- Hat - Haskell.org

... and another?
by Thomas Davie 01 Nov '05

01 Nov '05

I think there's a bug in unique naming for instances of (originally) derived classes. Putting this in one file: data Result = Result { resultModule :: TagString, resultName :: TagString, resultType :: TagString, resultInfo :: [Reason], resultScore :: Score } deriving Show and this in another: data ClassTable = ClassTable (Map.Map String [TypeMatch]) deriving Show generates code that has the function a25v28v25v31showsPrec in both modules, and thus causes this error: Hat/MatchClass.hs:30:12: Ambiguous occurrence `a25v28v25v31showsPrec' It could refer to either `Hat.MatchClass.a25v28v25v31showsPrec', defined at Hat/MatchClass.hs:368:0 or `Hat.Result.a25v28v25v31showsPrec', imported from Hat.Result at Hat/MatchClass.hs:11:0-16 Bob

2 1

Another hat-trans bug?
by Thomas Davie 01 Nov '05

01 Nov '05

Having worked around the other bug while Malcolm tries to figure it out (sorry Malcolm, didn't mean to land such a big chunk of work on your lap), I think I've hit another hat-trans bug. The following code: lookupAlias :: AliasTable -> Type -> Type lookupAlias (AliasTable table) (TLit x) = typ where (0, typ) = Map.findWithDefault (0, TLit x) x table lookupAlias a@(AliasTable table) (TList (TLit x:xs)) = if isJust res then some else none where res = Map.lookup x table none = TList (TLit x : map (lookupAlias a) xs) Just (n, typ) = res some = if n == length xs then mapNumber (xs !!) typ else error "lookupAlias: mismatch" lookupAlias _ x = x Gets translated to: ... lots of stuff ... hlookupAlias (T.R (AliasTable ftable) _) (T.R (TLit fx) _) p = gtyp p58v43v58v45 p where gtyp ptyp p = T.constUse ptyp p styp j59v11v59v18typ = case T.cguard p59v11v59v18 p 148 > (T.ap2 p59v12v59v12 p (p59v12v59v12 Hat.Prelude.!== p) fv59v12v59v12n (T.ap1 p59v12v59v12 p (Hat.PreludeBasic.gfromInteger p59v12v59v12 p) 150 > (T.conInteger p59v12v59v12 p 0))) (\ p -> h ftyp p) (\ p -> T.fatal p) of T.R (T.Tuple2 fv59v12v59v12n ftyp) ktyp -> (ktyp,ftyp) _ -> T.fatal p where h ftyp p = T.app3 p59v22v59v60 p59v22v59v40 p Hat.Map.afindWithDefault Hat.Map.hfindWithDefault (T.con2 p59v42v59v52 p T.Tuple2 T.aTuple2 (T.ap1 p59v43v59v43 p (Hat.PreludeBasic.gfromInteger p59v43v59v43 p) (T.conInteger p59v43v59v43 p 0)) (T.con1 p59v46v59v51 p TLit aTLit fx)) fx ftable h _ p = T.fatal p ... lots of stuff... Note in the case statement that the variable fv59v12v59v12n is bound in the second case, but used in the first... As is the variable ftyp. This of course leads to these errors: Hat/TypeAlias.hs:148:63: Not in scope: `fv59v12v59v12n' Hat/TypeAlias.hs:150:56: Not in scope: `ftyp' Bob

2 1

Any guesses?
by Thomas Davie 31 Oct '05

31 Oct '05

Anyone know how to fix the following type error is translated code? Hat/MySet.hs:636:60: Couldn't match `Hat.Hat.R (Hat.Hat.List a)' against `Hat.Hat.RefExp' Expected type: Hat.Hat.R (Hat.Hat.Fun a b2) -> Hat.Hat.RefExp -> Hat.Hat.R (Hat.Hat.Fun b1 (Hat.Hat.List b)) Inferred type: Hat.Hat.R (Hat.Hat.Fun a b2) -> Hat.Hat.R (Hat.Hat.List a) -> Hat.Hat.RefExp -> Hat.Hat.R (Hat.Hat.List b2) In the fifth argument of `Hat.Hat.app1', namely `Hat.PreludeBasic.hmap' In the fourth argument of `Hat.Hat.ap2', namely `(Hat.Hat.app1 p391v20v391v29 p391v20v391v27 p Hat.PreludeBasic.amap Hat.PreludeBasic.hmap ff)' The original code looks like this (and does not cause type errors when compiled normally). map :: (Ord a, Ord b) => (a->b) -> Set a -> Set b map f = fromList . List.map f . toList The translated code looks like this: gmap :: (Ord a,Ord b) => T.RefSrcPos -> T.RefExp -> T.R (T.Fun (T.Fun a b) (T.Fun (Set a) (Set b))) hmap :: (Ord a,Ord b) => (T.R (T.Fun a b)) -> T.RefExp -> T.R (T.Fun (Set a) (Set b)) gmap pmap p = T.fun1 amap pmap p hmap hmap ff p = T.ap2 p391v9v391v38 p (p391v18v391v18 !. p) (gfromList p391v9v391v16 p) (T.ap2 p391v20v391v38 p (p391v31v391v31 !. p) 636 > (T.app1 p391v20v391v29 p391v20v391v27 p Hat.List.amap Hat.List.hmap ff) (gtoList p391v33v391v38 p)) I have not a clue what I'm up to when it gets this fun. Bob

