- Projects

XRM 26: Add MetaModules and some other extensions.
by SIGOURE Benoit 20 May '06

20 May '06

https://svn.lrde.epita.fr/svn/xrm/trunk Index: ChangeLog from SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> Add MetaModules and some other extensions. * src/tools/parse-xrm.str: Factorize some code in common with parse-prism in parser-common.str. Improve use of addPosInfo. * src/tools/parse-prism.str: Likewise. * src/tools/parser-common.str: New. * src/lib/xrm/pp/xrm-module.str: Add MetaModule boxing. * src/lib/xrm/pp/xrm-to-abox.str: Import xrm-arrays. * src/lib/xrm/pp/xrm-arrays.meta: New. * src/lib/xrm/pp/xrm-arrays.str: New. * src/str/xrm-front.str: Use parse-xrm instead of receiving ATerms on stdin. Use addPosInfo by default. * src/str/prism-desugar.str: Update literals used in concrete syntax. * src/syn/xrm/XRM-Arrays.sdf: New. * src/syn/xrm/XRM-StaticIf.sdf: Add EBNF grammar in comments. * src/syn/xrm/XRM-Literals.sdf: New. * src/syn/xrm/XRM-Main.sdf: Import XRM-Literals. * src/syn/xrm/XRM-Module.sdf: Add MetaModule. * src/syn/prism/PRISM-MetaVars.sdf: Add a FIXME. * src/syn/prism/StrategoPRISM.sdf: Add a FIXME. Fix meta-escapes. * tests/xrm/outer-static-if.xpm: Update test case. * tests/xrm/outer-static-for-step.xpm, * tests/xrm/static-if.xpm, * tests/xrm/static-for-step.xpm, * tests/xrm/outer-static-for.xpm, * tests/xrm/static-for.xpm, * tests/test-xrm-front.sh.in: Adapt to xrm-front changes. src/lib/xrm/pp/xrm-arrays.meta | 1 src/lib/xrm/pp/xrm-arrays.str | 9 ++++ src/lib/xrm/pp/xrm-module.str | 8 +++ src/lib/xrm/pp/xrm-to-abox.str | 1 src/str/prism-desugar.str | 16 +++---- src/str/xrm-front.str | 29 +++++++++++- src/syn/prism/PRISM-MetaVars.sdf | 3 - src/syn/prism/StrategoPRISM.sdf | 7 +-- src/syn/xrm/XRM-Arrays.sdf | 19 ++++++++ src/syn/xrm/XRM-Literals.sdf | 12 +++++ src/syn/xrm/XRM-Main.sdf | 1 src/syn/xrm/XRM-Module.sdf | 9 ++++ src/syn/xrm/XRM-StaticIf.sdf | 18 ++++++++ src/tools/parse-prism.str | 81 ++---------------------------------- src/tools/parse-xrm.str | 81 ++---------------------------------- src/tools/parser-common.str | 77 ++++++++++++++++++++++++++++++++++ tests/test-xrm-front.sh.in | 12 ----- tests/xrm/outer-static-for-step.xpm | 5 -- tests/xrm/outer-static-for.xpm | 5 -- tests/xrm/outer-static-if.xpm | 7 --- tests/xrm/static-for-step.xpm | 5 -- tests/xrm/static-for.xpm | 5 -- tests/xrm/static-if.xpm | 7 --- 23 files changed, 214 insertions(+), 204 deletions(-) Index: src/tools/parse-xrm.str --- src/tools/parse-xrm.str (revision 25) +++ src/tools/parse-xrm.str (working copy) @@ -4,6 +4,7 @@ imports liblib tool-doc + parser-common strategies main-parse-xrm = @@ -11,7 +12,7 @@ xtc-io-wrap( parse-xrm-options , parse-xrm-usage - , parse-xrm-about + , parser-about , !["sglr", "implode-asfix", "XRM.tbl", "pp-aterm"] , parse-xrm ) @@ -25,18 +26,13 @@ strategies parse-xrm = - xtc-sglr-no-heuristics(get-parse-table, get-start-symbol) + where(?FILE(input-file-name) + !"stdin" => input-file-name) + ; xtc-sglr-no-heuristics(get-parse-table, get-start-symbol) ; if must-preserve-comments then xtc-transform(!"asfix-anno-comments", pass-verbose) end - /* - ** -p is supposed to be the original path of the file, used in the - ** position representation. The input of addPosInfo is an asfix file, - ** which is not the original source file. The -p option is almost - ** never used by addPosInfo but it's a "mandatory option" - */ ; if must-preserve-positions then - xtc-transform(!"addPosInfo", !