ghc-base.git
21 years ago[project @ 2002-09-02 16:40:55 by simonpj]
simonpj [Mon, 2 Sep 2002 16:40:55 +0000 (16:40 +0000)]
[project @ 2002-09-02 16:40:55 by simonpj]
Un-break head build due to (:) export

21 years ago[project @ 2002-09-02 16:33:31 by ross]
ross [Mon, 2 Sep 2002 16:33:31 +0000 (16:33 +0000)]
[project @ 2002-09-02 16:33:31 by ross]
non-GHC compilers should also import Numeric.

21 years ago[project @ 2002-09-02 11:20:50 by ross]
ross [Mon, 2 Sep 2002 11:20:50 +0000 (11:20 +0000)]
[project @ 2002-09-02 11:20:50 by ross]
Minor rearrangement: the implementation of System.Exit.exitWith is
GHC-specific, while exitFailure is portable.

21 years ago[project @ 2002-08-31 08:37:56 by simonpj]
simonpj [Sat, 31 Aug 2002 08:37:56 +0000 (08:37 +0000)]
[project @ 2002-08-31 08:37:56 by simonpj]
Remove GHC.Base.assert until I can figure out what happened

21 years ago[project @ 2002-08-30 14:54:58 by simonpj]
simonpj [Fri, 30 Aug 2002 14:54:58 +0000 (14:54 +0000)]
[project @ 2002-08-30 14:54:58 by simonpj]
Move defn of hs_fileno inside ifdef

21 years ago[project @ 2002-08-30 14:29:51 by simonpj]
simonpj [Fri, 30 Aug 2002 14:29:52 +0000 (14:29 +0000)]
[project @ 2002-08-30 14:29:51 by simonpj]
* Do not export built-in syntax (H98 conformance)
* Add 'assert' to GHC.Base

(no need to merge to stable branch)

21 years ago[project @ 2002-08-30 14:19:18 by simonpj]
simonpj [Fri, 30 Aug 2002 14:19:18 +0000 (14:19 +0000)]
[project @ 2002-08-30 14:19:18 by simonpj]
Extra space after $ to avoid conflict with Template Haskell

21 years ago[project @ 2002-08-30 13:43:57 by stolz]
stolz [Fri, 30 Aug 2002 13:43:57 +0000 (13:43 +0000)]
[project @ 2002-08-30 13:43:57 by stolz]
Haddock-ise with comments from library report

21 years ago[project @ 2002-08-30 13:27:42 by stolz]
stolz [Fri, 30 Aug 2002 13:27:42 +0000 (13:27 +0000)]
[project @ 2002-08-30 13:27:42 by stolz]
Haddock-ise.

21 years ago[project @ 2002-08-30 13:00:31 by simonpj]
simonpj [Fri, 30 Aug 2002 13:00:31 +0000 (13:00 +0000)]
[project @ 2002-08-30 13:00:31 by simonpj]
Remove bogus parError; MERGE TO STABLE

21 years ago[project @ 2002-08-30 12:32:44 by stolz]
stolz [Fri, 30 Aug 2002 12:32:44 +0000 (12:32 +0000)]
[project @ 2002-08-30 12:32:44 by stolz]
- Haddock-ise with comments from library report
- FIXME: Haddock doesn't support nested enumerations.

21 years ago[project @ 2002-08-30 12:25:15 by stolz]
stolz [Fri, 30 Aug 2002 12:25:15 +0000 (12:25 +0000)]
[project @ 2002-08-30 12:25:15 by stolz]
Express 'killThread' in terms of 'throwTo'

21 years ago[project @ 2002-08-30 10:46:10 by stolz]
stolz [Fri, 30 Aug 2002 10:46:10 +0000 (10:46 +0000)]
[project @ 2002-08-30 10:46:10 by stolz]
Haddock-ised. The previous version already did contain a lot of
documentation which just wasn't in the right shape. The
current documentation diverges slightly from the Library Report.

21 years ago[project @ 2002-08-30 07:56:48 by stolz]
stolz [Fri, 30 Aug 2002 07:56:48 +0000 (07:56 +0000)]
[project @ 2002-08-30 07:56:48 by stolz]
Fix sample way of expressing 'killThread' in docs.

From: Dean Herington <heringto@cs.unc.edu>

PLEASE MERGE

21 years ago[project @ 2002-08-29 16:39:42 by stolz]
stolz [Thu, 29 Aug 2002 16:39:42 +0000 (16:39 +0000)]
[project @ 2002-08-29 16:39:42 by stolz]
- Haddock-ise with comments from library report
- The chapter "Deriving Instances of Ix" doesn't end up in the "Contents"
    section (yet), although it should.

