From rick at rickmurphy.org Sun Jan 17 18:03:07 2010 From: rick at rickmurphy.org (Rick Murphy) Date: Sun, 17 Jan 2010 12:03:07 -0500 Subject: [Hets-users] install and config Message-ID: <4B5342CB.7040803@rickmurphy.org> The default network installer wasn't able to find Isabelle at the specified location. I elected not to install Isabelle because I already have an Isabelle2009 instance on my x86 opensolaris box. As expected the prover dialog now indicates no prover is running. I would like to point Hets at the Isabelle2009 instance already on my box. Is the location of Isabelle configurable in Hets? If so could you please indicate where I can configure its location? Best regards and thanks for such great work ! -- Rick cell: 703-201-9129 web: http://www.rickmurphy.org blog: http://phaneron.rickmurphy.org From Christian.Maeder at dfki.de Mon Jan 18 12:51:19 2010 From: Christian.Maeder at dfki.de (Christian Maeder) Date: Mon, 18 Jan 2010 12:51:19 +0100 Subject: [Hets-users] install and config In-Reply-To: <4B5342CB.7040803@rickmurphy.org> References: <4B5342CB.7040803@rickmurphy.org> Message-ID: <4B544B37.4070100@dfki.de> In order to use Isabelle 2009 set the environment variable HETS_ISABELLE to the isabelle executable plus the argument "emacs" so that the emacs interface pops up. I.e. HETS_ISABELLE='/usr/local/Isabelle2009/bin/isabelle emacs' Furthermore the included file Hets-lib/Isabelle/prelude.ML must replaced (overwritten) by the file Hets-lib/Isabelle/prelude2009.ML I assume you've installed hets-0.93-installer-pc-solaris.jar Feel free to ask further question via this list or directly to me. Cheers Christian Rick Murphy schrieb: > The default network installer wasn't able to find Isabelle at the > specified location. I elected not to install Isabelle because I already > have an Isabelle2009 instance on my x86 opensolaris box. > > As expected the prover dialog now indicates no prover is running. > > I would like to point Hets at the Isabelle2009 instance already on my > box. Is the location of Isabelle configurable in Hets? If so could you > please indicate where I can configure its location? > > Best regards and thanks for such great work ! > From Christian.Maeder at dfki.de Mon Jan 18 15:10:05 2010 From: Christian.Maeder at dfki.de (Christian Maeder) Date: Mon, 18 Jan 2010 15:10:05 +0100 Subject: [Hets-users] install and config In-Reply-To: <4B5342CB.7040803@rickmurphy.org> References: <4B5342CB.7040803@rickmurphy.org> Message-ID: <4B546BBD.7040705@dfki.de> Rick Murphy schrieb: > The default network installer wasn't able to find Isabelle at the > specified location. I elected not to install Isabelle because I already > have an Isabelle2009 instance on my x86 opensolaris box. Under Solaris there may be a problem with the "patch" program, that is used to insert proofs into .thy files (i.e. Hets-lib/RCCVerification_RCC_FO_in_MetricSpace_T.thy) The program "patch" should be the GNU version "gpatch" (possibly via a soft link). Christian From rick at rickmurphy.org Mon Jan 18 15:43:31 2010 From: rick at rickmurphy.org (Rick Murphy) Date: Mon, 18 Jan 2010 09:43:31 -0500 Subject: [Hets-users] install and config In-Reply-To: <4B544B37.4070100@dfki.de> References: <4B5342CB.7040803@rickmurphy.org> <4B544B37.4070100@dfki.de> Message-ID: <4B547393.8010808@rickmurphy.org> Excellent and thanks again for such a great project! Questions below ... Christian Maeder wrote: > In order to use Isabelle 2009 set the environment variable > HETS_ISABELLE to the isabelle executable plus the argument "emacs" so > that the emacs interface pops up. I.e. > > HETS_ISABELLE='/usr/local/Isabelle2009/bin/isabelle emacs' > > Furthermore the included file Hets-lib/Isabelle/prelude.ML must replaced > (overwritten) by the file Hets-lib/Isabelle/prelude2009.ML Do you mean simply rename the existing prelude.ML as prelude2009.ML? Assuming that's the case and using the Propositional/Birds example. If I examine the generated Birds_Penguin.thy it appears the "uses" line also needs to be updated to prelude2009.ML. Once I did that emacs and proof general loads. I can start Isabelle, but when I execute "process the next proof command" emacs return the following response: *** Error: Value or constructor (thms_of_proof) has not been declared Found near *** List.partition(fn s => mem( ...))( *** filter(fn s => ......)( *** filter(fn ... => ...)(filter(...(...))(Symtab.keys(...