["-p", "DummyFileName" + xtc-transform(!"addPosInfo", !["-p", input-file-name | <pass-verbose>()]) end ; xtc-implode-asfix @@ -44,18 +40,6 @@ xtc-transform(!"pp-aterm", pass-verbose) end - /* sglr options: - ** -fi: heuristic: injection count - ** -fe: heuristic: eagerness - ** -2: use AsFix2 output format - ** -A: ambiguities are treated as errors - ** -p <file>: parse table to use - ** -s <symbol>: start symbol to use - */ - xtc-sglr-no-heuristics(tbl, sort) = - xtc-transform(!"sglr", !["-fi", "-fe", "-2A", "-p", <tbl; xtc-find> (), - "-s", <sort> () | <pass-v-verbose> ()]) - strategies symbol-option = @@ -72,50 +56,6 @@ <get-config> "start-symbol" <+ !"ModulesFile" strategies - - preserve-comments-option = - Option("--preserve-comments" - , <set-preserve-comments> "yes" - , !HelpString("--preserve-comments", "Preserve source code - comments as annotations of the abstract syntax tree. [off]") - ) - - set-preserve-comments = - <set-config> ("preserve-comments", <id>) - - must-preserve-comments = - <get-config> "preserve-comments" => "yes" - -strategies - - preserve-positions-option = - Option("--preserve-positions" - , <set-preserve-positions> "yes" - , !HelpString("--preserve-positions", "Preserve source code positions in - the input file as annotations of the abstract syntax tree. - [off]") - ) - - set-preserve-positions = - <set-config> ("preserve-positions", <id>) - - must-preserve-positions = - <get-config> "preserve-positions" => "yes" - -strategies - pp-aterm-option = - Option("-A" + "--pp-aterm" - , <set-pp-aterm> "yes" - , !HelpString("-A | --pp-aterm", "Pretty print output with pp-aterm") - ) - - set-pp-aterm = - <set-config> ("pp-aterm", <id>) - - must-pp-aterm = - <get-config> "pp-aterm" => "yes" - -strategies /* ** We use two parse tables for performances. One of them (XRM.tbl) has a ** single start symbol (ModulesFile) and the other (XRM-StartSymbols) has @@ -145,14 +85,3 @@ , OptionUsage() , AutoReportBugs() ] - - parse-xrm-about = - <tool-doc> - [ AutoProgram() - , Author(Person("SIGOURE Benoit", "sigoure.benoit(a)lrde.epita.fr")) - , GNU_GPL("2006", "SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr>") - , Config([ - DefaultXTCRepository() - , CurrentXTCRepository() - ]) - ] Index: src/tools/parse-prism.str --- src/tools/parse-prism.str (revision 25) +++ src/tools/parse-prism.str (working copy) @@ -4,6 +4,7 @@ imports liblib tool-doc + parser-common strategies main-parse-prism = @@ -11,7 +12,7 @@ xtc-io-wrap( parse-prism-options , parse-prism-usage - , parse-prism-about + , parser-about , !["sglr", "implode-asfix", "PRISM.tbl", "pp-aterm"] , parse-prism ) @@ -25,18 +26,13 @@ strategies parse-prism = - xtc-sglr-no-heuristics(get-parse-table, get-start-symbol) + where(?FILE(input-file-name) + !"stdin" => input-file-name) + ; xtc-sglr-no-heuristics(get-parse-table, get-start-symbol) ; if must-preserve-comments then xtc-transform(!"asfix-anno-comments", pass-verbose) end - /* - ** -p is supposed to be the original path of the file, used in the - ** position representation. The input of addPosInfo is an asfix file, - ** which is not the original source file. The -p option is almost - ** never used by addPosInfo but it's a "mandatory option" - */ ; if must-preserve-positions then - xtc-transform(!"addPosInfo", !