21 years ago[project @ 2002-08-29 16:05:59 by stolz]
stolz [Thu, 29 Aug 2002 16:05:59 +0000 (16:05 +0000)]
[project @ 2002-08-29 16:05:59 by stolz]
Haddock-ise with comments from library report

21 years ago[project @ 2002-08-29 16:03:57 by stolz]
stolz [Thu, 29 Aug 2002 16:03:57 +0000 (16:03 +0000)]
[project @ 2002-08-29 16:03:57 by stolz]
Add replicateM[_] and foldM_

21 years ago[project @ 2002-08-29 11:49:10 by simonmar]
simonmar [Thu, 29 Aug 2002 11:49:11 +0000 (11:49 +0000)]
[project @ 2002-08-29 11:49:10 by simonmar]
Make the readline binding into a hierarchical library and put it in
its own package (for licensing reasons).

As part of this, I moved fdToHandle from PosixIO into GHC.Handle (and
updated the code to avoid using _casm_) since Readline requires it.

I've also updated Readline to use the latest FFI syntax.

21 years ago[project @ 2002-08-29 11:29:40 by simonmar]
simonmar [Thu, 29 Aug 2002 11:29:40 +0000 (11:29 +0000)]
[project @ 2002-08-29 11:29:40 by simonmar]
Foreign.C.TypesISO has been merged into Foreign.C.Types now.

21 years ago[project @ 2002-08-29 09:24:21 by simonmar]
simonmar [Thu, 29 Aug 2002 09:24:22 +0000 (09:24 +0000)]
[project @ 2002-08-29 09:24:21 by simonmar]
Move library-project documentation to a better place, and add the
current hierarchy spec.

21 years ago[project @ 2002-08-29 05:11:41 by ross]
ross [Thu, 29 Aug 2002 05:11:41 +0000 (05:11 +0000)]
[project @ 2002-08-29 05:11:41 by ross]
The functions rawSystem, withArgs and withProgName aren't supported
by Hugs yet.

21 years ago[project @ 2002-08-28 14:30:12 by simonpj]
simonpj [Wed, 28 Aug 2002 14:30:12 +0000 (14:30 +0000)]
[project @ 2002-08-28 14:30:12 by simonpj]
Fix a lexing bug: "\SOH" could mean "\SO" followed by "H"
or "\SOH".  The Report specifies the latter.

MERGE TO STABLE

21 years ago[project @ 2002-08-28 13:59:19 by simonmar]
simonmar [Wed, 28 Aug 2002 13:59:19 +0000 (13:59 +0000)]
[project @ 2002-08-28 13:59:19 by simonmar]
- Move rawSystem from SystemExts to System.Cmd.
- Move withArgv and withProgName from SystemExts to System.Environment

These functions are still exported by SystemExts for compatibility.

21 years ago[project @ 2002-08-27 14:44:31 by simonmar]
simonmar [Tue, 27 Aug 2002 14:44:31 +0000 (14:44 +0000)]
[project @ 2002-08-27 14:44:31 by simonmar]
We really should export the representations of Ptr and FunPtr from here.

21 years ago[project @ 2002-08-25 09:16:07 by panne]
panne [Sun, 25 Aug 2002 09:16:07 +0000 (09:16 +0000)]
[project @ 2002-08-25 09:16:07 by panne]
Fixed bug in documentation introduced by "Mr. Haddock" himself. :-)

21 years ago[project @ 2002-08-23 22:10:43 by sof]
sof [Fri, 23 Aug 2002 22:10:43 +0000 (22:10 +0000)]
[project @ 2002-08-23 22:10:43 by sof]
ioeGetErrorString: don't "show" the string from a user error.

Merge to STABLE.

21 years ago[project @ 2002-08-21 10:43:22 by simonmar]
simonmar [Wed, 21 Aug 2002 10:43:22 +0000 (10:43 +0000)]
[project @ 2002-08-21 10:43:22 by simonmar]
Oops, unpackCStringUtf8# had rotted at some point, and no-one noticed.
Some parentheses are required because the relative precedences of
uncheckIShiftL# and +# (the default precedences, AFAICT) give the
wrong meaning to some expressions.