3 3

Before I embark on a voyage doomed to failure
by Thomas Davie 28 Oct '05

28 Oct '05

I'm trying to get some reasonably bit examples to test hat-detect with... Problem is, most of them need Data.map to work... Is there any reason why I shouldn't embark on shoving it through hat-trans? Bob

3 3

Summary page
by Thomas Davie 21 Oct '05

21 Oct '05

I thought we really needed a page that summarises how each tool works, and I was bored on a Friday, so I had a play. http://www.cs.kent.ac.uk/people/rpg/tatd2/icons/ Who want's to criticise? All the text is taken from manuals at the mo, it probably needs some better copy written. Bob p.s. I know it renders right in Safari and Firefox, but if someone could confirm how badly internet explorer is broken, that'd be useful.

2 4

New hat logo
by Thomas Davie 19 Oct '05

19 Oct '05

The hat logo on the site is looking decidedly dated, so I've made a slightly updated version with a nice high quality line art outline... I just wanted to get people's approval before putting it into cvs. It's at http://www.cs.kent.ac.uk/people/rpg/tatd2/hat.pdf for now. Comments are most welcome. Bob

1 0

Pre-check-in warning
by Thomas Davie 14 Oct '05

14 Oct '05

Hi all, I just wanted to warn various people that I was going to do a rather monolithic check in to cvs in a little bit. Hopefully after this one, my changes should be in smaller and neater increments. There's a few relatively large changes. Most obvious to you will be that I've completely re-written hat-detect (and an EDT module to go behind it). Hat-delta now works (although it'll probably stack overflow on larger programs, I've got plans for fixing that), and will be checked in too. Other than that there are a few movements of bits of code. Most notably from the old HatDetect, and from HatExplore, into separate modules that can be loaded from elsewhere. Yell (relatively) fast if you have major objections to this. Bob

1 0

Compiling Hat on Windows
by Neil Mitchell 18 Sep '05

18 Sep '05

Hi, I decided to try compiling Hat on Windows, just the viewing tools, using only the standard Windows tools and those things which get installed by GHC for Windows. It worked with a few tweaks/changes, so I thought I'd document them here: cd ...\hattools gcc -c *.c -DFILEVERSION="\"2.04\"" -DVERSION="\"2.04\"" (hat-anim.c fails totally) ghc --make HatCover.hs -cpp -fglasgow-exts artutils.o finitemap.o pathutils.o ghc --make BlackHat.hs -cpp -fglasgow-exts artutils.o finitemap.o pathutils.o nontermutils.o gcc hat-check.c In addition I also commented out the 4th line in finitemap.h, #include "checkglib.h". Creating an empty checkglib.h would have solved this one. I also commented out the signal lines in hat-check. I also copied art.h into the tools directory, although an -I flag would have worked here instead. I also defined ntohl myself, since its not defined in any of the default C libraries on Windows. unsigned long ntohl(unsigned long netlong) { return netlong; } With these changes I could get compiled versions of most of the tools. Thanks Neil