["-p", "DummyFileName" + xtc-transform(!"addPosInfo", !["-p", input-file-name | <pass-verbose>()]) end ; xtc-implode-asfix @@ -44,18 +40,6 @@ xtc-transform(!"pp-aterm", pass-verbose) end - /* sglr options: - ** -fi: heuristic: injection count - ** -fe: heuristic: eagerness - ** -2: use AsFix2 output format - ** -A: ambiguities are treated as errors - ** -p <file>: parse table to use - ** -s <symbol>: start symbol to use - */ - xtc-sglr-no-heuristics(tbl, sort) = - xtc-transform(!"sglr", !["-fi", "-fe", "-2A", "-p", <tbl; xtc-find> (), - "-s", <sort> () | <pass-v-verbose> ()]) - strategies symbol-option = @@ -72,50 +56,6 @@ <get-config> "start-symbol" <+ !"ModulesFile" strategies - - preserve-comments-option = - Option("--preserve-comments" - , <set-preserve-comments> "yes" - , !HelpString("--preserve-comments", "Preserve source code - comments as annotations of the abstract syntax tree. [off]") - ) - - set-preserve-comments = - <set-config> ("preserve-comments", <id>) - - must-preserve-comments = - <get-config> "preserve-comments" => "yes" - -strategies - - preserve-positions-option = - Option("--preserve-positions" - , <set-preserve-positions> "yes" - , !HelpString("--preserve-positions", "Preserve source code positions in - the input file as annotations of the abstract syntax tree. - [off]") - ) - - set-preserve-positions = - <set-config> ("preserve-positions", <id>) - - must-preserve-positions = - <get-config> "preserve-positions" => "yes" - -strategies - pp-aterm-option = - Option("-A" + "--pp-aterm" - , <set-pp-aterm> "yes" - , !HelpString("-A | --pp-aterm", "Pretty print output with pp-aterm") - ) - - set-pp-aterm = - <set-config> ("pp-aterm", <id>) - - must-pp-aterm = - <get-config> "pp-aterm" => "yes" - -strategies /* ** We use two parse tables for performances. One of them (PRISM.tbl) has a ** single start symbol (ModulesFile) and the other (PRISM-StartSymbols) has @@ -145,14 +85,3 @@ , OptionUsage() , AutoReportBugs() ] - - parse-prism-about = - <tool-doc> - [ AutoProgram() - , Author(Person("SIGOURE Benoit", "sigoure.benoit(a)lrde.epita.fr")) - , GNU_GPL("2006", "SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr>") - , Config([ - DefaultXTCRepository() - , CurrentXTCRepository() - ]) - ] Index: src/tools/parser-common.str --- src/tools/parser-common.str (revision 0) +++ src/tools/parser-common.str (revision 0) @@ -0,0 +1,77 @@ +// used by parse-prism and parse-xrm to factorize common strategies + +module parser-common + +strategies + + /* sglr options: + ** -fi: heuristic: injection count + ** -fe: heuristic: eagerness + ** -2: use AsFix2 output format + ** -A: ambiguities are treated as errors + ** -p <file>: parse table to use + ** -s <symbol>: start symbol to use + */ + xtc-sglr-no-heuristics(tbl, sort) = + xtc-transform(!"sglr", !