June 2006 - Projects

Re: XRM 56: Minor cleanups in the grammar.
by Akim Demaille 13 Jun '06

13 Jun '06

>>> "SIGOURE" == SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: > Index: src/syn/prism/PRISM-Main.sdf > --- src/syn/prism/PRISM-Main.sdf (revision 55) > +++ src/syn/prism/PRISM-Main.sdf (working copy) > @@ -1,20 +1,4 @@ > module PRISM-Main > imports > - PRISM-Command > - PRISM-Constant > - PRISM-Declaration > - PRISM-Expression > - PRISM-Formula > - PRISM-Global > - PRISM-Identifier > - PRISM-Init > - PRISM-Label > PRISM-Layout > - PRISM-Literals > - PRISM-Module > PRISM-ModulesFile > - PRISM-ModulesFileType > - PRISM-Reward > - PRISM-SystemComposition > - PRISM-Update How come they are useless? There were included elsewhere?

2 1

Re: XRM 54: Add RIVF's equivalent in XRM meta-code.
by Akim Demaille 13 Jun '06

13 Jun '06

>>> "SIGOURE" == SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: > https://svn.lrde.epita.fr/svn/xrm/trunk > Index: ChangeLog > from SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> > Add RIVF's equivalent in XRM meta-code. > Usage: xrm-front -i rivf.xpm -o rivf.pm -D > -D stands for --desugar (it will add an additional pass after the > main pipeline to desugar as much as possible everything and perform > constant propagation). This is completely optional. > One thing has been omitted here: support for battery_mode=0 (which is > basically the same thing but without bothering with b[x][y] and > without having to check at every update whether or not we should go > in the OFF state because we run out of power). I omitted this for the > sake of clarity although it's fairly straightforward to add it (since > it's only a matter of commenting out some lines). It is really nice. However, the M4 solution remains clearer IMHO. Functions will be most welcome.

2 1

Re: XRM 53: Add meta-if at Expression level.
by Akim Demaille 13 Jun '06

13 Jun '06

>>> "SIGOURE" == SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: I don't think you need these two: > + EvalIf: > + |[ d ? e1 : e2 ]| -> |[ e1 ]| > + where <not(compare(real-eq))>(d, 0) > + > + EvalIf: > + |[ d ? e1 : e2 ]| -> |[ e2 ]| > + where <compare(real-eq)>(d, 0) > + Since > + LitToBool: > + |[ d ]| -> |[ false ]| > + where <compare(real-eq)>(d, 0) > + > + LitToBool: > + |[ d ]| -> |[ true ]| > + where <not(compare(real-eq))>(d, 0) and > ?MetaIf(condition, then-part, else-part) > ///*DEBUG*/; say(!" @@@ eval-meta-if: starting:") > ///*DEBUG*/; printf(|" condition = ", condition) > - ; where(<prism-desugar> condition => condition-value) > + ; where(<prism-desugar> condition > + ; try(LitToBool) => condition-value) > ///*DEBUG*/; printf(|" condition-value = ", condition-value) > ; if !condition-value => True() then > <eval-meta-code> then-part

2 1

Re: XRM 51: Add static constant propagations in static definitions.
by Akim Demaille 13 Jun '06

13 Jun '06

>>> "SIGOURE" == SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: > collect-static-const-decl = > ?ConstInt(idf, value) > - ; where(!value{Type("int")} => v) > ; if <ExpandStaticConsts> idf then > cannot-redefine-static-const(|idf) > end > ; if <ExpandFormulas> idf then > cannot-redefine-formula(|idf) > end Why don't you factor your checkings into a "<check-redefinitions> idf"? > + ; where(<innermost(ExpandStaticConsts + ExpandFormulas)> value => v' > + ; !v'{Type("int")} => v) > ; rules(ExpandStaticConsts: idf -> v) > collect-static-const-decl = > ?ConstDouble(idf, value) > - ; where(!value{Type("double")} => v) > ; if <ExpandStaticConsts> idf then > cannot-redefine-static-const(|idf) > end > ; if <ExpandFormulas> idf then > cannot-redefine-formula(|idf) > end > + ; where(<innermost(ExpandStaticConsts + ExpandFormulas)> value => v' > + ; !v'{Type("double")} => v) > ; rules(ExpandStaticConsts: idf -> v) > collect-static-const-decl = > ?ConstBool(idf, value) > - ; where(!value{Type("bool")} => v) > ; if <ExpandStaticConsts> idf then > cannot-redefine-static-const(|idf) > end > ; if <ExpandFormulas> idf then > cannot-redefine-formula(|idf) > end > + ; where(<innermost(ExpandStaticConsts + ExpandFormulas)> value => v' > + ; !v'{Type("bool")} => v) > ; rules(ExpandStaticConsts: idf -> v) Also, the truplicate where clause seems truplication of code...

