RE: confusing error message
Good point. Error message improved, regression test added. Thanks for the suggestion Simon | -Original Message- | From: Dean Herington [mailto:[EMAIL PROTECTED]] | Sent: 05 February 2003 19:14 | To: [EMAIL PROTECTED] | Subject: confusing error message | | buzzard(118)% cat Bug5.hs | import Control.Monad.State | data S = S Int | newtype M a = M (StateT S IO a) | deriving (Monad) | | main = return () | buzzard(119)% ghc -c Bug5.hs | | Bug5.hs:3: | Can't make a derived instance of `Monad M' | (too hard for cunning newtype deriving) | When deriving instances for type `M' | buzzard(120)% ghc --version | The Glorious Glasgow Haskell Compilation System, version 5.04.2 | | | The real problem above is that I forgot to enable extensions. It would | be helpful if the error message indicated that as the problem. | | | ___ | Glasgow-haskell-bugs mailing list | [EMAIL PROTECTED] | http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs ___ Glasgow-haskell-bugs mailing list [EMAIL PROTECTED] http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
Re: Confusing error message in case of layout-related syntax erro r
Simon wrote: I know about this one, but haven't got around to fixing it yet. The parse error is on a semicolon generated by the layout system, as you probably guessed, which is why there's no token available to print in the error message. Any thoughts on what a suitable error message should be? I'd thought something along the lines of Main.hs:3: parse error on input ';' (inserted by layout) My students find this kind of error message (Hugs says "unexpected ';'") the most confusing and frustrating thing in all of Haskelldom. If you know that the (virtual) semicolon has been inserted by layout, how about a message like: Main.hs:3: indentation error or Main.hs:3: incorrect indentation or Main.hs:3: parse error due to incorrect indentation (I think the term "indentation" means more to them than "layout", but there might be good reasons to say "layout error" instead). Mark. Dr Mark Utting, Senior Lecturer Department of Computer Science School of Computing and Mathematical Sciences The University of Waikato Tel: +64 7 838 4791 Private Bag 3105Fax: +64 7 838 4155 HamiltonEmail: [EMAIL PROTECTED] New Zealand Web: http://www.cs.waikato.ac.nz/~marku
RE: Confusing error message in case of layout-related syntax error
Main.hs:3: parse error on input ';' (inserted by layout) Incomplete expression on previous line or incorrect indentation of the current line. I think Manuel's suggestion is a definite improvement here. For interest, another Haskell compiler, nhc98, gives the message 4:5 Found ;' but expected one of ) , which is perhaps a little bit cryptic, but could be improved to 4:5 Found end-of-definition but expected one of ) , which I think is very clear. Regards, Malcolm
RE: Confusing error message in case of layout-related syntax erro r
For foo = let x = (1, 2 y = 3 in fst x + y GHC 4.04 gives me Main.hs:3: parse error on input `' Interesting, but not very informative ;-) I know about this one, but haven't got around to fixing it yet. The parse error is on a semicolon generated by the layout system, as you probably guessed, which is why there's no token available to print in the error message. Any thoughts on what a suitable error message should be? I'd thought something along the lines of Main.hs:3: parse error on input ';' (inserted by layout) re: the multi-line string literals, this one should be easier to fix. I'll take a look at it today. Cheers, Simon
Re: Confusing error message
I encountered a confusing error message, which you can reproduce with type P a = Maybe a instance Monad P where (=) = error "foo" return = error "bar" I get bug.hs:5: `P' should have 1 argument, but has been given 0 . Would it be better if it said Type synonym constructor P should have 1 argument, but has been given 0 Haskell requires that type synonyms are never partially applied; that's what's being complained about here. If you did fully apply it, GHC 3.1 (without -fglasow-exts) would then complain about making an instance of a type synonym. At the moment, though, it trips over the mal-formed type expression first. That's what I guessed, but I reckon that it may be a bit difficult to spot for people who are not so familiar with the details of constructor classes. But, maybe it is too much fuzz to check for this special situation explicitly. I wonder whether it would be helpful to add a comment like (or if this is a instance declaration, type synonyms are not allowed) to the message. Manuel
Re: Confusing error message
I encountered a confusing error message, which you can reproduce with type P a = Maybe a instance Monad P where (=) = error "foo" return = error "bar" I get bug.hs:5: `P' should have 1 argument, but has been given 0 . Would it be better if it said Type synonym constructor P should have 1 argument, but has been given 0 Haskell requires that type synonyms are never partially applied; that's what's being complained about here. If you did fully apply it, GHC 3.1 (without -fglasow-exts) would then complain about making an instance of a type synonym. At the moment, though, it trips over the mal-formed type expression first. Does that make sense? Any suggestions for improving the error message in a way that would have made sense to you at the time? Simon