(...)))))) (line 23 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") *** Error: Value or constructor (the_context) has not been declared Found near *** the_context(()) (line 35 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") *** Error: Value or constructor (thms_of) has not been declared in structure PureThy Found near *** PureThy.thms_of(thy) (line 36 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") *** Error: Value or constructor (the_context) has not been declared Found near *** Context.theory_name(the_context(())) (line 52 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") *** Exception- Fail "Static errors (pass2)" raised *** At command "theory". > I assume you've installed hets-0.93-installer-pc-solaris.jar Yes, but I ran the daily update yesterday so I'm running 0.94. > Feel free to ask further question via this list or directly to me. > > Cheers Christian > > > Rick Murphy schrieb: >> The default network installer wasn't able to find Isabelle at the >> specified location. I elected not to install Isabelle because I already >> have an Isabelle2009 instance on my x86 opensolaris box. >> >> As expected the prover dialog now indicates no prover is running. >> >> I would like to point Hets at the Isabelle2009 instance already on my >> box. Is the location of Isabelle configurable in Hets? If so could you >> please indicate where I can configure its location? >> >> Best regards and thanks for such great work ! >> > > From rick at rickmurphy.org Mon Jan 18 15:47:41 2010 From: rick at rickmurphy.org (Rick Murphy) Date: Mon, 18 Jan 2010 09:47:41 -0500 Subject: [Hets-users] install and config Message-ID: <4B54748D.1050406@rickmurphy.org> Excellent and thanks again for such a great project! Questions below ... Christian Maeder wrote: > In order to use Isabelle 2009 set the environment variable > HETS_ISABELLE to the isabelle executable plus the argument "emacs" so > that the emacs interface pops up. I.e. > > HETS_ISABELLE='/usr/local/Isabelle2009/bin/isabelle emacs' > > Furthermore the included file Hets-lib/Isabelle/prelude.ML must replaced > (overwritten) by the file Hets-lib/Isabelle/prelude2009.ML Do you mean simply rename the existing prelude.ML as prelude2009.ML? Assuming that's the case and using the Propositional/Birds example. If I examine the generated Birds_Penguin.thy it appears the "uses" line also needs to be updated to prelude2009.ML. Once I did that emacs and proof general loads. I can start Isabelle, but when I execute "process the next proof command" emacs return the following response: *** Error: Value or constructor (thms_of_proof) has not been declared Found near *** List.partition(fn s => mem( ...))( *** filter(fn s => ......)( *** filter(fn ... => ...)(filter(...(...))(Symtab.keys(...(...)))))) (line 23 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") *** Error: Value or constructor (the_context) has not been declared Found near *** the_context(()) (line 35 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") *** Error: Value or constructor (thms_of) has not been declared in structure PureThy Found near *** PureThy.thms_of(thy) (line 36 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") *** Error: Value or constructor (the_context) has not been declared Found near *** Context.theory_name(the_context(())) (line 52 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") *** Exception- Fail "Static errors (pass2)" raised *** At command "theory". > I assume you've installed hets-0.93-installer-pc-solaris.jar Yes, but I ran the daily update yesterday so I'm running 0.94. > Feel free to ask further question via this list or directly to me. > > Cheers Christian > > > Rick Murphy schrieb: >> The default network installer wasn't able to find Isabelle at the >> specified location. I elected not to install Isabelle because I already >> have an Isabelle2009 instance on my x86 opensolaris box. >> >> As expected the prover dialog now indicates no prover is running. >> >> I would like to point Hets at the Isabelle2009 instance already on my >> box. Is the location of Isabelle configurable in Hets? If so could you >> please indicate where I can configure its location? >> >> Best regards and thanks for such great work ! >> > > -- Rick cell: 703-201-9129 web: http://www.rickmurphy.org blog: http://phaneron.rickmurphy.org From rick at rickmurphy.org Mon Jan 18 15:48:19 2010 From: rick at rickmurphy.org (Rick Murphy) Date: Mon, 18 Jan 2010 09:48:19 -0500 Subject: [Hets-users] install and config In-Reply-To: <4B547393.8010808@rickmurphy.org> References: <4B5342CB.7040803@rickmurphy.org> <4B544B37.4070100@dfki.de> <4B547393.8010808@rickmurphy.org> Message-ID: <4B5474B3.3050702@rickmurphy.org> BTW - the terminal also displays the following: "illegal change of theory header" -- Rick Rick Murphy wrote: > Excellent and thanks again for such a great project! Questions below ... > > Christian Maeder wrote: >> In order to use Isabelle 2009 set the environment variable >> HETS_ISABELLE to the isabelle executable plus the argument "emacs" so >> that the emacs interface pops up. I.e. >> >> HETS_ISABELLE='/usr/local/Isabelle2009/bin/isabelle emacs' >> >> Furthermore the included file Hets-lib/Isabelle/prelude.ML must replaced >> (overwritten) by the file Hets-lib/Isabelle/prelude2009.ML > > Do you mean simply rename the existing prelude.ML as prelude2009.ML? > > Assuming that's the case and using the Propositional/Birds example. If I > examine the generated Birds_Penguin.thy it appears the "uses" line also > needs to be updated to prelude2009.ML. Once I did that emacs and proof > general loads. I can start Isabelle, but when I execute "process the > next proof command" emacs return the following response: > > *** Error: Value or constructor (thms_of_proof) has not been declared > Found near > *** List.partition(fn s => mem( ...))( > *** filter(fn s => ......)( > *** filter(fn ... => ...)(filter(...(...))(Symtab.keys(...(...)))))) > (line 23 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") > *** Error: Value or constructor (the_context) has not been declared > Found near > *** the_context(()) (line 35 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") > *** Error: Value or constructor (thms_of) has not been declared in > structure PureThy Found near > *** PureThy.thms_of(thy) (line 36 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") > *** Error: Value or constructor (the_context) has not been declared > Found near > *** Context.theory_name(the_context(())) (line 52 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") > *** Exception- Fail "Static errors (pass2)" raised > *** At command "theory". > >> I assume you've installed hets-0.93-installer-pc-solaris.jar > > Yes, but I ran the daily update yesterday so I'm running 0.94. > >> Feel free to ask further question via this list or directly to me. >> >> Cheers Christian >> >> >> Rick Murphy schrieb: >>> The default network installer wasn't able to find Isabelle at the >>> specified location. I elected not to install Isabelle because I already >>> have an Isabelle2009 instance on my x86 opensolaris box. >>> >>> As expected the prover dialog now indicates no prover is running. >>> >>> I would like to point Hets at the Isabelle2009 instance already on my >>> box. Is the location of Isabelle configurable in Hets? If so could you >>> please indicate where I can configure its location? >>> >>> Best regards and thanks for such great work ! >>> >> >> > From Christian.Maeder at dfki.de Mon Jan 18 16:26:57 2010 From: Christian.Maeder at dfki.de (Christian Maeder) Date: Mon, 18 Jan 2010 16:26:57 +0100 Subject: [Hets-users] install and config In-Reply-To: <4B54748D.1050406@rickmurphy.org> References: <4B54748D.1050406@rickmurphy.org> Message-ID: <4B547DC1.6000903@dfki.de> Rick Murphy schrieb: >> Furthermore the included file Hets-lib/Isabelle/prelude.ML must replaced >> (overwritten) by the file Hets-lib/Isabelle/prelude2009.ML > > Do you mean simply rename the existing prelude.ML as prelude2009.ML? No, the other way round. Remove prelude.ML and rename prelude2009.ML to prelude.ML > Assuming that's the case and using the Propositional/Birds example. If I > examine the generated Birds_Penguin.thy it appears the "uses" line also > needs to be updated to prelude2009.ML. Once I did that emacs and proof This line should not be changed! "illegal change of theory header" should not appear then. > general loads. I can start Isabelle, but when I execute "process the > next proof command" emacs return the following response: > > *** Error: Value or constructor (thms_of_proof) has not been declared > Found near That's the old line from prelude.ML that only worked with Isabelle2008. > *** List.partition(fn s => mem( ...))( > *** filter(fn s => ......)( > *** filter(fn ... => ...)(filter(...(...))(Symtab.keys(...