[PATCH] D147357: [clang-tidy] Add bugprone-optional-value-conversion check

[Nathan James via Phabricator via cfe-commits]

njames93 added a comment.

In D147357#4256636 <https://reviews.llvm.org/D147357#4256636>, @PiotrZSL wrote:

> As for `takeOptionalValue(std::move(*param));`, it could be added as an configuration option for PassthroughFunctions or UtilityFunctions. I can thing about that, that should be easy to implement.

That's overcomplicating things, just look for `std::move` and you'll cover 99.9% of instances in real world code.



================
Comment at: clang-tools-extra/docs/clang-tidy/checks/bugprone/optional-value-conversion.rst:34
+Value extraction using ``operator *`` is matched by default.
+Check does not provide auto-fixes.
+
----------------
PiotrZSL wrote:
> njames93 wrote:
> > Any reason for this limitation, given in most cases the fix is to just remove the `*` or `.value()` call
> Sometimes problem is not with optional but with called function that take optional when it could take value.
> At least for such issues I run, still probably could be safe to introduce such auto-fix to just remove */.value().
> But then question would be if it should be bugprone or readability check.
> For me like 10% of these issues show some possible bug around that code.
That's a good point, if they are hiding a bug we don't want to auto fix it.
Perhaps the fix should be added as a note
```
note: Remove the call to '%FuncName%' to silence this warning
note: Remove '*' to silence this warning
```


================
Comment at: clang-tools-extra/test/clang-tidy/checkers/bugprone/optional-value-conversion.cpp:16
+  };
+}
+
----------------
PiotrZSL wrote:
> njames93 wrote:
> > It'd be good to have tests demonstating `boost::optional` and `absl::optional` as well as the (non standard) get accessor
> I added them to config just to add them, as they should work, I could add some tests for them, but behavior should be same...
On the subject of boost(unsure about absl as I've never used that library)
Would you ever add support for `T& boost::get(boost::optional<T> &)` or its const ref cousin?


================
Comment at: clang-tools-extra/test/clang-tidy/checkers/bugprone/optional-value-conversion.cpp:37
+  // CHECK-MESSAGES: :[[@LINE-1]]:26: warning: conversion from 'std::optional<int>' into 'int' and back into 'std::optional<int>', remove potentially error-prone optional dereference [bugprone-optional-value-conversion]
+}
+
----------------
What happens with this pathological case
`takeOptionalValue(param.operator*());`
Looks like it isn't being handled, should probably handle(and test) for it


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D147357/new/

https://reviews.llvm.org/D147357