[InstallAPI] Pick up input headers by directory traversal (llvm#94508)

Match TAPI behavior and allow input headers to be resolved via a passed
directory, which is expected to be a library sitting in a build
directory.

Author: cyndyishida

clang/include/clang/Basic/DiagnosticInstallAPIKinds.td

@@ -26,6 +26,8 @@ def err_unsupported_environment : Error<"environment '%0' is not supported: '%1'
 def err_unsupported_os : Error<"os '%0' is not supported: '%1'">;
 def err_cannot_read_input_list : Error<"could not read %0 input list '%1': %2">;
 def err_invalid_label: Error<"label '%0' is reserved: use a different label name for -X<label>">;
+def err_directory_scanning: Error<"could not read directory '%0': %1">;
+def err_more_than_one_library: Error<"more than one framework/dynamic library found">;
 } // end of command line category.
 
 let CategoryName = "Verification" in {

clang/include/clang/InstallAPI/DirectoryScanner.h

@@ -0,0 +1,81 @@
+//===- InstallAPI/DirectoryScanner.h ----------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// The DirectoryScanner for collecting library files on the file system.
+///
+//===----------------------------------------------------------------------===//
+#ifndef LLVM_CLANG_INSTALLAPI_DIRECTORYSCANNER_H
+#define LLVM_CLANG_INSTALLAPI_DIRECTORYSCANNER_H
+
+#include "clang/Basic/FileManager.h"
+#include "clang/InstallAPI/Library.h"
+
+namespace clang::installapi {
+
+enum ScanMode {
+  /// Scanning Framework directory.
+  ScanFrameworks,
+  /// Scanning Dylib directory.
+  ScanDylibs,
+};
+
+class DirectoryScanner {
+public:
+  DirectoryScanner(FileManager &FM, ScanMode Mode = ScanMode::ScanFrameworks)
+      : FM(FM), Mode(Mode) {}
+
+  /// Scan for all input files throughout directory.
+  ///
+  /// \param Directory Path of input directory.
+  llvm::Error scan(StringRef Directory);
+
+  /// Take over ownership of stored libraries.
+  std::vector<Library> takeLibraries() { return std::move(Libraries); };
+
+  /// Get all the header files in libraries.
+  ///
+  /// \param Libraries Reference of collection of libraries.
+  static HeaderSeq getHeaders(ArrayRef<Library> Libraries);
+
+private:
+  /// Collect files for dylibs in usr/(local)/lib within directory.
+  llvm::Error scanForUnwrappedLibraries(StringRef Directory);
+
+  /// Collect files for any frameworks within directory.
+  llvm::Error scanForFrameworks(StringRef Directory);
+
+  /// Get a library from the libraries collection.
+  Library &getOrCreateLibrary(StringRef Path, std::vector<Library> &Libs) const;
+
+  /// Collect multiple frameworks from directory.
+  llvm::Error scanMultipleFrameworks(StringRef Directory,
+                                     std::vector<Library> &Libs) const;
+  /// Collect files from nested frameworks.
+  llvm::Error scanSubFrameworksDirectory(StringRef Directory,
+                                         std::vector<Library> &Libs) const;
+
+  /// Collect files from framework path.
+  llvm::Error scanFrameworkDirectory(StringRef Path, Library &Framework) const;
+
+  /// Collect header files from path.
+  llvm::Error scanHeaders(StringRef Path, Library &Lib, HeaderType Type,
+                          StringRef BasePath,
+                          StringRef ParentPath = StringRef()) const;
+
+  /// Collect files from Version directories inside Framework directories.
+  llvm::Error scanFrameworkVersionsDirectory(StringRef Path,
+                                             Library &Lib) const;
+  FileManager &FM;
+  ScanMode Mode;
+  StringRef RootPath;
+  std::vector<Library> Libraries;
+};
+
+} // namespace clang::installapi
+
+#endif // LLVM_CLANG_INSTALLAPI_DIRECTORYSCANNER_H