(...)))))) > (line 23 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") > *** Error: Value or constructor (the_context) has not been declared > Found near > *** the_context(()) (line 35 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") > *** Error: Value or constructor (thms_of) has not been declared in > structure PureThy Found near > *** PureThy.thms_of(thy) (line 36 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") > *** Error: Value or constructor (the_context) has not been declared > Found near > *** Context.theory_name(the_context(())) (line 52 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") > *** Exception- Fail "Static errors (pass2)" raised > *** At command "theory". > >> I assume you've installed hets-0.93-installer-pc-solaris.jar > > Yes, but I ran the daily update yesterday so I'm running 0.94. Ah, this version ("hets --version") generates lines: imports "$HETS_ISABELLE_LIB/MainHC" uses "$HETS_ISABELLE_LIB/prelude" in .thy-files, so HETS_ISABELLE_LIB must be set to /usr/local/Hets/Hets-lib/Isabelle. export HETS_ISABELLE_LIB=/usr/local/Hets/Hets-lib/Isabelle For the new version you may also want to update Hets-lib for our RCC-example: svn co https://svn-agbkb.informatik.uni-bremen.de/Hets-lib/trunk \ Hets-lib Christian From rick at rickmurphy.org Tue Jan 19 03:19:37 2010 From: rick at rickmurphy.org (Rick Murphy) Date: Mon, 18 Jan 2010 21:19:37 -0500 Subject: [Hets-users] install and config In-Reply-To: <4B547DC1.6000903@dfki.de> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> Message-ID: <4B5516B9.7070506@rickmurphy.org> Thank you for your kind support. I have a few questions below. Christian Maeder wrote: > Rick Murphy schrieb: >>> Furthermore the included file Hets-lib/Isabelle/prelude.ML must replaced >>> (overwritten) by the file Hets-lib/Isabelle/prelude2009.ML >> Do you mean simply rename the existing prelude.ML as prelude2009.ML? > > No, the other way round. Remove prelude.ML and rename prelude2009.ML to > prelude.ML Ok, I co'd version 1352 of Hets-lib from svn, removed prelude.ML then renamed prelude2009.ML to prelude.ML. >> Assuming that's the case and using the Propositional/Birds example. If I >> examine the generated Birds_Penguin.thy it appears the "uses" line also >> needs to be updated to prelude2009.ML. Once I did that emacs and proof > > This line should not be changed! > "illegal change of theory header" should not appear then. Understood. >> general loads. I can start Isabelle, but when I execute "process the >> next proof command" emacs return the following response: >> >> *** Error: Value or constructor (thms_of_proof) has not been declared >> Found near > > That's the old line from prelude.ML that only worked with Isabelle2008. Understood. >> *** List.partition(fn s => mem( ...))( >> *** filter(fn s => ......)( >> *** filter(fn ... => ...)(filter(...(...))(Symtab.keys(...(...)))))) >> (line 23 of "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") >> *** Error: Value or constructor (the_context) has not been declared >> Found near >> *** the_context(()) (line 35 of >> "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") >> *** Error: Value or constructor (thms_of) has not been declared in >> structure PureThy Found near >> *** PureThy.thms_of(thy) (line 36 of >> "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") >> *** Error: Value or constructor (the_context) has not been declared >> Found near >> *** Context.theory_name(the_context(())) (line 52 of >> "/usr/local/Hets/Hets-lib/Isabelle/prelude2009.ML") >> *** Exception- Fail "Static errors (pass2)" raised >> *** At command "theory". >> >>> I assume you've installed hets-0.93-installer-pc-solaris.jar >> Yes, but I ran the daily update yesterday so I'm running 0.94. > > Ah, this version ("hets --version") generates lines: > > imports "$HETS_ISABELLE_LIB/MainHC" > uses "$HETS_ISABELLE_LIB/prelude" I see the RCCVerification file in the root of Hets-lib has these lines. > in .thy-files, so HETS_ISABELLE_LIB must be set to > /usr/local/Hets/Hets-lib/Isabelle. > > export HETS_ISABELLE_LIB=/usr/local/Hets/Hets-lib/Isabelle I have added this, but FWIW I use tcsh ... > For the new version you may also want to update Hets-lib for our > RCC-example: AS above I've co'd 1352, but I have not been able to validate that translations are creating the imports and uses lines that you have specified. > imports "$HETS_ISABELLE_LIB/MainHC" > uses "$HETS_ISABELLE_LIB/prelude" For example, if I execute the following command ... bin/hets -v3 -g Hets-lib/Calculi/Space/SimpleRCC.het then the CASL2Isabelle translation produces the following output ... theory SimpleRCC_ExtRCC_FO imports Main uses "$HETS_LIB/Isabelle/prelude" begin ML "Header.initialize [\"C_reflex\", \"C_sym\", \"C_id\", \"P_def\", \"P_impl_C\"]" typedecl Region consts X__C__X :: "Region => Region => bool" ("(_/ C/ _)" [44,44] 42) X__P__X :: "Region => Region => bool" ("(_/ P/ _)" [44,44] 42) axioms C_reflex [rule_format] : "ALL x. x C x" C_sym [rule_format] : "ALL x. ALL y. x C y --> y C x" C_id [rule_format] : "ALL x. ALL y. (ALL z. (z C x) = (z C y)) --> x = y" P_def [rule_format] : "ALL x. ALL y. (ALL z. z C x --> z C y) = (x P y)" declare C_reflex [simp] declare C_sym [simp] theorem P_impl_C : "ALL x. ALL y. x P y --> x C y" using P_def by (auto) ML "Header.record \"P_impl_C\"" end So it appears the imports and uses statements are not as expected, correct? I can start Isabelle in ProofGeneral, but when I execute "process next proof command" the emacs response is ... *** Error: Value or constructor (thms_of_proof) has not been declared Found near *** List.partition(fn s => mem( ...))( *** filter(fn s => ......)( *** filter(fn ... => ...)(filter(...(...))(Symtab.keys(...(...)))))) (line 23 of "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") *** Error: Value or constructor (the_context) has not been declared Found near *** the_context(()) (line 35 of "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") *** Error: Value or constructor (thms_of) has not been declared in structure PureThy Found near *** PureThy.thms_of(thy) (line 36 of "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") *** Error: Value or constructor (the_context) has not been declared Found near *** Context.theory_name(the_context(())) (line 52 of "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") *** Exception- Fail "Static errors (pass2)" raised *** At command "theory". > svn co https://svn-agbkb.informatik.uni-bremen.de/Hets-lib/trunk \ > Hets-lib As above, I now have 1352. BTW - Thanks for the other help. That's working great. > Christian > > > From Christian.Maeder at dfki.de Tue Jan 19 13:40:58 2010 From: Christian.Maeder at dfki.de (Christian Maeder) Date: Tue, 19 Jan 2010 13:40:58 +0100 Subject: [Hets-users] install and config In-Reply-To: <4B5516B9.7070506@rickmurphy.org> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> <4B5516B9.7070506@rickmurphy.org> Message-ID: <4B55A85A.8090009@dfki.de> Rick Murphy schrieb: >> Ah, this version ("hets --version") generates lines: >> >> imports "$HETS_ISABELLE_LIB/MainHC" >> uses "$HETS_ISABELLE_LIB/prelude" HETS_ISABELLE_LIB is used in hets-0.94 after Sept. 1st 2009 So you'll properly have hets-0.94 from 2009-04-28. This version should work fine for you with HETS_LIB set, only. Getting a daily snapshot did not work as expected. I'll explain below how to fix this. > For example, if I execute the following command ... > > bin/hets -v3 -g Hets-lib/Calculi/Space/SimpleRCC.het > > then the CASL2Isabelle translation produces the following output ... > > theory SimpleRCC_ExtRCC_FO > imports Main > uses "$HETS_LIB/Isabelle/prelude" > begin > > ML "Header.initialize > [\"C_reflex\", \"C_sym\", \"C_id\", \"P_def\", \"P_impl_C\"]" > > typedecl Region > > consts > X__C__X :: "Region => Region => bool" ("(_/ C/ _)" [44,44] 42) > X__P__X :: "Region => Region => bool" ("(_/ P/ _)" [44,44] 42) > > axioms > C_reflex [rule_format] : "ALL x. x C x" > > C_sym [rule_format] : "ALL x. ALL y. x C y --> y C x" > > C_id [rule_format] : > "ALL x. ALL y. (ALL z. (z C x) = (z C y)) --> x = y" > > P_def [rule_format] : > "ALL x. ALL y. (ALL z. z C x --> z C y) = (x P y)" > > declare C_reflex [simp] > declare C_sym [simp] > > theorem P_impl_C : "ALL x. ALL y. x P y --> x C y" > using P_def by (auto) > > ML "Header.record \"P_impl_C\"" > > end This file is fine for your hets-0.94 version. It should only fail for the generated proof at "auto". > So it appears the imports and uses statements are not as expected, > correct? I can start Isabelle in ProofGeneral, but when I execute > "process next proof command" the emacs response is ... The imports and uses statements are fine, too, for your hets version. The following error is still caused by the Isabelle/prelude.ML not being the version for Isabelle 2009! > *** Error: Value or constructor (thms_of_proof) has not been declared > Found near > *** List.partition(fn s => mem( ...))( > *** filter(fn s => ......)( > *** filter(fn ... => ...)(filter(...(...))(Symtab.keys(...