[llvm-branch-commits] [mlir] [mlir][Parser] Deduplicate floating-point parsing functionality (PR #116172)

[River Riddle via llvm-branch-commits]

================
@@ -99,6 +99,29 @@ FailureOr<APFloat> detail::parseFloatFromIntegerLiteral(
   return APFloat(semantics, truncatedValue);
 }
 
+FailureOr<APFloat>
+detail::parseFloatFromLiteral(function_ref<InFlightDiagnostic()> emitError,
----------------
River707 wrote:

Why not add this method to Parser instead of a floating function? All of the users seem to already have a parser instance? That would remove the need for the `emitError` function and make this a bit cleaner.

https://github.com/llvm/llvm-project/pull/116172