2 1

Fwd: buggy sorting function
by Thomas Davie 16 Sep '05

16 Sep '05

Evening all... I appear to have broken hat-trans... Stephan Kahrs sent me this program to try hat-delta on, but it appears that hat barfs before I even get that far... lappybob:~/Documents/Work/Hat test cases/Hell Sort tatd2$ hmake -hat hellSort hat-trans hellSort.hs Creating directories Hat Wrote Hat/hellSort.hs ghc -c -package hat -o Hat/hellSort.o Hat/hellSort.hs ghc -package hat -o hellSort Hat/hellSort.o lappybob:~/Documents/Work/Hat test cases/Hell Sort tatd2$ ./hellSort [6,4,2,9,10,9,3] > [2,3,4,6,9,10,9] ^CKilled lappybob:~/Documents/Work/Hat test cases/Hell Sort tatd2$ hat-detect hellSort hat-detect (error): file hellSort.hat is too short lappybob:~/Documents/Work/Hat test cases/Hell Sort tatd2$ hat-check hellSort hat-check (error): file hellSort.hat is too short Begin forwarded message: > From: "S.M.Kahrs" <S.M.Kahrs(a)kent.ac.uk> > Date: 16 September 2005 15:34:07 BDT > To: tatd2(a)kent.ac.uk > Cc: S.M.Kahrs(a)kent.ac.uk > Subject: buggy sorting function > > > The sorting function is called 'treesort'. > > module Main where > {- Ulrich Furbach, Lothar Schmitz -} > > treesort :: Ord a => [a] -> [a] > treesort = sort . establish > > establish f = eaux (length f) ((0,[1]),f) > > eaux n a@((u,ss),f) = > let sm = head ss > sm1 = ss !! 1 > sm2 = ss !! 2 > in > if u==n-1 then ((u,ss),siftav a) > else if u>=sm && sm == pre sm1 then eaux n ((u+1,((sm+sm1 > +1):drop 2 ss)),siftbin((u,sm),f)) > else if n-u>=pre sm then eaux n((u+1,1:ss),siftbin((u,sm),f)) > else eaux n((u+1,1:ss), siftav((u,ss),f)) > > sort ((u,ss),f) = > if u==0 then take 1 f > else sort (rearrange((u,ss),f)) ++ [f !! u] > > rearrange((u,sm:ss),f) = > if sm==1 then ((u-1,ss),f) > else ((u-1,tt),siftsp((u-1,tt),siftsp((u-1-tlast,tt0),f))) > where > tt0 = t1: ss > tt = tlast: tt0 > (tlast,t1)=preaux sm 0 1 1 > sm1 = ss !! 0 > sm2 = ss !! 1 > > siftsp ((u,ss),f) = > let sm = head ss > sm1 = ss !! 1 > sm2 = ss !! 2 > m = length ss > in > if m==1 || (f !! (u-sm)) <= (f !! u) then f > else siftav ((u-sm,tail ss),swap f (u-sm)u) > > siftav((u,ss),f) = > if u<sm || (f !! (u-sm))<= (f !! u) then siftbin((u,sm),f) > else if sm==1 || (f !! g)<= (f !! (u-sm)) then siftav((u- > sm,tail ss),swap f (u-sm) u) > else siftbin ((g,sg),swap f u g) > where > (p2,p1)=preaux sm 0 1 1 > (g,sg)= if (f !! (u-1))<=(f !!(u-1-p2)) then (u-1-p2,p1) > else (u-1,p2) > sm = head ss > sm1 = ss !! 1 > sm2 = ss !! 2 > > > siftbin ((u,s),f) = > if s==1 || (f !! g)<=(f!! u) then f > else siftbin((g,sg),swap f g u) > where > (p2,p1)=preaux s 0 1 1 > (g,sg)= if (f !! (u-1))<= (f !!(u-1-p2)) then (u-1-p2,p1) > else (u-1,p2) > > > swap :: [a] -> Int -> Int -> [a] > swap xs i j = if i>j then swap' xs j i else swap' xs i j > swap' :: [a] -> Int -> Int -> [a] > swap' xs i j = {- i<j -} > let (a,b)=splitAt i xs > (c,d)=splitAt (j-i) b > in a++[head d]++ tail c ++ [head c] ++ tail d > > > pre n = snd (preaux n 0 1 1) > > preaux n a b c = if n<=c then (a,b) > else preaux n b c (b+c+1) > > > tsort :: [Int] -> [Int] > tsort = treesort > > main = > do > putStr "> " > xs <- getLine > if null xs then return () > else ( print (tsort (read xs))) >> main >

4 7

Data.Generics missing from Hat
by Ken T Takusagawa 15 Aug '05

15 Aug '05

I get the error hat-trans: /usr/local/include/hat/Data/Generics.hx: openFile: does not exist (No such file or directory) when compiling a file with "import Data.Generics". Is there a fix or a workaround? --ken

3 2