[flang-commits] [flang] [flang] Accept a compiler directive sentinel after a semicolon (PR #96966)

[Peter Klausler via flang-commits]

https://github.com/klausler updated https://github.com/llvm/llvm-project/pull/96966

>From 111d269f3b011f91f70e08014ddda5e46110f820 Mon Sep 17 00:00:00 2001
From: Peter Klausler <pklausler at nvidia.com>
Date: Thu, 27 Jun 2024 13:52:40 -0700
Subject: [PATCH] [flang] Accept a compiler directive sentinel after a
 semicolon

Don't treat !DIR$ or an active !$ACC, !$OMP, &c. as a comment when they
appear after a semicolon, but instead treat them as a compiler directive
sentinel.
---
 flang/lib/Parser/prescan.cpp                  | 54 ++++++++++---------
 flang/lib/Parser/prescan.h                    |  3 ++
 flang/lib/Parser/token-sequence.cpp           | 11 ++--
 .../Preprocessing/sentinel-after-semi.F90     |  7 +++
 4 files changed, 42 insertions(+), 33 deletions(-)
 create mode 100644 flang/test/Preprocessing/sentinel-after-semi.F90

diff --git a/flang/lib/Parser/prescan.cpp b/flang/lib/Parser/prescan.cpp
index 42aa829e0ed5b..d44bdde70f380 100644
--- a/flang/lib/Parser/prescan.cpp
+++ b/flang/lib/Parser/prescan.cpp
@@ -471,7 +471,7 @@ bool Prescanner::MustSkipToEndOfLine() const {
   if (inFixedForm_ && column_ > fixedFormColumnLimit_ && !tabInCurrentLine_) {
     return true; // skip over ignored columns in right margin (73:80)
   } else if (*at_ == '!' && !inCharLiteral_) {
-    return true; // inline comment goes to end of source line
+    return !IsCompilerDirectiveSentinel(at_);
   } else {
     return false;
   }
@@ -1345,32 +1345,12 @@ Prescanner::IsFixedFormCompilerDirectiveLine(const char *start) const {
 
 std::optional<Prescanner::LineClassification>
 Prescanner::IsFreeFormCompilerDirectiveLine(const char *start) const {
-  char sentinel[8];
-  const char *p{SkipWhiteSpace(start)};
-  if (*p++ != '!') {
-    return std::nullopt;
-  }
-  for (std::size_t j{0}; j + 1 < sizeof sentinel; ++p, ++j) {
-    if (*p == '\n') {
-      break;
-    }
-    if (*p == ' ' || *p == '\t' || *p == '&') {
-      if (j == 0) {
-        break;
-      }
-      sentinel[j] = '\0';
-      p = SkipWhiteSpace(p + 1);
-      if (*p == '!') {
-        break;
-      }
-      if (const char *sp{IsCompilerDirectiveSentinel(sentinel, j)}) {
-        std::size_t offset = p - start;
-        return {LineClassification{
-            LineClassification::Kind::CompilerDirective, offset, sp}};
-      }
-      break;
+  if (const char *p{SkipWhiteSpace(start)}; p && *p++ == '!') {
+    if (auto maybePair{IsCompilerDirectiveSentinel(p)}) {
+      auto offset{static_cast<std::size_t>(maybePair->second - start)};
+      return {LineClassification{LineClassification::Kind::CompilerDirective,
+          offset, maybePair->first}};
     }
-    sentinel[j] = ToLowerCaseLetter(*p);
   }
   return std::nullopt;
 }
@@ -1415,6 +1395,28 @@ const char *Prescanner::IsCompilerDirectiveSentinel(CharBlock token) const {
   return end > p && IsCompilerDirectiveSentinel(p, end - p) ? p : nullptr;
 }
 
