================
@@ -698,11 +698,32 @@ void LoopConvertCheck::doConversion(
std::string ReplaceText;
SourceRange Range = Usage.Range;
if (Usage.Expression) {
+ // Decide whether `ReplaceText` needs to be pre-appended with a space
+ // or not. In cases like `delete*it`, we can't just change it to
+ // `deleteit` we need to introduce space after `delete` to make it
+ // `delete it`.
+ Token StarToken;
+ if (Usage.Kind == Usage::UK_Default &&
+ !Lexer::getRawToken(Usage.Range.getBegin(), StarToken,
+ Context->getSourceManager(), getLangOpts(),
+ false) &&
+ StarToken.is(tok::star)) {
+ std::optional<Token> PrevToken = Lexer::findPreviousToken(
+ Usage.Range.getBegin(), Context->getSourceManager(),
+ getLangOpts(), true);
+ if (PrevToken) {
----------------
gxyd wrote:
I'm getting a warning from code linter:
```
clang-tools-extra/clang-tidy/modernize/LoopConvertCheck.cpp:566:38: warning:
unchecked access to optional value [bugprone-unchecked-optional-access]\n 566
| const bool IsAdjacentToPrevToken = PrevToken->getEndLoc() ==
BeginLocation;\n |
^~~~~~~~~\nclang-tools-extra/clang-tidy/modernize/LoopConvertCheck.cpp:567:10:
warning: unchecked access to optional value
[bugprone-unchecked-optional-access]\n 567 | return
PrevToken->isAnyIdentifier() && IsAdjacentToPrevToken;
```
should I revert back to use of `if (PrevToken)` to fix the linter warning or as
an AI suggested to me to use `if (PrevToken.has_value())`?
https://github.com/llvm/llvm-project/pull/202015
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