MERGE TO STABLE

21 years ago[project @ 2002-08-20 10:03:05 by simonmar]
simonmar [Tue, 20 Aug 2002 10:03:05 +0000 (10:03 +0000)]
[project @ 2002-08-20 10:03:05 by simonmar]
The Typeable instances were missing in the __GLASGOW_HASKELL__ case;
add them.

21 years ago[project @ 2002-08-16 11:38:57 by simonmar]
simonmar [Fri, 16 Aug 2002 11:38:57 +0000 (11:38 +0000)]
[project @ 2002-08-16 11:38:57 by simonmar]
Some extra docs for addFinalizer, describing why using addFinalizer on
a ForeignPtr isn't the same as using addForeignPtrFinalizer.

21 years ago[project @ 2002-08-16 11:38:04 by simonmar]
simonmar [Fri, 16 Aug 2002 11:38:04 +0000 (11:38 +0000)]
[project @ 2002-08-16 11:38:04 by simonmar]
Data.Array.IArray should export Array too

21 years ago[project @ 2002-08-08 22:29:28 by reid]
reid [Thu, 8 Aug 2002 22:29:28 +0000 (22:29 +0000)]
[project @ 2002-08-08 22:29:28 by reid]
Hugs provides makeForeignPtr instead of newForeignPtr.

It is hoped that these macros overcome the difference.

+ #ifdef __HUGS__
+ #define MAKE_ARRAY(x) makeForeignPtr (x) free
+ #else
+ #define MAKE_ARRAY(x) newForeignPtr (x) (free (x))
+ #endif

I could probably get away with introducing a Haskell functions instead
of a macro..

[Untested since Data.Array.Base requires the pattern guard extension
so I can't load it.  Still, I think this will be ready to go once we
fix D.A.B]

21 years ago[project @ 2002-08-08 06:25:05 by ross]
ross [Thu, 8 Aug 2002 06:25:05 +0000 (06:25 +0000)]
[project @ 2002-08-08 06:25:05 by ross]
Add hTell to Hugs.

21 years ago[project @ 2002-08-07 12:17:59 by ross]
ross [Wed, 7 Aug 2002 12:17:59 +0000 (12:17 +0000)]
[project @ 2002-08-07 12:17:59 by ross]
Renamed HugsStorable -> Hugs.Storable (copy of lib/exts/HugsStorable.hs)

21 years ago[project @ 2002-08-05 08:48:07 by simonmar]
simonmar [Mon, 5 Aug 2002 08:48:08 +0000 (08:48 +0000)]
[project @ 2002-08-05 08:48:07 by simonmar]
Remove the literate bits from this file: I'm trying to keep .lhs out
of the non-GHC parts of the libraries tree, we don't use it any more
and it clutters up the code.

21 years ago[project @ 2002-08-05 08:44:53 by simonmar]
simonmar [Mon, 5 Aug 2002 08:44:53 +0000 (08:44 +0000)]
[project @ 2002-08-05 08:44:53 by simonmar]
Move the import of System.IO.Unsafe into the #else part of #ifdef
__GLASGOW_HASKELL__, because it was causing an import loop.
(Alastair: that #else should probably be #elif __HUGS__, right?)

21 years ago[project @ 2002-08-05 08:29:17 by simonmar]
simonmar [Mon, 5 Aug 2002 08:29:17 +0000 (08:29 +0000)]
[project @ 2002-08-05 08:29:17 by simonmar]
Add missing GHC imports

21 years ago[project @ 2002-08-03 21:33:15 by reid]
reid [Sat, 3 Aug 2002 21:33:15 +0000 (21:33 +0000)]
[project @ 2002-08-03 21:33:15 by reid]
Directives like this

   {-# OPTIONS -#include "HsBase.h" #-}

aren't portable so I added this:

   foreign import ccall unsafe "HsBase.h ghcErrno" _errno :: Ptr CInt
                                ^^^^^^^^
                                ^^^^^^^^
                                ^^^^^^^^

21 years ago[project @ 2002-08-03 21:01:26 by reid]
reid [Sat, 3 Aug 2002 21:01:26 +0000 (21:01 +0000)]
[project @ 2002-08-03 21:01:26 by reid]
Final bout of changes to make things work with Hugs.

- I'd forgotten to include MachDep.h to the SIZEOF_* and ALIGNMENT_*
  macros (breaking GHC as well)
- I'd missed the use of one of GHC's type extensions.

