[flang] More clean-up of CookedSource API

The std::string holding the content of a CookedSource no longer
needs to be exposed in its API after the recent work that allows
the parsing context to hold multiple instances of a CookedSource.
So clean the API.  These changes were extracted from some work in
progress that was made easier by the API changes.

Differential Revision: https://reviews.llvm.org/D87635

flang/include/flang/Parser/parse-state.h

@@ -36,7 +36,7 @@ class ParseState {
 public:
   // TODO: Add a constructor for parsing a normalized module file.
   ParseState(const CookedSource &cooked)
-      : p_{&cooked.data().front()}, limit_{&cooked.data().back() + 1} {}
+      : p_{cooked.AsCharBlock().begin()}, limit_{cooked.AsCharBlock().end()} {}
   ParseState(const ParseState &that)
       : p_{that.p_}, limit_{that.limit_}, context_{that.context_},
         userState_{that.userState_}, inFixedForm_{that.inFixedForm_},

flang/include/flang/Parser/provenance.h

@@ -167,6 +167,7 @@ public:
       const std::string &message, bool echoSourceLine = false) const;
   const SourceFile *GetSourceFile(
       Provenance, std::size_t *offset = nullptr) const;
+  const char *GetSource(ProvenanceRange) const;
   std::optional<SourcePosition> GetSourcePosition(Provenance) const;
   std::optional<ProvenanceRange> GetFirstFileProvenance() const;
   std::string GetPath(Provenance) const; // __FILE__
@@ -219,16 +220,7 @@ private:
 // single instances of CookedSource.
 class CookedSource {
 public:
-  const std::string &data() const { return data_; }
-
-  bool Contains(const char *p) const {
-    return p >= &data_.front() && p <= &data_.back() + 1;
-  }
-  bool Contains(CharBlock range) const {
-    return !range.empty() && Contains(range.begin()) &&
-        Contains(range.end() - 1);
-  }
-
+  CharBlock AsCharBlock() const { return CharBlock{data_}; }
   std::optional<ProvenanceRange> GetProvenanceRange(CharBlock) const;
   std::optional<CharBlock> GetCharBlock(ProvenanceRange) const;
 
@@ -253,7 +245,6 @@ public:
   std::size_t BufferedBytes() const;
   void Marshal(AllSources &); // marshals text into one contiguous block
   void CompileProvenanceRangeToOffsetMappings(AllSources &);
-  std::string AcquireData() { return std::move(data_); }
   llvm::raw_ostream &Dump(llvm::raw_ostream &) const;
 
 private:
@@ -276,7 +267,7 @@ public:
   template <typename A> // const char * or CharBlock
   const CookedSource *Find(A x) const {
     for (const auto &c : cooked_) {
-      if (c.Contains(x)) {
+      if (c.AsCharBlock().Contains(x)) {
         return &c;
       }
     }

flang/include/flang/Semantics/semantics.h

@@ -204,10 +204,10 @@ private:
 class Semantics {
 public:
   explicit Semantics(SemanticsContext &context, parser::Program &program,
-      const parser::CookedSource &cooked, bool debugModuleWriter = false)
+      parser::CharBlock charBlock, bool debugModuleWriter = false)
       : context_{context}, program_{program} {
     context.set_debugModuleWriter(debugModuleWriter);
-    context.globalScope().AddSourceRange(parser::CharBlock{cooked.data()});
+    context.globalScope().AddSourceRange(charBlock);
   }
 
   SemanticsContext &context() const { return context_; }

flang/lib/Lower/OpenACC.cpp

@@ -1,4 +1,4 @@
-//===-- OpenMP.cpp -- OpenACC directive lowering --------------------------===//
+//===-- OpenACC.cpp -- OpenACC directive lowering -------------------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.

flang/lib/Parser/prescan.cpp

@@ -62,11 +62,8 @@ static void NormalizeCompilerDirectiveCommentMarker(TokenSequence &dir) {
 
 void Prescanner::Prescan(ProvenanceRange range) {
   startProvenance_ = range.start();
-  std::size_t offset{0};
-  const SourceFile *source{
-      allSources_.GetSourceFile(startProvenance_, &offset)};
-  CHECK(source);
-  start_ = source->content().data() + offset;
+  start_ = allSources_.GetSource(range);
+  CHECK(start_);
   limit_ = start_ + range.size();
   nextLine_ = start_;
   const bool beganInFixedForm{inFixedForm_};
@@ -75,7 +72,7 @@ void Prescanner::Prescan(ProvenanceRange range) {
         "too many nested INCLUDE/#include files, possibly circular"_err_en_US);
     return;
   }
-  while (nextLine_ < limit_) {
+  while (!IsAtEnd()) {
     Statement();
   }
   if (inFixedForm_ != beganInFixedForm) {
@@ -232,7 +229,7 @@ void Prescanner::Statement() {
 }
 
