing separator.
+    return new File(resolved + path.substring(2));
+  }
+
+  private static boolean isDriveLetter(char c) {
+    return (c >= 'A' && c <= 'Z') || (c >= 'a' && c <= 'z');
+  }
+}

java/ql/integration-tests/java/subst/file.expected

@@ -1,5 +1,5 @@
+| code/Test.class:0:0:0:0 | Test | relative |
+| code/test1.java:0:0:0:0 | test1 | relative |
+| code/test2.kt:0:0:0:0 | test2 | relative |
 | file://:0:0:0:0 |  |  |
 | file://:0:0:0:0 |  |  |
-| file://Z:/Test.class:0:0:0:0 | Test | relative |
-| file://Z:/test1.java:0:0:0:0 | test1 | relative |
-| file://Z:/test2.kt:0:0:0:0 | test2 | relative |

javascript/ql/integration-tests/subst/file.expected

@@ -1,2 +1,2 @@
-| file://Z:/main.js:0:0:0:0 | Z:/main.js | relative |
-| file://Z:/test.ts:0:0:0:0 | Z:/test.ts | relative |
+| code/main.js:0:0:0:0 | code/main.js | relative |
+| code/test.ts:0:0:0:0 | code/test.ts | relative |

python/ql/integration-tests/subst/file.expected

@@ -1 +1 @@
-| code/main.py:0:0:0:0 | code/main.py |  |
+| code/main.py:0:0:0:0 | code/main.py | relative |

ruby/ql/integration-tests/subst/file.expected

@@ -1,2 +1,2 @@
-| code/test.rb:0:0:0:0 | code/test.rb |  |
+| code/test.rb:0:0:0:0 | code/test.rb | relative |
 | file://:0:0:0:0 |  |  |

rust/ql/integration-tests/subst/file.expected

@@ -1,2 +1,2 @@
-| code/test.rs:0:0:0:0 | code/test.rs |  |
+| code/test.rs:0:0:0:0 | code/test.rs | relative |
 | file://:0:0:0:0 |  |  |