21 years ago[project @ 2002-08-03 20:14:23 by reid]
reid [Sat, 3 Aug 2002 20:14:24 +0000 (20:14 +0000)]
[project @ 2002-08-03 20:14:23 by reid]
oops again

21 years ago[project @ 2002-08-03 19:32:49 by reid]
reid [Sat, 3 Aug 2002 19:32:49 +0000 (19:32 +0000)]
[project @ 2002-08-03 19:32:49 by reid]
Oops

21 years ago[project @ 2002-08-03 19:32:16 by reid]
reid [Sat, 3 Aug 2002 19:32:17 +0000 (19:32 +0000)]
[project @ 2002-08-03 19:32:16 by reid]
Changes to make libs work with Hugs/FFI.

Highlights:

  Moved most of the body of base/GHC/Storable.lhs
  into base/Foreign/Storable.lhs since it is mostly portable.

  base/include/CTypes.h and base/Foreign/C/Types.hs both generated
  Typeable instances for CChar and friends until I fixed them.

21 years ago[project @ 2002-08-02 12:26:36 by simonmar]
simonmar [Fri, 2 Aug 2002 12:26:36 +0000 (12:26 +0000)]
[project @ 2002-08-02 12:26:36 by simonmar]
Don't need -monly-3-regs now

21 years ago[project @ 2002-08-02 12:25:30 by simonmar]
simonmar [Fri, 2 Aug 2002 12:25:30 +0000 (12:25 +0000)]
[project @ 2002-08-02 12:25:30 by simonmar]
Don't need -monly-3-regs now.

21 years ago[project @ 2002-08-02 12:24:36 by simonmar]
simonmar [Fri, 2 Aug 2002 12:24:36 +0000 (12:24 +0000)]
[project @ 2002-08-02 12:24:36 by simonmar]
I can get away without -monly-3-regs for this file, now that GHC is
passing -fno-builtin to gcc.

21 years ago[project @ 2002-08-01 12:50:31 by simonpj]
simonpj [Thu, 1 Aug 2002 12:50:31 +0000 (12:50 +0000)]
[project @ 2002-08-01 12:50:31 by simonpj]
Add parens to make precence clear

21 years ago[project @ 2002-07-29 09:57:17 by simonmar]
simonmar [Mon, 29 Jul 2002 09:57:17 +0000 (09:57 +0000)]
[project @ 2002-07-29 09:57:17 by simonmar]
Remove non-existent and empty dirs from $(ALL_DIRS)

21 years ago[project @ 2002-07-29 09:54:40 by ross]
ross [Mon, 29 Jul 2002 09:54:41 +0000 (09:54 +0000)]
[project @ 2002-07-29 09:54:40 by ross]
Tweaked imports and includes for compatibility.

21 years ago[project @ 2002-07-26 13:23:38 by malcolm]
malcolm [Fri, 26 Jul 2002 13:23:38 +0000 (13:23 +0000)]
[project @ 2002-07-26 13:23:38 by malcolm]
Must start tweaking the libraries for NHC soon.

21 years ago[project @ 2002-07-26 12:12:33 by stolz]
stolz [Fri, 26 Jul 2002 12:12:33 +0000 (12:12 +0000)]
[project @ 2002-07-26 12:12:33 by stolz]
Fix typo.

21 years ago[project @ 2002-07-26 10:05:04 by stolz]
stolz [Fri, 26 Jul 2002 10:05:05 +0000 (10:05 +0000)]
[project @ 2002-07-26 10:05:04 by stolz]
Add some references. Requires upgrade to current haddock version (>2002-7-25)!

21 years ago[project @ 2002-07-26 02:36:57 by sof]
sof [Fri, 26 Jul 2002 02:36:57 +0000 (02:36 +0000)]
[project @ 2002-07-26 02:36:57 by sof]
gmtoff: according to the (POSIX  / Single Unix and MSVC) documentation of tzset(),
'timezone' is > 0 west of the Prime Meridian, so extend prev. commit to apply
to all platforms, not just Win32.

21 years ago[project @ 2002-07-25 14:14:36 by simonmar]
simonmar [Thu, 25 Jul 2002 14:14:36 +0000 (14:14 +0000)]
[project @ 2002-07-25 14:14:36 by simonmar]
Fix comment typos, from Reuben Thomas.

21 years ago[project @ 2002-07-24 09:57:21 by simonmar]
simonmar [Wed, 24 Jul 2002 09:57:21 +0000 (09:57 +0000)]
[project @ 2002-07-24 09:57:21 by simonmar]
tiny doc fixes

21 years ago[project @ 2002-07-23 22:04:36 by sof]
sof [Tue, 23 Jul 2002 22:04:36 +0000 (22:04 +0000)]
[project @ 2002-07-23 22:04:36 by sof]
inputReady(): using MsgWaitForMultipleObjects() instead of
WaitForMultipleObjects() on file handles is nicer from within a
message pump, but here it is less confusing to use the latter (and
simply just block message delivery for its duration.)