 TokenSequence Prescanner::TokenizePreprocessorDirective() {
-  CHECK(nextLine_ < limit_ && !inPreprocessorDirective_);
+  CHECK(!IsAtEnd() && !inPreprocessorDirective_);
   inPreprocessorDirective_ = true;
   BeginStatementAndAdvance();
   TokenSequence tokens;
@@ -360,7 +357,7 @@ void Prescanner::SkipCComments() {
         break;
       }
     } else if (inPreprocessorDirective_ && at_[0] == '\\' && at_ + 2 < limit_ &&
-        at_[1] == '\n' && nextLine_ < limit_) {
+        at_[1] == '\n' && !IsAtEnd()) {
       BeginSourceLineAndAdvance();
     } else {
       break;
@@ -804,7 +801,7 @@ bool Prescanner::IsNextLinePreprocessorDirective() const {
 }
 
 bool Prescanner::SkipCommentLine(bool afterAmpersand) {
-  if (nextLine_ >= limit_) {
+  if (IsAtEnd()) {
     if (afterAmpersand && prescannerNesting_ > 0) {
       // A continuation marker at the end of the last line in an
       // include file inhibits the newline for that line.
@@ -843,7 +840,7 @@ bool Prescanner::SkipCommentLine(bool afterAmpersand) {
 }
 
 const char *Prescanner::FixedFormContinuationLine(bool mightNeedSpace) {
-  if (nextLine_ >= limit_) {
+  if (IsAtEnd()) {
     return nullptr;
   }
   tabInCurrentLine_ = false;
@@ -995,7 +992,7 @@ bool Prescanner::FreeFormContinuation() {
 // arguments to span multiple lines.
 bool Prescanner::IsImplicitContinuation() const {
   return !inPreprocessorDirective_ && !inCharLiteral_ &&
-      delimiterNesting_ > 0 && nextLine_ < limit_ &&
+      delimiterNesting_ > 0 && !IsAtEnd() &&
       ClassifyLine(nextLine_).kind == LineClassification::Kind::Source;
 }
 

flang/lib/Parser/provenance.cpp

@@ -301,6 +301,14 @@ const SourceFile *AllSources::GetSourceFile(
       origin.u);
 }
 
+const char *AllSources::GetSource(ProvenanceRange range) const {
+  Provenance start{range.start()};
+  const Origin &origin{MapToOrigin(start)};
+  return origin.covers.Contains(range)
+      ? &origin[origin.covers.MemberOffset(start)]
+      : nullptr;
+}
+
 std::optional<SourcePosition> AllSources::GetSourcePosition(
     Provenance prov) const {
   const Origin &origin{MapToOrigin(prov)};
@@ -402,7 +410,7 @@ const AllSources::Origin &AllSources::MapToOrigin(Provenance at) const {
 
 std::optional<ProvenanceRange> CookedSource::GetProvenanceRange(
     CharBlock cookedRange) const {
-  if (!Contains(cookedRange)) {
+  if (!AsCharBlock().Contains(cookedRange)) {
     return std::nullopt;
   }
   ProvenanceRange first{provenanceMap_.Map(cookedRange.begin() - &data_[0])};

flang/tools/f18/f18.cpp

@@ -251,7 +251,7 @@ std::string CompileFortran(std::string path, Fortran::parser::Options options,
       driver.dumpSymbols || driver.dumpUnparseWithSymbols ||
       driver.getDefinition || driver.getSymbolsSources) {
     Fortran::semantics::Semantics semantics{semanticsContext, parseTree,
-        parsing.cooked(), driver.debugModuleWriter};
+        parsing.cooked().AsCharBlock(), driver.debugModuleWriter};
     semantics.Perform();
     semantics.EmitMessages(llvm::errs());
     if (driver.dumpSymbols) {

flang/unittests/Evaluate/intrinsics.cpp

@@ -26,10 +26,10 @@ public:
   }
   void Marshal() { cooked_.Marshal(allSources_); }
   parser::CharBlock operator()(const std::string &s) {
-    return {cooked_.data().data() + offsets_[s], s.size()};
+    return {cooked_.AsCharBlock().begin() + offsets_[s], s.size()};
   }
   parser::ContextualMessages Messages(parser::Messages &buffer) {
-    return parser::ContextualMessages{cooked_.data(), &buffer};
+    return parser::ContextualMessages{cooked_.AsCharBlock(), &buffer};
   }
   void Emit(llvm::raw_ostream &o, const parser::Messages &messages) {
     messages.Emit(o, allCookedSources_);