> And yes, happy is up-to-date, perhaps too much so.
> I get various mismatch errors. For example % is used in
> ParseIFace.hs, but GHC can't
> find it. When I tried getting rid of that, I got other messages:
> ParseIface.hs:5633:
> Couldn't match `EncodedFS' against `[Char]'
> Expected type: EncodedFS
> Inferred type: [Char]
> In the first argument of `HappyAbsSyn54', namely `(("!"))'
> In the right-hand side of an equation for `reduction':
> HappyAbsSyn54 (("!"))
very strange - it looks like the SLIT macro is expanding to the wrong thing.
This should only happen if you're compiling with a very old version of GHC,
take a look at fptools/ghc/compiler/HsVersions.h for the details.
Basically, something has gone wrong while cpp'ing ParseIface.hs. You'll
have to send me a full -v output from the compilation so we can track this
down.
Cheers,
Simon