21 years ago[project @ 2002-07-23 18:50:54 by sof]
sof [Tue, 23 Jul 2002 18:50:54 +0000 (18:50 +0000)]
[project @ 2002-07-23 18:50:54 by sof]
gmtoff: MS CRT implementation of _tzset() (and _timezone) assumes that >0
represent positions west of the Prime Meridian. This module assumes the
opposite, so flip the sign of 'timezone' when compiling for mingw.

21 years ago[project @ 2002-07-23 14:52:46 by simonpj]
simonpj [Tue, 23 Jul 2002 14:52:47 +0000 (14:52 +0000)]
[project @ 2002-07-23 14:52:46 by simonpj]
Various precedence errors in the code
for read and show.  A couple (the show instances
for Ratio and Array) were actually errors in
the Library Report.   A couple more were to
do with whether the precedence of application is
9 (wrong) or 10 (right).

21 years ago[project @ 2002-07-23 10:46:27 by ross]
ross [Tue, 23 Jul 2002 10:46:27 +0000 (10:46 +0000)]
[project @ 2002-07-23 10:46:27 by ross]
Make Data.Unique work with Hugs.

21 years ago[project @ 2002-07-22 13:30:43 by simonmar]
simonmar [Mon, 22 Jul 2002 13:30:43 +0000 (13:30 +0000)]
[project @ 2002-07-22 13:30:43 by simonmar]
Make the new mergesort stable.

21 years ago[project @ 2002-07-22 11:37:38 by ross]
ross [Mon, 22 Jul 2002 11:37:39 +0000 (11:37 +0000)]
[project @ 2002-07-22 11:37:38 by ross]
Make Data.Array.{IArray,MArray,IO,ST} work with Hugs, splitting off
the IOArray part of Hugs.IOExts into Hugs.IOArray and (sadly) adding
Hugs.Array.Base, a cut-down version of Data.Array.Base.

21 years ago[project @ 2002-07-18 22:01:50 by sof]
sof [Thu, 18 Jul 2002 22:01:50 +0000 (22:01 +0000)]
[project @ 2002-07-18 22:01:50 by sof]
win32 console-based implementations of setCooked, {get,set}Echo

21 years ago[project @ 2002-07-18 22:01:07 by sof]
sof [Thu, 18 Jul 2002 22:01:07 +0000 (22:01 +0000)]
[project @ 2002-07-18 22:01:07 by sof]
helper functions for mucking about with Win32 consoles

21 years ago[project @ 2002-07-17 10:47:01 by ross]
ross [Wed, 17 Jul 2002 10:47:01 +0000 (10:47 +0000)]
[project @ 2002-07-17 10:47:01 by ross]
Make the libraries' Numeric module usable by Hugs.  Also deleted a chunk
of code inside #ifdef __HUGS__ -- this was unused, and the copy in Hugs
(now Hugs.Numeric) has since been improved.  (Since these functions are
portable, ideally they would be merged with the versions in GHC.Float)

21 years ago[project @ 2002-07-17 09:22:20 by simonmar]
simonmar [Wed, 17 Jul 2002 09:22:20 +0000 (09:22 +0000)]
[project @ 2002-07-17 09:22:20 by simonmar]
Need to #include <math.h> now that Stg.h doesn't.

21 years ago[project @ 2002-07-16 22:42:28 by sof]
sof [Tue, 16 Jul 2002 22:42:28 +0000 (22:42 +0000)]
[project @ 2002-07-16 22:42:28 by sof]
handleFinalizer: (also) use the handle's type/state
to decide whether or not go ahead with close()ing.

Little bit tidier than magic fd values.

21 years ago[project @ 2002-07-16 16:08:58 by ross]
ross [Tue, 16 Jul 2002 16:09:00 +0000 (16:09 +0000)]
[project @ 2002-07-16 16:08:58 by ross]
Add imports of Hugs.* modules (wrapped in #ifdef __HUGS__) to make these
modules work with Hugs.

21 years ago[project @ 2002-07-16 15:47:25 by ross]
ross [Tue, 16 Jul 2002 15:47:26 +0000 (15:47 +0000)]
[project @ 2002-07-16 15:47:25 by ross]
First stage of making the new libraries work with Hugs:

* wrap #ifndef __HUGS__ around stuff that Hugs defines in its monster
  prelude (because Hugs can't handle recursive imports).