(...)))))) > (line 23 of "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") > *** Error: Value or constructor (the_context) has not been declared > Found near > *** the_context(()) (line 35 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") > *** Error: Value or constructor (thms_of) has not been declared in > structure PureThy Found near > *** PureThy.thms_of(thy) (line 36 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") > *** Error: Value or constructor (the_context) has not been declared > Found near > *** Context.theory_name(the_context(())) (line 52 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") > *** Exception- Fail "Static errors (pass2)" raised > *** At command "theory". > >> svn co https://svn-agbkb.informatik.uni-bremen.de/Hets-lib/trunk \ >> Hets-lib > > As above, I now have 1352. Fine, but you still have to copy Hets-lib/Isabelle/prelude2009.ML to Hets-lib/Isabelle/prelude.ML in order to get you going. How to use a daily snapshot properly? The file /usr/local/Hets/bin/hets is a (bash) script that must set the variable HETS to the actual hets binary that was downloaded. This line (about line number 51) must be adjusted manually with an editor currently. The original line should like: HETS=/usr/local/Hets/lib/hets/hets-2009-04-28 You can also call the downloaded binary directly, but the script also sets some environment variables, like i.e. HETS_LIB. (In hets-0.94 versions after Sept. 1st 2009 HETS_LIB may be a colon-separated list of directories, therefore we had to introduce the new HETS_ISABELLE_LIB environment variable.) You may download the latest x86 solaris hets version from http://www.dfki.de/sks/hets/pc-solaris/versions/ And yes, we should make a new release (maybe next month). I hope I could help you Christian P.S. Sorry for assuming bash (tcsh uses setenv to set environment variables and no equal-sign) From rick at rickmurphy.org Wed Jan 20 04:59:02 2010 From: rick at rickmurphy.org (Rick Murphy) Date: Tue, 19 Jan 2010 22:59:02 -0500 Subject: [Hets-users] install and config In-Reply-To: <4B55A85A.8090009@dfki.de> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> <4B5516B9.7070506@rickmurphy.org> <4B55A85A.8090009@dfki.de> Message-ID: <4B567F86.50608@rickmurphy.org> Thank you for your kind help. I have made some progress, but not quite there just yet. See below ... Christian Maeder wrote: > This file is fine for your hets-0.94 version. It should only fail for > the generated proof at "auto". By auto do you mean that I should de-select Isabelle, Settings, AutoSolve? >> As above, I now have 1352. > > Fine, but you still have to copy Hets-lib/Isabelle/prelude2009.ML to > Hets-lib/Isabelle/prelude.ML in order to get you going. I believe I have now corrected this. Sorry for the mistake. > How to use a daily snapshot properly? The file /usr/local/Hets/bin/hets > is a (bash) script that must set the variable HETS to the actual hets > binary that was downloaded. This line (about line number 51) must be > adjusted manually with an editor currently. The original line should like: > > HETS=/usr/local/Hets/lib/hets/hets-2009-04-28 > > You can also call the downloaded binary directly, but the script also > sets some environment variables, like i.e. HETS_LIB. Ah, I did not know this. Thank you. I have now updated the bash script with the HETS and HETS_ISABELLE_LIB environment variables, as well as downloaded hets-2010-01-18. > (In hets-0.94 versions after Sept. 1st 2009 HETS_LIB may be ayes > colon-separated list of directories, therefore we had to introduce the > new HETS_ISABELLE_LIB environment variable.) Understood. The file produced by the transformation is now as follows: theory SimpleRCC_ExtRCC_FO imports Main uses "$HETS_ISABELLE_LIB/prelude" begin ML "Header.initialize [\"C_reflex\", \"C_sym\", \"C_id\", \"P_def\", \"P_impl_C\"]" typedecl Region consts X__C__X :: "Region => Region => bool" ("(_/ C/ _)" [44,44] 42) X__P__X :: "Region => Region => bool" ("(_/ P/ _)" [44,44] 42) axioms C_reflex [rule_format] : "ALL x. x C x" C_sym [rule_format] : "ALL x. ALL y. x C y --> y C x" C_id [rule_format] : "ALL x. ALL y. (ALL z. (z C x) = (z C y)) --> x = y" P_def [rule_format] : "ALL x. ALL y. (ALL z. z C x --> z C y) = (x P y)" declare C_reflex [simp] declare C_sym [simp] theorem P_impl_C : "ALL x. ALL y. x P y --> x C y" using P_def by (auto) ML "Header.record \"P_impl_C\"" end On process next proof command, Isabelle still produces the following response. *** Error: Value or constructor (the_context) has not been declared Found near *** the_context(()) (line 10 of "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") *** Error: Value or constructor (the_context) has not been declared Found near *** the_context(()) (line 30 of "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") *** Exception- Fail "Static errors (pass2)" raised *** At command "theory". > You may download