[flang-commits] [flang] [flang] Handle continuation line edge case (PR #74751)

[Peter Klausler via flang-commits]

https://github.com/klausler updated https://github.com/llvm/llvm-project/pull/74751

>From c71eb11affb05231559ee2c02f1fff656ea6507e Mon Sep 17 00:00:00 2001
From: Peter Klausler <pklausler at nvidia.com>
Date: Thu, 7 Dec 2023 11:14:18 -0800
Subject: [PATCH] [flang] Handle continuation line edge case

For a character literal that is split over more than one
source line with free form line continuation using '&'
at the end of one line but missing the standard-required
'&' on the continuation line, also handle the case of spaces
at the beginning of the continuation line.

For example,

PRINT *, 'don'&
 't poke the bear'

now prints "don't poke the bear", like nearly all other Fortran
compilers do.

This is not strictly standard conforming behavior, and the compiler
emits a portability warning with -pedantic.

Fixes llvm-test-suite/Fortran/gfortran/regression/continuation_1.f90,
.../continuation_12.f90, and .../continuation_13.f90.
---
 flang/lib/Parser/prescan.cpp                  | 20 ++++++++++---------
 flang/lib/Parser/prescan.h                    |  1 +
 .../test/Parser/continuation-before-quote.f90 | 10 ----------
 .../Parser/continuation-without-ampersand.f90 | 13 ++++++++++++
 4 files changed, 25 insertions(+), 19 deletions(-)
 delete mode 100644 flang/test/Parser/continuation-before-quote.f90
 create mode 100644 flang/test/Parser/continuation-without-ampersand.f90

diff --git a/flang/lib/Parser/prescan.cpp b/flang/lib/Parser/prescan.cpp
index 449ea60144424..79cdaccf1fbfe 100644
--- a/flang/lib/Parser/prescan.cpp
+++ b/flang/lib/Parser/prescan.cpp
@@ -706,6 +706,7 @@ void Prescanner::QuotedCharacterLiteral(
   char quote{*at_};
   const char *end{at_ + 1};
   inCharLiteral_ = true;
+  continuationInCharLiteral_ = true;
   const auto emit{[&](char ch) { EmitChar(tokens, ch); }};
   const auto insert{[&](char ch) { EmitInsertedChar(tokens, ch); }};
   bool isEscaped{false};
@@ -749,16 +750,9 @@ void Prescanner::QuotedCharacterLiteral(
         break;
       }
       inCharLiteral_ = true;
-      if (insertASpace_) {
-        if (features_.ShouldWarn(
-                common::LanguageFeature::MiscSourceExtensions)) {
-          Say(GetProvenanceRange(at_, end),
-              "Repeated quote mark in character literal continuation line should have been preceded by '&'"_port_en_US);
-        }
-        insertASpace_ = false;
-      }
     }
   }
+  continuationInCharLiteral_ = false;
   inCharLiteral_ = false;
 }
 
@@ -1122,7 +1116,15 @@ const char *Prescanner::FreeFormContinuationLine(bool ampersand) {
     } else if (*p == '!' || *p == '\n' || *p == '#') {
       return nullptr;
     } else if (ampersand || IsImplicitContinuation()) {
-      if (p > nextLine_) {
+      if (continuationInCharLiteral_) {
+        // 'a'&            -> 'a''b' == "a'b"
+        //   'b'
+        if (features_.ShouldWarn(
+                common::LanguageFeature::MiscSourceExtensions)) {
+          Say(GetProvenanceRange(p, p + 1),
+              "Character literal continuation line should have been preceded by '&'"_port_en_US);
+        }
+      } else if (p > nextLine_) {
         --p;
       } else {
         insertASpace_ = true;
diff --git a/flang/lib/Parser/prescan.h b/flang/lib/Parser/prescan.h
index 16b2c6165f611..84e046c1b102f 100644
--- a/flang/lib/Parser/prescan.h
+++ b/flang/lib/Parser/prescan.h
@@ -218,6 +218,7 @@ class Prescanner {
   bool slashInCurrentStatement_{false};
   bool preventHollerith_{false}; // CHARACTER*4HIMOM not Hollerith
   bool inCharLiteral_{false};
+  bool continuationInCharLiteral_{false};
   bool inPreprocessorDirective_{false};
 
   // In some edge cases of compiler directive continuation lines, it
diff --git a/flang/test/Parser/continuation-before-quote.f90 b/flang/test/Parser/continuation-before-quote.f90
deleted file mode 100644
index 66252010d89c4..0000000000000
--- a/flang/test/Parser/continuation-before-quote.f90
+++ /dev/null
@@ -1,10 +0,0 @@
-! RUN: %flang_fc1 -fsyntax-only -pedantic %s 2>&1 | FileCheck %s
-! Continuation between repeated quotation marks
-subroutine test
-!CHECK: portability: Repeated quote mark in character literal continuation line should have been preceded by '&'
-  print *, 'needs an '&
-'ampersand'''
-!CHECK-NOT: portability: Repeated quote mark in character literal continuation line should have been preceded by '&'
-  print *, 'has an '&
-&'ampersand'''
-end
diff --git a/flang/test/Parser/continuation-without-ampersand.f90 b/flang/test/Parser/continuation-without-ampersand.f90
new file mode 100644
index 0000000000000..5c3f23235edcc
--- /dev/null
+++ b/flang/test/Parser/continuation-without-ampersand.f90
@@ -0,0 +1,13 @@
+! RUN: %flang_fc1 -fsyntax-only -pedantic %s 2>&1 | FileCheck %s
+! Continuation between repeated quotation marks
+subroutine test
+!CHECK: portability: Character literal continuation line should have been preceded by '&'
+  print *, 'needs an '&
+'ampersand'''
+!CHECK: portability: Character literal continuation line should have been preceded by '&'
+  print *, 'also needs an '&
+ 'ampersand'''
+!CHECK-NOT: portability: Character literal continuation line should have been preceded by '&'
+  print *, 'has an '&
+&'ampersand'''
+end