* remove non-Haskell 98 export of [](..) from Data.List

* make the definitions of tuples in Data.Tuple GHC-only.

* make the export of the new Read stuff in Text.Read GHC-only, like its
  definition is.

21 years ago[project @ 2002-07-15 16:15:14 by simonmar]
simonmar [Mon, 15 Jul 2002 16:15:43 +0000 (16:15 +0000)]
[project @ 2002-07-15 16:15:14 by simonmar]
extra comma in export list

21 years ago[project @ 2002-07-15 16:02:11 by simonmar]
simonmar [Mon, 15 Jul 2002 16:02:12 +0000 (16:02 +0000)]
[project @ 2002-07-15 16:02:11 by simonmar]
Remove for extra commas in export lists (GHC is a little too lenient).
Patch from Ross Paterson.

21 years ago[project @ 2002-07-15 12:25:24 by simonmar]
simonmar [Mon, 15 Jul 2002 12:25:24 +0000 (12:25 +0000)]
[project @ 2002-07-15 12:25:24 by simonmar]
Remove duplication in the extra_libraries list for mingw32

21 years ago[project @ 2002-07-08 10:43:10 by simonmar]
simonmar [Mon, 8 Jul 2002 10:43:10 +0000 (10:43 +0000)]
[project @ 2002-07-08 10:43:10 by simonmar]
Fix a bug in getDirectoryEntries where the directory stream wasn't
always being closed.  This one shows up on Solaris as a "too many open
files" failure when trying to run the test suite.

21 years ago[project @ 2002-07-04 16:22:02 by simonmar]
simonmar [Thu, 4 Jul 2002 16:22:03 +0000 (16:22 +0000)]
[project @ 2002-07-04 16:22:02 by simonmar]
Flesh out the documentation a bit.

21 years ago[project @ 2002-07-04 13:33:24 by simonmar]
simonmar [Thu, 4 Jul 2002 13:33:25 +0000 (13:33 +0000)]
[project @ 2002-07-04 13:33:24 by simonmar]
Sigh, enabling _POSIX_PTHREAD_SEMANTICS causes the Solaris header
files to drop several silly little C functions into the source code.
We don't want these duplicated in every Haskell-compiled object.

So plan B: just define _POSIX_PTHREAD_SEMANTICS in dirUtils.c where it
is needed to make readdir_r work.  I've checked, and readdir_r is the
only function we use that is affected by _POSIX_PTHREAD_SEMANTICS.  If
we ever use any more of these functions, then we'll have to be careful
to give them a C wrapper.

21 years ago[project @ 2002-07-04 12:57:39 by simonmar]
simonmar [Thu, 4 Jul 2002 12:57:39 +0000 (12:57 +0000)]
[project @ 2002-07-04 12:57:39 by simonmar]
We have to define _POSIX_PTHREAD_SEMANTICS on Solaris in order to get
the right versions of the _r functions.  Otherwise we get
Solaris-specific versions of these, which puts a spanner in the works.

Fixes problems with getDirectoryContents on Solaris in the HEAD.

21 years ago[project @ 2002-07-04 10:42:32 by simonmar]
simonmar [Thu, 4 Jul 2002 10:42:32 +0000 (10:42 +0000)]
[project @ 2002-07-04 10:42:32 by simonmar]
Explicitly import the Prelude, and add a few types signatures to make
these modules produce better documentation.

21 years ago[project @ 2002-07-03 13:05:28 by simonmar]
simonmar [Wed, 3 Jul 2002 13:05:28 +0000 (13:05 +0000)]
[project @ 2002-07-03 13:05:28 by simonmar]
Export dynApply and dynApp

21 years ago[project @ 2002-07-02 13:13:36 by simonmar]
simonmar [Tue, 2 Jul 2002 13:13:37 +0000 (13:13 +0000)]
[project @ 2002-07-02 13:13:36 by simonmar]
Add descriptions to the top level of each package documentation

21 years ago[project @ 2002-07-02 10:34:52 by simonmar]
simonmar [Tue, 2 Jul 2002 10:34:52 +0000 (10:34 +0000)]
[project @ 2002-07-02 10:34:52 by simonmar]
Point the punters to the online report at haskell.org until the
documentation for the Prelude is up to scratch.

21 years ago[project @ 2002-07-02 10:33:23 by simonmar]
simonmar [Tue, 2 Jul 2002 10:33:23 +0000 (10:33 +0000)]
[project @ 2002-07-02 10:33:23 by simonmar]
Documentation for getProgName.