2 1

Re: XRM 49: Add the static_rand builtin.
by Akim Demaille 13 Jun '06

13 Jun '06

>>> "SIGOURE" == SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: > Add the static_rand builtin. > NOTE: Some C was needed to implement rand(from, to) because for some > reason I didn't succeed in writing something equivalent in Stratego. > For some reason, the numbers I generated were all too small. Finalement, tu sais pourquoi ?

2 1

rivf.xpm
by Akim Demaille 13 Jun '06

13 Jun '06

Est-ce que les modifications que je propose sont bonnes ?

1 0

Re: XRM 53: Add meta-if at Expression level.
by Akim Demaille 13 Jun '06

13 Jun '06

>>> "SIGOURE" == SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: > + EvalIf: > + |[ true ? e1 : e2 ]| -> |[ e1 ]| > + > + EvalIf: > + |[ false ? e1 : e2 ]| -> |[ e2 ]| > + > + EvalIf: > + |[ d ? e1 : e2 ]| -> |[ e1 ]| > + where <not(compare(real-eq))>(d, 0) > + > + EvalIf: > + |[ d ? e1 : e2 ]| -> |[ e2 ]| > + where <compare(real-eq)>(d, 0) How about testing whether e1 and e2 evaluate to the same value? I guess a previous phase did it already, as for the truth value, so EvalIf: |[ d ? e : e ]| -> |[ e ]|

2 2

Re: XRM 55: Add documentation draft for XRM.
by michael.cadilhac＠lrde.org 12 Jun '06

12 Jun '06

SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: > Index: doc/user-guide.txt > --- doc/user-guide.txt (revision 0) > +++ doc/user-guide.txt (revision 0) > @@ -0,0 +1,392 @@ I'm sure you want to know more about RST [1], isn't it ? Footnotes: [1] http://docutils.sourceforge.net/rst.html -- | La culture c'est comme la confiture, Michaël `Micha' Cadilhac | | c'est meilleur avec du pain. cadilh_m - Epita 2007 - CSI | | -- MOI59 JID: micha(a)amessage.be | `-- - - - - --'

2 1

Re: XRM 67: Fix pretty-printing.
by Akim Demaille 12 Jun '06

12 Jun '06

>>> "SIGOURE" == SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: > +for file in `find "@srcdir@" -name '*.pctl' -o -name '*.csl' | sort`; do Honestly, using @var@ in a script reads awful. I strongly suggest adding strdir='@srcdir@' in the prolog, and using $srcdir.

1 0

Re: XRM 66: Add parameterized formulas.
by Akim Demaille 12 Jun '06

12 Jun '06

>>> "SIGOURE" == SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> writes: > Index: tests/test-parse-xrm.sh.in > --- tests/test-parse-xrm.sh.in (revision 65) > +++ tests/test-parse-xrm.sh.in (working copy) > @@ -18,7 +18,7 @@ > "@top_builddir@/src/tools/parse-xrm" -i "$file" -o /dev/null > rv=$? > test_cnt=$((test_cnt + 1)) > - if [ $? -eq 0 ]; then > + if [ $rv -eq 0 ]; then > echo 'OK, no ambiguities found' > test_pass=$((test_pass + 1)) > else > Index: tests/test-parse-xpctl.sh.in > --- tests/test-parse-xpctl.sh.in (revision 65) > +++ tests/test-parse-xpctl.sh.in (working copy) > @@ -18,7 +18,7 @@ > "@top_builddir@/src/tools/parse-xpctl" -i "$file" -o /dev/null > rv=$? > test_cnt=$((test_cnt + 1)) > - if [ $? -eq 0 ]; then > + if [ $rv -eq 0 ]; then > echo 'OK, no ambiguities found' > test_pass=$((test_pass + 1)) > else > I I let you guess what my thoughts are... :)

1 0