the latest x86 solaris hets version from > http://www.dfki.de/sks/hets/pc-solaris/versions/ Thank you. As above I have now completed this step as well. > And yes, we should make a new release (maybe next month). I look forward to it ! Thanks again for your patience. > I hope I could help you > Christian > > P.S. Sorry for assuming bash (tcsh uses setenv to set environment > variables and no equal-sign) > > > From Christian.Maeder at dfki.de Wed Jan 20 11:31:45 2010 From: Christian.Maeder at dfki.de (Christian Maeder) Date: Wed, 20 Jan 2010 11:31:45 +0100 Subject: [Hets-users] install and config In-Reply-To: <4B567F86.50608@rickmurphy.org> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> <4B5516B9.7070506@rickmurphy.org> <4B55A85A.8090009@dfki.de> <4B567F86.50608@rickmurphy.org> Message-ID: <4B56DB91.4030005@dfki.de> Rick Murphy schrieb: > Thank you for your kind help. I have made some progress, but not quite > there just yet. See below ... > > Christian Maeder wrote: >> This file is fine for your hets-0.94 version. It should only fail for >> the generated proof at "auto". > > By auto do you mean that I should de-select Isabelle, Settings, AutoSolve? no, you have to replace the generated proof (in the file below) by a valid proof. That is the main and usually difficult prove-task left to the user. (Well, you can prove this node using SPASS more easily.) [...] > The file produced by the transformation is now as follows: the file is correct! > theory SimpleRCC_ExtRCC_FO > imports Main > uses "$HETS_ISABELLE_LIB/prelude" > begin > > ML "Header.initialize > [\"C_reflex\", \"C_sym\", \"C_id\", \"P_def\", \"P_impl_C\"]" [...] > theorem P_impl_C : "ALL x. ALL y. x P y --> x C y" you have to prove this manually! > using P_def by (auto) this line is just a generated wild guess of a proof that works in a few cases. > ML "Header.record \"P_impl_C\"" > > end > > On process next proof command, Isabelle still produces the following > response. You seem to have an Isabelle 2009 problem with logic HOL. Do you have a HOL heap in your Isabelle2009 installation (i.e. under /usr/local/Isabelle/heaps/polyml*)? If not this logic needs to be compiled using the Isabelle/build script. (So far, I've only done this for Isabelle2008 under solaris, but you may wait until I've done so for Isabelle2009, too) > *** Error: Value or constructor (the_context) has not been declared > Found near > *** the_context(()) (line 10 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") > *** Error: Value or constructor (the_context) has not been declared > Found near > *** the_context(()) (line 30 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") > *** Exception- Fail "Static errors (pass2)" raised > *** At command "theory". The file prelude.ML is correct and it works for me, if I call "isabelle-process" and enter (at "ML>" prompt): ML> use "$HETS_ISABELLE_LIB/prelude"; structure Header : sig val initialize : string list -> unit val record : string -> unit end val it = () : unit val it = () : unit (Ctrl-D to exit) I'll let you know as soon as I've a working Isabelle2009 installation under solaris. Cheers Christian From Christian.Maeder at dfki.de Wed Jan 20 13:32:38 2010 From: Christian.Maeder at dfki.de (Christian Maeder) Date: Wed, 20 Jan 2010 13:32:38 +0100 Subject: [Hets-users] install and config In-Reply-To: <4B567F86.50608@rickmurphy.org> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> <4B5516B9.7070506@rickmurphy.org> <4B55A85A.8090009@dfki.de> <4B567F86.50608@rickmurphy.org> Message-ID: <4B56F7E6.6050207@dfki.de> Rick Murphy schrieb: > On process next proof command, Isabelle still produces the following > response. > > *** Error: Value or constructor (the_context) has not been declared > Found near > *** the_context(()) (line 10 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") > *** Error: Value or constructor (the_context) has not been declared > Found near > *** the_context(()) (line 30 of > "/usr/local/Hets/Hets-lib/Isabelle/prelude.ML") > *** Exception- Fail "Static errors (pass2)" raised > *** At command "theory". It seems, you've Isabelle2009-1, where I've only considered Isabelle2009. The NEWS say: * Legacy goal package: reduced interface to the bare minimum required to keep existing proof scripts running. Most other user-level functions are now part of the OldGoals structure, which is *not* open by default (consider isatool expandshort before open OldGoals). Removed top_sg, prin, printyp, pprint_term/typ altogether, because these tend to cause confusion about