[BOLT] Allow merge-fdata to take a directory as input

[kongyi]

Summary

... and recursively merge all files under said directory. This is similar to llvm-profdata merge behaviour.

[rafauler]

LGTM, thanks!

bolt/tools/merge-fdata/merge-fdata.cpp

 //
 //   $ merge-fdata 1.fdata 2.fdata 3.fdata > merged.fdata
 //
 //===----------------------------------------------------------------------===//
 
 #include "bolt/Profile/ProfileYAMLMapping.h"
 #include "llvm/ADT/StringMap.h"
 #include "llvm/Support/CommandLine.h"
+#include "llvm/Support/FileSystem.h"
 #include "llvm/Support/ManagedStatic.h"
 #include "llvm/Support/PrettyStackTrace.h"
 #include "llvm/Support/Signals.h"
 #include <unordered_map>
 
 using namespace llvm;
 using namespace llvm::yaml::bolt;
 
bool isYAML(const StringRef Filename) {
   if (std::error_code EC = MB.getError())
     report_error(Filename, EC);
   StringRef Buffer = MB.get()->getBuffer();
   if (Buffer.startswith("---\n"))
     return true;
   return false;
 }
 
-void mergeLegacyProfiles(const cl::list<std::string> &Filenames) {
+void mergeLegacyProfiles(const SmallVectorImpl<std::string> &Filenames) {
   errs() << "Using legacy profile format.\n";
   bool BoltedCollection = false;
   bool First = true;
   StringMap<uint64_t> Entries;
   for (const std::string &Filename : Filenames) {
     if (isYAML(Filename))
       report_error(Filename, "cannot mix YAML and legacy formats");
     ErrorOr<std::unique_ptr<MemoryBuffer>> MB =
int main(int argc, char **argv) {
 
   cl::HideUnrelatedOptions(opts::MergeFdataCategory);
 
   cl::ParseCommandLineOptions(argc, argv,
                               "merge multiple fdata into a single file");
 
   ToolName = argv[0];
 
-  if (!isYAML(opts::InputDataFilenames.front())) {
-    mergeLegacyProfiles(opts::InputDataFilenames);
+  // Recursively expand input directories into input file lists.
+  SmallVector<std::string> Inputs;
+  for (std::string &InputDataFilename : opts::InputDataFilenames) {
+    if (!llvm::sys::fs::exists(InputDataFilename))
+      report_error(InputDataFilename,
+                   std::make_error_code(std::errc::no_such_file_or_directory));
+    if (llvm::sys::fs::is_regular_file(InputDataFilename))
+      Inputs.emplace_back(InputDataFilename);
+    else if (llvm::sys::fs::is_directory(InputDataFilename)) {
+      std::error_code EC;
+      for (llvm::sys::fs::recursive_directory_iterator F(InputDataFilename, EC),
+           E;
+           F != E && !EC; F.increment(EC))
+        if (llvm::sys::fs::is_regular_file(F->path()))
+          Inputs.emplace_back(F->path());
+      if (EC)
+        report_error(InputDataFilename, EC);
+    }
+  }
+
+  if (!isYAML(Inputs.front())) {
+    mergeLegacyProfiles(Inputs);
     return 0;
   }
 
   // Merged header.
   BinaryProfileHeader MergedHeader;
   MergedHeader.Version = 1;
 
   // Merged information for all functions.
   StringMap<BinaryFunctionProfile> MergedBFs;
 
-  for (std::string &InputDataFilename : opts::InputDataFilenames) {
+  for (std::string &InputDataFilename : Inputs) {
     ErrorOr<std::unique_ptr<MemoryBuffer>> MB =
         MemoryBuffer::getFileOrSTDIN(InputDataFilename);
     if (std::error_code EC = MB.getError())
       report_error(InputDataFilename, EC);
     yaml::Input YamlInput(MB.get()->getBuffer());
 
     errs() << "Merging data from " << InputDataFilename << "...\n";