- Cvs-nhc98 - Haskell.org

Just formatting.
by malcolm＠sparud.net 13 Feb '02

13 Feb '02

malcolm: Wed Feb 13 17:12:15 CET 2002 Update of /usr/src/master/nhc/src/hat/tools In directory hinken:/tmp/cvs-serv21062 Modified Files: PrettyExp.hs Log Message: Just formatting.

1 0

Some formatting, some tidying of interactive prompts, and
by malcolm＠sparud.net 13 Feb '02

13 Feb '02

malcolm: Wed Feb 13 17:11:33 CET 2002 Update of /usr/src/master/nhc/src/hat/tools In directory hinken:/tmp/cvs-serv21033 Modified Files: HatDetect.hs HatObserve.hs Log Message: Some formatting, some tidying of interactive prompts, and added tunnels to jump to hat-explore.

1 0

Support newtype data constructors and ~ in -T tracing vers
by olaf＠sparud.net 13 Feb '02

13 Feb '02

olaf: Wed Feb 13 17:03:33 CET 2002 Update of /usr/src/master/nhc/src/compiler98 In directory hinken:/tmp/cvs-serv20974/src/compiler98 Modified Files: DbgTrans.hs Info.hs Log Message: Support newtype data constructors and ~ in -T tracing version. The traced code used to be more strict than it should be in arguments of constructors of newtypes and variables occuring in ~ patterns. Variables in irrefutable patterns may never be evaluated. Hence these variables need to be treated as let not as lambda bound. Now before actual trace transformation replace every irrefutable pattern in a refutable pattern by a new variable and add an appropriate pattern binding. This leads to a tracing program with the original strictness. The created trace is basically right, except that the newly introduced variable appears in the trace. This is because I reused the normal transformation of pattern bindings; to obtain the desired trace special handling would be necessary. Then change from lambda to let-bound variables is a bit of a hack. Still need to do something similar in the portable version. Example program that did not work before: newtype Seq a = Seq ([a] -> [a]) (+++) :: Seq a -> Seq a -> Seq a (Seq fxs) +++ (Seq fys) = Seq (fxs . fys) left (Seq fxs) = head (fxs []) main = print (left (Seq (True:) +++ undefined))

1 0

Added module TraceDerive of the new deriving mechanism for
by olaf＠sparud.net 12 Feb '02

12 Feb '02

olaf: Tue Feb 12 11:44:36 CET 2002 Update of /usr/src/master/nhc/src/compiler98 In directory hinken:/tmp/cvs-serv9901/src/compiler98 Modified Files: PrettySyntax.hs Added Files: TraceDerive.hs Log Message: Added module TraceDerive of the new deriving mechanism for tracing, which I had forgotten last time. Also corrected a bug in pretty printing (newtype).

1 0

Add explicit import list for Flags.
by malcolm＠sparud.net 12 Feb '02

12 Feb '02

malcolm: Tue Feb 12 11:42:47 CET 2002 Update of /usr/src/master/nhc/src/compiler98 In directory hinken:/tmp/cvs-serv9870 Modified Files: Depend.hs Log Message: Add explicit import list for Flags.

1 0

Fix export list so it will compile.
by malcolm＠sparud.net 12 Feb '02

12 Feb '02

malcolm: Mon Feb 11 22:39:34 CET 2002 Update of /usr/src/master/nhc/src/compiler98 In directory hinken:/tmp/cvs-serv19335 Modified Files: TraceId.hs Log Message: Fix export list so it will compile.

1 0

Implemented "deriving Eq" in portable transformation.
by olaf＠sparud.net 09 Feb '02

09 Feb '02

olaf: Fri Feb 8 19:55:31 CET 2002 Update of /usr/src/master/nhc/src/compiler98 In directory hinken:/tmp/cvs-serv19128/src/compiler98 Modified Files: HatTrans.hs Makefile PrettySyntax.hs Syntax.hs TokenId.hs TraceId.hs TraceTrans.hs Log Message: Implemented "deriving Eq" in portable transformation. The general mechanism and the specific one for the class Eq are implemented. Deriving for the remaining classes still needs to be done. The derived code is untraced. I use a hack for determining the context of an instance. It is the context that will "commonly" be needed, but it is not correct in all cases. A correct implementation would require much more effort: context reduction (with knowledge about all classes, instances and type synonyms) and a fixed point iteration.

1 0

Implemented "deriving Eq" in portable transformation.
by olaf＠sparud.net 09 Feb '02

09 Feb '02

olaf: Fri Feb 8 19:55:32 CET 2002 Update of /usr/src/master/nhc/src/hat/lib In directory hinken:/tmp/cvs-serv19128/src/hat/lib Modified Files: PreludeBasic.hs Log Message: Implemented "deriving Eq" in portable transformation. The general mechanism and the specific one for the class Eq are implemented. Deriving for the remaining classes still needs to be done. The derived code is untraced. I use a hack for determining the context of an instance. It is the context that will "commonly" be needed, but it is not correct in all cases. A correct implementation would require much more effort: context reduction (with knowledge about all classes, instances and type synonyms) and a fixed point iteration.

1 0

Updated Hat user manual.
by olaf＠sparud.net 08 Feb '02

08 Feb '02

olaf: Thu Feb 7 20:34:49 CET 2002 Update of /usr/src/master/nhc/docs/hat In directory hinken:/tmp/cvs-serv10954 Modified Files: hatuser.tex Log Message: Updated Hat user manual.

1 0

Improvements to hat-explore.
by malcolm＠sparud.net 07 Feb '02

07 Feb '02

malcolm: Thu Feb 7 18:27:09 CET 2002 Update of /usr/src/master/nhc/src/hat/tools In directory hinken:/tmp/cvs-serv10105 Modified Files: HatExplore.hs Makefile SExp.hs utils.c utils.h Added Files: LowLevel.hs Log Message: Improvements to hat-explore. * Use a different low-level interface which gives correct parent information for identifiers, constructors, and literals. * In the state, store a stack of sub-expression contexts, not just expressions. Thus, when you go backwards up the display, the cursor returns to the previously selected subexpression, not the whole line.

1 0