+std::optional<std::pair<const char *, const char *>>
+Prescanner::IsCompilerDirectiveSentinel(const char *p) const {
+  char sentinel[8];
+  for (std::size_t j{0}; j + 1 < sizeof sentinel && *p != '\n'; ++p, ++j) {
+    if (*p == ' ' || *p == '\t' || *p == '&') {
+      if (j > 0) {
+        sentinel[j] = '\0';
+        p = SkipWhiteSpace(p + 1);
+        if (*p != '!') {
+          if (const char *sp{IsCompilerDirectiveSentinel(sentinel, j)}) {
+            return std::make_pair(sp, p);
+          }
+        }
+      }
+      break;
+    } else {
+      sentinel[j] = ToLowerCaseLetter(*p);
+    }
+  }
+  return std::nullopt;
+}
+
 constexpr bool IsDirective(const char *match, const char *dir) {
   for (; *match; ++match) {
     if (*match != ToLowerCaseLetter(*dir++)) {
diff --git a/flang/lib/Parser/prescan.h b/flang/lib/Parser/prescan.h
index 421ba97d324c9..a64df5377e7e0 100644
--- a/flang/lib/Parser/prescan.h
+++ b/flang/lib/Parser/prescan.h
@@ -74,6 +74,9 @@ class Prescanner {
 
   const char *IsCompilerDirectiveSentinel(const char *, std::size_t) const;
   const char *IsCompilerDirectiveSentinel(CharBlock) const;
+  // 'first' is the sentinel, 'second' is beginning of payload
+  std::optional<std::pair<const char *, const char *>>
+  IsCompilerDirectiveSentinel(const char *p) const;
 
   template <typename... A> Message &Say(A &&...a) {
     return messages_.Say(std::forward<A>(a)...);
diff --git a/flang/lib/Parser/token-sequence.cpp b/flang/lib/Parser/token-sequence.cpp
index 133e60ba4f009..c73bca1df33de 100644
--- a/flang/lib/Parser/token-sequence.cpp
+++ b/flang/lib/Parser/token-sequence.cpp
@@ -266,7 +266,7 @@ TokenSequence &TokenSequence::ClipComment(
     if (std::size_t blanks{tok.CountLeadingBlanks()};
         blanks < tok.size() && tok[blanks] == '!') {
       // Retain active compiler directive sentinels (e.g. "!dir$")
-      for (std::size_t k{j + 1}; k < tokens && tok.size() < blanks + 5; ++k) {
+      for (std::size_t k{j + 1}; k < tokens && tok.size() <= blanks + 5; ++k) {
         if (tok.begin() + tok.size() == TokenAt(k).begin()) {
           tok.ExtendToCover(TokenAt(k));
         } else {
@@ -274,12 +274,9 @@ TokenSequence &TokenSequence::ClipComment(
         }
       }
       bool isSentinel{false};
-      if (tok.size() == blanks + 5) {
-        char sentinel[4];
-        for (int k{0}; k < 4; ++k) {
-          sentinel[k] = ToLowerCaseLetter(tok[blanks + k + 1]);
-        }
-        isSentinel = prescanner.IsCompilerDirectiveSentinel(sentinel, 4);
+      if (tok.size() > blanks + 5) {
+        isSentinel = prescanner.IsCompilerDirectiveSentinel(&tok[blanks + 1])
+                         .has_value();
       }
       if (isSentinel) {
       } else if (skipFirst) {
diff --git a/flang/test/Preprocessing/sentinel-after-semi.F90 b/flang/test/Preprocessing/sentinel-after-semi.F90
new file mode 100644
index 0000000000000..75060ac1db032
--- /dev/null
+++ b/flang/test/Preprocessing/sentinel-after-semi.F90
@@ -0,0 +1,7 @@
+! RUN: %flang_fc1 -fdebug-unparse -fopenacc %s 2>&1 | FileCheck %s
+! CHECK: !$ACC DECLARE COPYIN(var)
+#define ACCDECLARE(var) integer :: var; \
+  !$acc declare copyin(var)
+program main
+  ACCDECLARE(var)
+end program