21 years ago[project @ 2002-07-02 10:31:35 by simonmar]
simonmar [Tue, 2 Jul 2002 10:31:35 +0000 (10:31 +0000)]
[project @ 2002-07-02 10:31:35 by simonmar]
Documentation for system

21 years ago[project @ 2002-07-02 10:28:54 by simonmar]
simonmar [Tue, 2 Jul 2002 10:28:54 +0000 (10:28 +0000)]
[project @ 2002-07-02 10:28:54 by simonmar]
Documentation for hSetBinaryMode

21 years ago[project @ 2002-07-01 18:50:54 by sof]
sof [Mon, 1 Jul 2002 18:50:54 +0000 (18:50 +0000)]
[project @ 2002-07-01 18:50:54 by sof]
closesocket() is stdcall-based

21 years ago[project @ 2002-07-01 11:37:35 by simonmar]
simonmar [Mon, 1 Jul 2002 11:37:35 +0000 (11:37 +0000)]
[project @ 2002-07-01 11:37:35 by simonmar]
Replace divInt# with a version which doesn't suffer from overflow
problems (thanks to Dylan Thurston for the code).

21 years ago[project @ 2002-06-27 17:19:01 by sof]
sof [Thu, 27 Jun 2002 17:19:01 +0000 (17:19 +0000)]
[project @ 2002-06-27 17:19:01 by sof]
theStdGen: initialise it via mkStdRNG (as was done in <= 5.02)

21 years ago[project @ 2002-06-27 15:38:58 by simonmar]
simonmar [Thu, 27 Jun 2002 15:38:58 +0000 (15:38 +0000)]
[project @ 2002-06-27 15:38:58 by simonmar]
Finally fix foreign export and foreign import "wrapper" so that
exceptions raised during the call are handled properly rather than
causing the RTS to bomb out.

In particular, calling System.exitWith in a foreign export will cause
the program to terminate cleanly with the desired exit code.  All
other exceptions are printed on stderr (and the program is
terminated).

Details:

GHC.TopHandler.runMain is now called runIO, and has type IO a -> IO a
(previously it had type IO a -> IO (), but that's not general enough
for a foreign export).  The stubs for foreign export and forein import
"wrapper" now automatically wrap the computation in runIO or its dual,
runNonIO.  It turned out to be simpler to do it this way than to do
the wrapping in Haskell land (plain foreign exports don't have
wrappers in Haskell).

21 years ago[project @ 2002-06-27 13:40:37 by stolz]
stolz [Thu, 27 Jun 2002 13:40:37 +0000 (13:40 +0000)]
[project @ 2002-06-27 13:40:37 by stolz]
- forkProcess[Prim] wrapper
- document forkProcess[Prim] & labelThread

21 years ago[project @ 2002-06-26 08:18:45 by stolz]
stolz [Wed, 26 Jun 2002 08:18:45 +0000 (08:18 +0000)]
[project @ 2002-06-26 08:18:45 by stolz]
- Make TSO "stable" again: The thread label was changing the size of the
   TSO if you were building a debugging-RTS, leading to binary
   incompatibility. Now we map TSOs to strings using Hash.c.

- API change for labelThread: Label arbitrary threads.

21 years ago[project @ 2002-06-24 14:40:02 by simonmar]
simonmar [Mon, 24 Jun 2002 14:40:02 +0000 (14:40 +0000)]
[project @ 2002-06-24 14:40:02 by simonmar]
Makefile updates for generating docs with Haddock:

- explicitly depend on package base
- set SRC_HADDOCK_OPTS in some places

21 years ago[project @ 2002-06-21 14:00:32 by simonmar]
simonmar [Fri, 21 Jun 2002 14:00:32 +0000 (14:00 +0000)]
[project @ 2002-06-21 14:00:32 by simonmar]
Mention the package name in the Haddock title

21 years ago[project @ 2002-06-21 09:12:37 by simonmar]
simonmar [Fri, 21 Jun 2002 09:12:37 +0000 (09:12 +0000)]
[project @ 2002-06-21 09:12:37 by simonmar]
Andy Gill asked me to remove his license to keep things simple.

