From A.M.Gimblett at swansea.ac.uk Wed Oct 11 17:57:26 2006 From: A.M.Gimblett at swansea.ac.uk (Andy Gimblett) Date: Wed, 11 Oct 2006 16:57:26 +0100 Subject: [Hets-devel] GHC.Prim problem Message-ID: <20061011155726.GF13319@cspcag2.swan.ac.uk> Hi all, A cry for help: I wish to import LIB_ITEM from Syntax/AS_Library.hs, but adding a line of form import Syntax.AS_Library (LIB_ITEM) to my module causes the following compilation error: hmake -Wall -fglasgow-exts -package parsec -package fgl -i/home/gimbo/work/research/mphil/tools/HetCATS ccparse MkProg: Can't find module GHC.Prim in user directories . /home/gimbo/work/research/mphil/tools/HetCATS Or in installed libraries/packages at /usr/local/lib/ghc-6.4.1/imports /usr/local/lib/ghc-6.4.1/imports /usr/local/lib/ghc-6.4.1/imports Asked for by: /home/gimbo/work/research/mphil/tools/HetCATS/Common/ATerm/AbstractSyntax.hs Fix using the -I, -P, or -package flags. This is some code I'm working on outside the hets codebase, by the way. Obviously it works fine in hets. :-) I assume I'm failing to specify some path somewhere, but having tried a few things out (mainly from looking at Makefiles in hets) and getting nowhere, I'm still a bit stumped, so I thought I'd ask for help... Finally, is this list an appropriate forum for questions of this type? If not, apologies, and where is? Many thanks, -Andy -- Andy Gimblett Computer Science Department University of Wales Swansea http://www.cs.swan.ac.uk/~csandy/ From maeder at tzi.de Wed Oct 11 19:16:51 2006 From: maeder at tzi.de (Christian Maeder) Date: Wed, 11 Oct 2006 19:16:51 +0200 Subject: [Hets-devel] GHC.Prim problem In-Reply-To: <20061011155726.GF13319@cspcag2.swan.ac.uk> References: <20061011155726.GF13319@cspcag2.swan.ac.uk> Message-ID: <452D2703.6010109@tzi.de> Andy Gimblett schrieb: > Hi all, > > A cry for help: I wish to import LIB_ITEM from Syntax/AS_Library.hs, > but adding a line of form > > import Syntax.AS_Library (LIB_ITEM) > > to my module causes the following compilation error: > > hmake -Wall -fglasgow-exts -package parsec -package fgl -i/home/gimbo/work/research/mphil/tools/HetCATS ccparse > MkProg: Can't find module GHC.Prim in user directories > . > /home/gimbo/work/research/mphil/tools/HetCATS > Or in installed libraries/packages at > /usr/local/lib/ghc-6.4.1/imports > /usr/local/lib/ghc-6.4.1/imports > /usr/local/lib/ghc-6.4.1/imports > Asked for by: /home/gimbo/work/research/mphil/tools/HetCATS/Common/ATerm/AbstractSyntax.hs > Fix using the -I, -P, or -package flags. This seems to be a problem related to hmake. GHC.Prim is in package base (maybe adding "-package base" helps). Also in our installation is no file GHC/Prim.hi under imports but only a file GHC/PrimopWrappers.hi. I suggest to use "ghc --make ..." instead. > This is some code I'm working on outside the hets codebase, by the > way. fine! If you have the whole HetCATS tree from cvs, then adding a toplevel file SomeWhere/myfile.hs should be easy. "make SomeWhere/myfile" should create a corresponding binary. > Finally, is this list an appropriate forum for questions of this type? yes. > Many thanks, > > -Andy You're welcome Christian From maeder at tzi.de Thu Oct 12 13:55:08 2006 From: maeder at tzi.de (Christian Maeder) Date: Thu, 12 Oct 2006 13:55:08 +0200 Subject: [Hets-devel] programatica and ghc-6.6 Message-ID: <452E2D1C.6040902@tzi.de> Hi, I will remove "-package data" from our Makefile. If you get a corresponding error (usually only after "make dist_clean") you'll have to replace "FiniteMap" with "Data.FiniteMap" in: ../programatica/tools/base/parse2/LexerGen/CompileRegExp.hs If you switch to ghc-6.6, you'll have to get uni and programatica (with today's date) anew from http://www.informatik.uni-bremen.de/agbkb/forschung/formal_methods/CoFI/hets/src-distribution/versions/ Cheers Christian From A.M.Gimblett at swansea.ac.uk Fri Oct 13 12:55:16 2006 From: A.M.Gimblett at swansea.ac.uk (Andy Gimblett) Date: Fri, 13 Oct 2006 11:55:16 +0100 Subject: [Hets-devel] GHC.Prim problem In-Reply-To: <452D2703.6010109@tzi.de> References: <20061011155726.GF13319@cspcag2.swan.ac.uk> <452D2703.6010109@tzi.de> Message-ID: <20061013105516.GB1270@cspcag2.swan.ac.uk> On Wed, Oct 11, 2006 at 07:16:51PM +0200, Christian Maeder wrote: > > This seems to be a problem related to hmake. GHC.Prim is in package > base (maybe adding "-package base" helps). That doesn't help. > Also in our installation is no file GHC/Prim.hi under imports but > only a file GHC/PrimopWrappers.hi. > > I suggest to use "ghc --make ..." instead. That does help. :-) Perfect - many thanks! > > This is some code I'm working on outside the hets codebase, by the > > way. > > fine! If you have the whole HetCATS tree from cvs, then adding a > toplevel file SomeWhere/myfile.hs should be easy. "make > SomeWhere/myfile" should