the actual goal (!) context being used here, which is not necessarily the same as the_context(). I think, replacing "theContext" with "OldGoals.the_context" in prelude.ML should do the trick! I've checked it in as Revision 1353 of our Hets-lib Cheers Christian From Christian.Maeder at dfki.de Wed Jan 20 13:46:36 2010 From: Christian.Maeder at dfki.de (Christian Maeder) Date: Wed, 20 Jan 2010 13:46:36 +0100 Subject: [Hets-users] Isabelle2009-1 In-Reply-To: <4B56F7E6.6050207@dfki.de> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> <4B5516B9.7070506@rickmurphy.org> <4B55A85A.8090009@dfki.de> <4B567F86.50608@rickmurphy.org> <4B56F7E6.6050207@dfki.de> Message-ID: <4B56FB2C.5050602@dfki.de> Christian Maeder schrieb: > I think, replacing "theContext" with "OldGoals.the_context" in > prelude.ML should do the trick! No, the_context was removed from Isabelle2009-1 entirely. We need to fix this differently. Sorry Christian From rick at rickmurphy.org Thu Jan 21 03:03:29 2010 From: rick at rickmurphy.org (Rick Murphy) Date: Wed, 20 Jan 2010 21:03:29 -0500 Subject: [Hets-users] Isabelle2009-1 In-Reply-To: <4B56FB2C.5050602@dfki.de> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> <4B5516B9.7070506@rickmurphy.org> <4B55A85A.8090009@dfki.de> <4B567F86.50608@rickmurphy.org> <4B56F7E6.6050207@dfki.de> <4B56FB2C.5050602@dfki.de> Message-ID: <4B57B5F1.5010608@rickmurphy.org> Ok, my fault. I should have been more specific about my Isabelle instance. I'll watch for updates. Most likely doing some interim Haskell. -- Rick Christian Maeder wrote: > Christian Maeder schrieb: >> I think, replacing "theContext" with "OldGoals.the_context" in >> prelude.ML should do the trick! > > No, the_context was removed from Isabelle2009-1 entirely. We need to fix > this differently. > > Sorry Christian > > From Christian.Maeder at dfki.de Wed Jan 27 14:39:58 2010 From: Christian.Maeder at dfki.de (Christian Maeder) Date: Wed, 27 Jan 2010 14:39:58 +0100 Subject: [Hets-users] Isabelle 2009(-1) In-Reply-To: <4B567F86.50608@rickmurphy.org> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> <4B5516B9.7070506@rickmurphy.org> <4B55A85A.8090009@dfki.de> <4B567F86.50608@rickmurphy.org> Message-ID: <4B60422E.40706@dfki.de> Dear Hets users, the daily development snapshots of hets have been changed to support Isabelle 2009 by default now. Note that there are two Isabelle 2009 versions: Isabelle2009: April 2009 Isabelle2009-1: December 2009 The generated .thy-files only differ in using "setup" rather than "ML" lines. Apart from a new binary (i.e. hets-2010-01-27.bz2) you also need to update Hets-lib (to Revision 1361 or higher): svn co https://svn-agbkb.informatik.uni-bremen.de/Hets-lib/trunk \ Hets-lib Cheers Christian http://www.dfki.de/sks/hets/installation_e.htm In order to use Isabelle2008 (June 2008) the file Hets-lib/Isabelle/prelude.ML must be overwritten with the contents of Hets-lib/Isabelle/prelude2008.ML and the environment variable HETS_ISABELLE must be set to Isabelle (with a capital "I"), i.e. under a bash: export HETS_ISABELLE=Isabelle otherwise "isabelle emacs" will be called. From rick at rickmurphy.org Sat Jan 30 02:08:35 2010 From: rick at rickmurphy.org (Rick Murphy) Date: Fri, 29 Jan 2010 20:08:35 -0500 Subject: [Hets-users] Isabelle 2009(-1) In-Reply-To: <4B60422E.40706@dfki.de> References: <4B54748D.1050406@rickmurphy.org> <4B547DC1.6000903@dfki.de> <4B5516B9.7070506@rickmurphy.org> <4B55A85A.8090009@dfki.de> <4B567F86.50608@rickmurphy.org> <4B60422E.40706@dfki.de> Message-ID: <4B638693.5070903@rickmurphy.org> +1 Rick Christian Maeder wrote: > Dear Hets users, > > the daily development snapshots of hets have been changed to support > Isabelle 2009 by default now. Note that there are two Isabelle 2009 > versions: > > Isabelle2009: April 2009 > Isabelle2009-1: December 2009 > > The generated .thy-files only differ in using "setup" rather than "ML" > lines. Apart from a new binary (i.e. hets-2010-01-27.bz2) you also need > to update Hets-lib (to Revision 1361 or higher): > > svn co https://svn-agbkb.informatik.uni-bremen.de/Hets-lib/trunk \ > Hets-lib > > Cheers Christian > > http://www.dfki.de/sks/hets/installation_e.htm > > In order to use Isabelle2008 (June 2008) the file > Hets-lib/Isabelle/prelude.ML must be overwritten with the contents of > Hets-lib/Isabelle/prelude2008.ML and the environment variable > HETS_ISABELLE must be set to Isabelle (with a capital "I"), > i.e. under a bash: > > export HETS_ISABELLE=Isabelle > > otherwise "isabelle emacs" will be called. > > >