#8607: Invalid location reported for type constructors -------------------------------------+------------------------------------ Reporter: edsko | Owner: goldfire Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.7 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by edsko): FYI, I sent the following email to SPJ, but he won't be back until next year: In the commit message you say "The only wrinkle is that, since we don't have the original declaration, we don't have its `SrcSpan` to put in the error message". But the function you are talking about is checkValidDecl, right? And that takes a Located Name as argument -- would it be possible to use the nameSrcSpan of the Name instead of the error message? I.e., have the def site of the Name be the entire declaration of the type, but the location of the identifier itself still the actual location of the identifier? Changing the `SrcSpan` associated with the identifier itself is troublesome for IDEs that want to know information about identifiers at particular locations. I'm not 100% how to go about this though, because the parser obviously generates a `RdrName`, not a Name, so we don't yet have the distinction between the location of a name (Located Name) and its nameSrcSpan. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/8607#comment:7 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler