Replaced manual mmap with llvm::MemoryBuffer

include/flang/Parser/provenance.h

@@ -150,8 +150,8 @@ class AllSources {
 
   void PushSearchPathDirectory(std::string);
   std::string PopSearchPathDirectory();
-  const SourceFile *Open(std::string path, std::stringstream *error);
-  const SourceFile *ReadStandardInput(std::stringstream *error);
+  const SourceFile *Open(std::string path, llvm::raw_ostream &error);
+  const SourceFile *ReadStandardInput(llvm::raw_ostream &error);
 
   ProvenanceRange AddIncludedFile(
       const SourceFile &, ProvenanceRange, bool isModule = false);

include/flang/Parser/source.h

@@ -20,6 +20,7 @@
 #include <string>
 #include <utility>
 #include <vector>
+#include "llvm/Support/MemoryBuffer.h"
 
 namespace Fortran::parser {
 
@@ -39,34 +40,30 @@ class SourceFile {
   explicit SourceFile(Encoding e) : encoding_{e} {}
   ~SourceFile();
   std::string path() const { return path_; }
-  const char *content() const { return content_; }
-  std::size_t bytes() const { return bytes_; }
+  llvm::ArrayRef<char> content() const { return buf_->getBuffer().slice(bom_end_, buf_end_ - bom_end_); }
+  std::size_t bytes() const { return content().size(); }
   std::size_t lines() const { return lineStart_.size(); }
   Encoding encoding() const { return encoding_; }
 
-  bool Open(std::string path, std::stringstream *error);
-  bool ReadStandardInput(std::stringstream *error);
+  bool Open(std::string path, llvm::raw_ostream &error);
+  bool ReadStandardInput(llvm::raw_ostream &error);
   void Close();
   SourcePosition FindOffsetLineAndColumn(std::size_t) const;
   std::size_t GetLineStartOffset(int lineNumber) const {
     return lineStart_.at(lineNumber - 1);
   }
 
 private:
-  bool ReadFile(std::string errorPath, std::stringstream *error);
+  void ReadFile();
   void IdentifyPayload();
   void RecordLineStarts();
 
   std::string path_;
-  int fileDescriptor_{-1};
-  bool isMemoryMapped_{false};
-  const char *address_{nullptr};  // raw content
-  std::size_t size_{0};
-  const char *content_{nullptr};  // usable content
-  std::size_t bytes_{0};
+  std::unique_ptr<llvm::WritableMemoryBuffer> buf_;
   std::vector<std::size_t> lineStart_;
-  std::string normalized_;
-  Encoding encoding_{Encoding::UTF_8};
+  std::size_t bom_end_ {0};
+  std::size_t buf_end_;
+  Encoding encoding_;
 };
 }
 #endif  // FORTRAN_PARSER_SOURCE_H_

lib/Parser/CMakeLists.txt

@@ -33,7 +33,7 @@ add_library(FortranParser
 )
 
 target_link_libraries(FortranParser
-  FortranCommon
+  FortranCommon LLVMSupport
 )
 
 install (TARGETS FortranParser

lib/Parser/parsing.cpp

@@ -29,12 +29,13 @@ const SourceFile *Parsing::Prescan(const std::string &path, Options options) {
     }
   }
 
-  std::stringstream fileError;
+  std::string fileError_buf;
+  llvm::raw_string_ostream fileError{fileError_buf};
   const SourceFile *sourceFile;
   if (path == "-") {
-    sourceFile = allSources.ReadStandardInput(&fileError);
+    sourceFile = allSources.ReadStandardInput(fileError);
   } else {
-    sourceFile = allSources.Open(path, &fileError);
+    sourceFile = allSources.Open(path, fileError);
   }
   if (!fileError.str().empty()) {
     ProvenanceRange range{allSources.AddCompilerInsertion(path)};

lib/Parser/preprocessor.cpp

@@ -574,8 +574,9 @@ void Preprocessor::Directive(const TokenSequence &dir, Prescanner *prescanner) {
           "#include: empty include file name"_err_en_US);
       return;
     }
-    std::stringstream error;
-    const SourceFile *included{allSources_.Open(include, &error)};
+    std::string error_buf;
+    llvm::raw_string_ostream error{error_buf};
+    const SourceFile *included{allSources_.Open(include, error)};
     if (!included) {
       prescanner->Say(dir.GetTokenProvenanceRange(dirOffset),
           "#include: %s"_err_en_US, error.str());

lib/Parser/prescan.cpp

@@ -63,7 +63,7 @@ void Prescanner::Prescan(ProvenanceRange range) {
   std::size_t offset{0};
   const SourceFile *source{allSources.GetSourceFile(startProvenance_, &offset)};
   CHECK(source);
-  start_ = source->content() + offset;
+  start_ = source->content().data() + offset;
   limit_ = start_ + range.size();
   nextLine_ = start_;
   const bool beganInFixedForm{inFixedForm_};
@@ -736,14 +736,15 @@ void Prescanner::FortranInclude(const char *firstQuote) {
     Say(GetProvenanceRange(garbage, p),
         "excess characters after path name"_en_US);
   }
-  std::stringstream error;
+  std::string error_buf;
+  llvm::raw_string_ostream error{error_buf};
   Provenance provenance{GetProvenance(nextLine_)};
   AllSources &allSources{cooked_.allSources()};
   const SourceFile *currentFile{allSources.GetSourceFile(provenance)};
   if (currentFile) {
     allSources.PushSearchPathDirectory(DirectoryName(currentFile->path()));
   }
-  const SourceFile *included{allSources.Open(path, &error)};
+  const SourceFile *included{allSources.Open(path, error)};
   if (currentFile) {
     allSources.PopSearchPathDirectory();
   }

lib/Parser/provenance.cpp

@@ -166,7 +166,7 @@ std::string AllSources::PopSearchPathDirectory() {
   return directory;
 }
 
-const SourceFile *AllSources::Open(std::string path, std::stringstream *error) {
+const SourceFile *AllSources::Open(std::string path, llvm::raw_ostream &error) {
   std::unique_ptr<SourceFile> source{std::make_unique<SourceFile>(encoding_)};
   if (source->Open(LocateSourceFile(path, searchPath_), error)) {
     return ownedSourceFiles_.emplace_back(std::move(source)).get();
@@ -175,7 +175,7 @@ const SourceFile *AllSources::Open(std::string path, std::stringstream *error) {
   }
 }
 
-const SourceFile *AllSources::ReadStandardInput(std::stringstream *error) {
+const SourceFile *AllSources::ReadStandardInput(llvm::raw_ostream &error) {
   std::unique_ptr<SourceFile> source{std::make_unique<SourceFile>(encoding_)};
   if (source->ReadStandardInput(error)) {
     return ownedSourceFiles_.emplace_back(std::move(source)).get();
@@ -227,7 +227,7 @@ void AllSources::EmitMessage(std::ostream &o,
             o << ':' << pos.line << ':' << pos.column;
             o << ": " << message << '\n';
             if (echoSourceLine) {
-              const char *text{inc.source.content() +
+              const char *text{inc.source.content().data() +
                   inc.source.GetLineStartOffset(pos.line)};
               o << "  ";
               for (const char *p{text}; *p != '\n'; ++p) {