clang/include/clang/InstallAPI/HeaderFile.h

@@ -97,6 +97,19 @@ class HeaderFile {
                                           Other.Excluded, Other.Extra,
                                           Other.Umbrella);
   }
+
+  bool operator<(const HeaderFile &Other) const {
+    /// For parsing of headers based on ordering,
+    /// group by type, then whether its an umbrella.
+    /// Capture 'extra' headers last.
+    /// This optimizes the chance of a sucessful parse for
+    /// headers that violate IWYU.
+    if (isExtra() && Other.isExtra())
+      return std::tie(Type, Umbrella) < std::tie(Other.Type, Other.Umbrella);
+
+    return std::tie(Type, Umbrella, Extra, FullPath) <
+           std::tie(Other.Type, Other.Umbrella, Other.Extra, Other.FullPath);
+  }
 };
 
 /// Glob that represents a pattern of header files to retreive.

clang/include/clang/InstallAPI/Library.h

@@ -0,0 +1,65 @@
+//===- InstallAPI/Library.h -------------------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// Defines the content of a library, such as public and private
+/// header files, and whether it is a framework.
+///
+//===----------------------------------------------------------------------===//
+#ifndef LLVM_CLANG_INSTALLAPI_LIBRARY_H
+#define LLVM_CLANG_INSTALLAPI_LIBRARY_H
+
+#include "clang/InstallAPI/HeaderFile.h"
+#include "clang/InstallAPI/MachO.h"
+
+namespace clang::installapi {
+
+class Library {
+public:
+  Library(StringRef Directory) : BaseDirectory(Directory) {}
+
+  /// Capture the name of the framework by the install name.
+  ///
+  /// \param InstallName The install name of the library encoded in a dynamic
+  /// library.
+  static StringRef getFrameworkNameFromInstallName(StringRef InstallName);
+
+  /// Get name of library by the discovered file path.
+  StringRef getName() const;
+
+  /// Get discovered path of library.
+  StringRef getPath() const { return BaseDirectory; }
+
+  /// Add a header file that belongs to the library.
+  ///
+  /// \param FullPath Path to header file.
+  /// \param Type Access level of header.
+  /// \param IncludePath The way the header should be included.
+  void addHeaderFile(StringRef FullPath, HeaderType Type,
+                     StringRef IncludePath = StringRef()) {
+    Headers.emplace_back(FullPath, Type, IncludePath);
+  }
+
+  /// Determine if library is empty.
+  bool empty() {
+    return SubFrameworks.empty() && Headers.empty() &&
+           FrameworkVersions.empty();
+  }
+
+private:
+  std::string BaseDirectory;
+  HeaderSeq Headers;
+  std::vector<Library> SubFrameworks;
+  std::vector<Library> FrameworkVersions;
+  bool IsUnwrappedDylib{false};
+
+  friend class DirectoryScanner;
+};
+
+} // namespace clang::installapi
+
+#endif // LLVM_CLANG_INSTALLAPI_LIBRARY_H

clang/include/clang/InstallAPI/MachO.h

@@ -31,6 +31,7 @@ using RecordLinkage = llvm::MachO::RecordLinkage;
 using Record = llvm::MachO::Record;
 using EncodeKind = llvm::MachO::EncodeKind;
 using GlobalRecord = llvm::MachO::GlobalRecord;
+using InterfaceFile = llvm::MachO::InterfaceFile;
 using ObjCContainerRecord = llvm::MachO::ObjCContainerRecord;
 using ObjCInterfaceRecord = llvm::MachO::ObjCInterfaceRecord;
 using ObjCCategoryRecord = llvm::MachO::ObjCCategoryRecord;

clang/lib/InstallAPI/CMakeLists.txt

@@ -8,10 +8,12 @@ set(LLVM_LINK_COMPONENTS
 
 add_clang_library(clangInstallAPI
   DiagnosticBuilderWrappers.cpp
+  DirectoryScanner.cpp
   DylibVerifier.cpp
   FileList.cpp
   Frontend.cpp
   HeaderFile.cpp
+  Library.cpp
   Visitor.cpp
 
   LINK_LIBS