[clang] [clang][Dependency Scanning] Refactor Scanning Compiler Instance Initialization (PR #161300)

[Jan Svoboda via cfe-commits]

================
@@ -356,43 +369,131 @@ void clang::tooling::dependencies::sanitizeDiagOpts(
   });
 }
 
-bool DependencyScanningAction::runInvocation(
-    std::shared_ptr<CompilerInvocation> Invocation,
-    IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS,
-    std::shared_ptr<PCHContainerOperations> PCHContainerOps,
-    DiagnosticConsumer *DiagConsumer) {
-  // Making sure that we canonicalize the defines before we create the deep
-  // copy to avoid unnecessary variants in the scanner and in the resulting
-  // explicit command lines.
-  if (any(Service.getOptimizeArgs() & ScanningOptimizations::Macros))
-    canonicalizeDefines(Invocation->getPreprocessorOpts());
+std::pair<std::unique_ptr<driver::Driver>, std::unique_ptr<driver::Compilation>>
+clang::tooling::dependencies::buildCompilation(
+    ArrayRef<std::string> ArgStrs, DiagnosticsEngine &Diags,
+    IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS) {
+  SmallVector<const char *, 256> Argv;
+  Argv.reserve(ArgStrs.size());
+  for (const std::string &Arg : ArgStrs)
+    Argv.push_back(Arg.c_str());
+
+  std::unique_ptr<driver::Driver> Driver = std::make_unique<driver::Driver>(
+      Argv[0], llvm::sys::getDefaultTargetTriple(), Diags,
+      "clang LLVM compiler", FS);
+  Driver->setTitle("clang_based_tool");
+
+  llvm::BumpPtrAllocator Alloc;
+  bool CLMode = driver::IsClangCL(
+      driver::getDriverMode(Argv[0], ArrayRef(Argv).slice(1)));
+
+  if (llvm::Error E =
+          driver::expandResponseFiles(Argv, CLMode, Alloc, FS.get())) {
+    Diags.Report(diag::err_drv_expand_response_file)
+        << llvm::toString(std::move(E));
+    return std::make_pair(nullptr, nullptr);
+  }
 
-  // Make a deep copy of the original Clang invocation.
-  CompilerInvocation OriginalInvocation(*Invocation);
+  std::unique_ptr<driver::Compilation> Compilation(
+      Driver->BuildCompilation(llvm::ArrayRef(Argv)));
+  if (!Compilation)
+    return std::make_pair(nullptr, nullptr);
 
-  if (Scanned) {
-    // Scanning runs once for the first -cc1 invocation in a chain of driver
-    // jobs. For any dependent jobs, reuse the scanning result and just
-    // update the LastCC1Arguments to correspond to the new invocation.
-    // FIXME: to support multi-arch builds, each arch requires a separate scan
-    setLastCC1Arguments(std::move(OriginalInvocation));
-    return true;
+  if (Compilation->containsError())
+    return std::make_pair(nullptr, nullptr);
+
+  return std::make_pair(std::move(Driver), std::move(Compilation));
+}
+
+llvm::IntrusiveRefCntPtr<llvm::vfs::FileSystem>
+clang::tooling::dependencies::initVFSForTUBufferScanning(
----------------
jansvoboda11 wrote:

This is essentially returning one part of the result as a return value and the other as an out-parameter. WDYT about making the return type `std::optional<std::pair<IntrusiveRefCntPtr<llvm::vfs::FileSystem>, std::vector<std::string>>` or taking the VFS as an out-parameter as well?

https://github.com/llvm/llvm-project/pull/161300