create a corresponding binary. That's impressive. Alas, it fails in this case because of other modules in that folder which ghc can't then find. However, this is no big issue, because I will shortly fold this code into the HetCATS CVS tree. Thanks again for your help! -Andy -- Andy Gimblett Computer Science Department University of Wales Swansea http://www.cs.swan.ac.uk/~csandy/ From maeder at tzi.de Wed Oct 25 10:36:31 2006 From: maeder at tzi.de (Christian Maeder) Date: Wed, 25 Oct 2006 10:36:31 +0200 Subject: [Hets-devel] strange blank lines Message-ID: <453F220F.2010905@tzi.de> Does anybody have an explanation for the blank lines being printed when requesting output (with -o th or -o thy)? It must be change within the last two days (possibly by myself) Thanks Christian maeder at turing:~/casl/CASL-lib> hets -v2 -o thy Basic/Numbers.casl Reading file Basic/Numbers.casl Analyzing library Basic/Numbers version 1.0 Analyzing spec Nat Analyzing spec Int Analyzing spec Rat Analyzing spec DecimalFraction Writing file: Basic/Numbers_DecimalFraction.thy Writing file: Basic/Numbers_Int.thy Writing file: Basic/Numbers_Nat.thy Writing file: Basic/Numbers_Rat.thy From maeder at tzi.de Wed Oct 25 11:05:09 2006 From: maeder at tzi.de (Christian Maeder) Date: Wed, 25 Oct 2006 11:05:09 +0200 Subject: [Hets-devel] strange blank lines In-Reply-To: <453F220F.2010905@tzi.de> References: <453F220F.2010905@tzi.de> Message-ID: <453F28C5.9060707@tzi.de> Christian Maeder schrieb: > Does anybody have an explanation for the blank lines being printed when > requesting output (with -o th or -o thy)? > > It must be change within the last two days (possibly by myself) fixed now, it was an empty trace message! (not to be checked in) Christian From A.M.Gimblett at swansea.ac.uk Fri Oct 27 15:49:23 2006 From: A.M.Gimblett at swansea.ac.uk (Andy Gimblett) Date: Fri, 27 Oct 2006 14:49:23 +0100 Subject: [Hets-devel] Emacs question: ignoring patterns in find-file? Message-ID: <20061027134923.GA10772@cspcag2.swan.ac.uk> Someone out there must know the answer to this... I never, ever, want to open up a ".hi" file in emacs. However, when opening a ".hs" file, tab-completing in the minibuffer typically takes me as far as ".h" and then I have to type the "s" to get the Haskell source. If I forget and just hit returm, emacs helpfully opens the ".h" version of the file, which of course doesn't exist, and populates it with my C header file template. This happens to me frequently enough to be annoying, and it seems to me that there might (must!) be a way in emacs to say "always ignore filenames matching the following pattern". Does anyone know it? Thanks! -Andy -- Andy Gimblett Computer Science Department University of Wales Swansea http://www.cs.swan.ac.uk/~csandy/ From till at tzi.de Fri Oct 27 16:33:32 2006 From: till at tzi.de (Till Mossakowski) Date: Fri, 27 Oct 2006 16:33:32 +0200 Subject: [Hets-devel] Emacs question: ignoring patterns in find-file? In-Reply-To: <20061027134923.GA10772@cspcag2.swan.ac.uk> References: <20061027134923.GA10772@cspcag2.swan.ac.uk> Message-ID: <454218BC.3080604@tzi.de> Good question! I have the same problem, but do not know the answer either... aynone? Greetings, Till Andy Gimblett schrieb: > Someone out there must know the answer to this... > > I never, ever, want to open up a ".hi" file in emacs. However, when > opening a ".hs" file, tab-completing in the minibuffer typically takes > me as far as ".h" and then I have to type the "s" to get the Haskell > source. If I forget and just hit returm, emacs helpfully opens the > ".h" version of the file, which of course doesn't exist, and populates > it with my C header file template. > > This happens to me frequently enough to be annoying, and it seems to > me that there might (must!) be a way in emacs to say "always ignore > filenames matching the following pattern". > > Does anyone know it? > > Thanks! > > -Andy > From A.M.Gimblett at swansea.ac.uk Sun Oct 29 20:42:29 2006 From: A.M.Gimblett at swansea.ac.uk (Andy Gimblett) Date: Sun, 29 Oct 2006 19:42:29 +0000 Subject: [Hets-devel] Emacs question: ignoring patterns in find-file? In-Reply-To: <454218BC.3080604@tzi.de> References: <20061027134923.GA10772@cspcag2.swan.ac.uk> <454218BC.3080604@tzi.de> Message-ID: <20061029194229.GA26555@cspcag2.swan.ac.uk> On Fri, Oct 27, 2006 at 04:33:32PM +0200, Till Mossakowski wrote: > I have the same problem, but do not know the answer > either... aynone? I have it! The variable "completion-ignored-extensions" is what we're after. http://www.gnu.org/software/emacs/manual/html_node/Completion-Options.html If you like to use emacs' built-in customisation interface, you'll find it under Environment->Dired. Cheers! -Andy -- Andy Gimblett Computer Science Department University of Wales Swansea http://www.cs.swan.ac.uk/~csandy/