21 years ago[project @ 2002-06-20 16:12:58 by simonmar]
simonmar [Thu, 20 Jun 2002 16:12:58 +0000 (16:12 +0000)]
[project @ 2002-06-20 16:12:58 by simonmar]
Haddock stuff has moved to ../../mk/package.mk

21 years ago[project @ 2002-06-20 16:11:45 by simonmar]
simonmar [Thu, 20 Jun 2002 16:11:45 +0000 (16:11 +0000)]
[project @ 2002-06-20 16:11:45 by simonmar]
use $(includedir) and INSTALL_INCLUDES rather than overriding $(datadir)

21 years ago[project @ 2002-06-20 14:56:11 by malcolm]
malcolm [Thu, 20 Jun 2002 14:56:11 +0000 (14:56 +0000)]
[project @ 2002-06-20 14:56:11 by malcolm]
Remove spurious comma in export list.

21 years ago[project @ 2002-06-20 13:54:40 by simonmar]
simonmar [Thu, 20 Jun 2002 13:54:40 +0000 (13:54 +0000)]
[project @ 2002-06-20 13:54:40 by simonmar]
Add license from Andy Gill's HTML library

21 years ago[project @ 2002-06-19 14:33:24 by simonmar]
simonmar [Wed, 19 Jun 2002 14:33:24 +0000 (14:33 +0000)]
[project @ 2002-06-19 14:33:24 by simonmar]
Remove this hack.  Happy 1.13 is now required to build GHC.

21 years ago[project @ 2002-06-18 13:58:22 by simonpj]
simonpj [Tue, 18 Jun 2002 13:58:23 +0000 (13:58 +0000)]
[project @ 2002-06-18 13:58:22 by simonpj]
---------------------------------------
    Rehash the handling of SeqOp
---------------------------------------

See the comments in the commentary (Cunning Prelude Code).

* Expunge SeqOp altogether

* Add GHC.Base.lazy :: a -> a
  to GHC.Base

* Add GHC.Base.lazy
  to basicTypes/MkId.  The idea is that this defn will over-ride
  the info from GHC.Base.hi, thereby hiding strictness and
  unfolding

* Make stranal/WorkWrap do a "manual inlining" for GHC.Base.lazy
  This happens nicely after the strictness analyser has run.

* Expunge the SeqOp/ParOp magic in CorePrep

* Expunge the RULE for seq in PrelRules

* Change the defns of pseq/par in GHC.Conc to:

{-# INLINE pseq  #-}
        pseq :: a -> b -> b
        pseq  x y = x `seq` lazy y

        {-# INLINE par  #-}
        par :: a -> b -> b
        par  x y = case (par# x) of { _ -> lazy y }

21 years ago[project @ 2002-06-18 13:01:43 by simonmar]
simonmar [Tue, 18 Jun 2002 13:01:43 +0000 (13:01 +0000)]
[project @ 2002-06-18 13:01:43 by simonmar]
Fix bug in the implementation of hGetLine: on finding the EOF when we
have a partial line in our hands, we weren't resetting the state of
the buffer to empty, so the same partial line would be returned for
each subsequent call to hGetLine.

21 years ago[project @ 2002-06-18 09:31:05 by simonpj]
simonpj [Tue, 18 Jun 2002 09:31:06 +0000 (09:31 +0000)]
[project @ 2002-06-18 09:31:05 by simonpj]
--------------------------
Deal with NaN and Infinity
--------------------------

This commit arranges that

* GHC.Real exports infinity   :: Rational = 1 :% 0
notANumber :: Rational = 0 :% 0

* GHC.Float.fromRat converts these Rational non-numbers into
  their Float/Double equivalents

As a result, arith015/arith016 start to work again.  We can
read NaN and Infinity into Float/Double.

21 years ago[project @ 2002-06-14 08:17:08 by simonpj]
simonpj [Fri, 14 Jun 2002 08:17:08 +0000 (08:17 +0000)]
[project @ 2002-06-14 08:17:08 by simonpj]
Add comment

21 years ago[project @ 2002-06-13 23:26:55 by sof]
sof [Thu, 13 Jun 2002 23:26:56 +0000 (23:26 +0000)]
[project @ 2002-06-13 23:26:55 by sof]
Provide STRefs over both lazy and strict ST monads as,

   Data.STRef.Lazy
   Data.STRef.Strict
   Data.STRef      <-- the strict version.

i.e., mirror how the strict and lazy ST monads are
organised over in Control.Monad.ST

21 years ago[project @ 2002-06-13 10:41:31 by simonmar]
simonmar [Thu, 13 Jun 2002 10:41:31 +0000 (10:41 +0000)]
[project @ 2002-06-13 10:41:31 by simonmar]
Some extra docs, from Jon Fairbairn