["-fi", "-fe", "-2A", "-p", <tbl; xtc-find> (), + "-s", <sort> () | <pass-v-verbose> ()]) + +strategies + + preserve-comments-option = + Option("--preserve-comments" + , <set-preserve-comments> "yes" + , !HelpString("--preserve-comments", "Preserve source code + comments as annotations of the abstract syntax tree. [off]") + ) + + set-preserve-comments = + <set-config> ("preserve-comments", <id>) + + must-preserve-comments = + <get-config> "preserve-comments" => "yes" + +strategies + + preserve-positions-option = + Option("--preserve-positions" + , <set-preserve-positions> "yes" + , !HelpString("--preserve-positions", "Preserve source code positions in + the input file as annotations of the abstract syntax tree. + [off]") + ) + + set-preserve-positions = + <set-config> ("preserve-positions", <id>) + + must-preserve-positions = + <get-config> "preserve-positions" => "yes" + +strategies + pp-aterm-option = + Option("-A" + "--pp-aterm" + , <set-pp-aterm> "yes" + , !HelpString("-A | --pp-aterm", "Pretty print output with pp-aterm") + ) + + set-pp-aterm = + <set-config> ("pp-aterm", <id>) + + must-pp-aterm = + <get-config> "pp-aterm" => "yes" + +/** + * Documentation + */ +strategies + + parser-about = + <tool-doc> + [ AutoProgram() + , Author(Person("SIGOURE Benoit", "sigoure.benoit(a)lrde.epita.fr")) + , GNU_GPL("2006", "SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr>") + , Config([ + DefaultXTCRepository() + , CurrentXTCRepository() + ]) + ] Index: src/lib/xrm/pp/xrm-module.str --- src/lib/xrm/pp/xrm-module.str (revision 25) +++ src/lib/xrm/pp/xrm-module.str (working copy) @@ -10,3 +10,11 @@ ] KW["endmodule"] ] ]| + + prism-to-box: + MetaModule(array-access, dec-or-cmd-list) + -> box |[ V[ V is=2 [ H hs=1 [KW["module"] ~array-access ] + ~*dec-or-cmd-list + ] + KW["endmodule"] + ] ]| Index: src/lib/xrm/pp/xrm-to-abox.str --- src/lib/xrm/pp/xrm-to-abox.str (revision 25) +++ src/lib/xrm/pp/xrm-to-abox.str (working copy) @@ -23,6 +23,7 @@ xrm-module xrm-staticfor xrm-staticif + xrm-arrays strategies Index: src/lib/xrm/pp/xrm-arrays.meta --- src/lib/xrm/pp/xrm-arrays.meta (revision 0) +++ src/lib/xrm/pp/xrm-arrays.meta (revision 0) @@ -0,0 +1 @@ +Meta([Syntax("Stratego-Box")]) Index: src/lib/xrm/pp/xrm-arrays.str --- src/lib/xrm/pp/xrm-arrays.str (revision 0) +++ src/lib/xrm/pp/xrm-arrays.str (revision 0) @@ -0,0 +1,9 @@ +module xrm-arrays + +rules + + // Identifier ArraySubscript+ -> ArrayAccess + prism-to-box: + ArrayAccess(idf, subscripts) -> H hs=0 [ ~idf ~*subscripts-impl ] + where + <mapconcat(\ x -> [S("["), x, S("]")] \)> subscripts => subscripts-impl Index: src/str/xrm-front.str --- src/str/xrm-front.str (revision 25) +++ src/str/xrm-front.str (working copy) @@ -14,14 +14,18 @@ xrm-front-options , xrm-front-usage , xrm-front-about - , !["pp-prism", "pp-aterm"] + , !["parse-xrm", "pp-prism", "pp-aterm"] , main-wrapped ) main-wrapped = check-options - ; read-from + ; xtc-transform(!"parse-xrm", // parse input (returns a FILE) + !["-b", "--preserve-positions" | <pass-verbose>()]) + ; read-from // read parsed input + ; strip-annos ; xrm-front-pipeline // transformations + ; if not(must-keep-attributes) then strip-annos end ; if <get-config> "-b" then write-to // output binary ATerms else @@ -51,6 +55,7 @@ xrm-front-options = pp-prism-option + pp-aterm-option + + keep-attributes-option + desugar-option // check the options are consistent @@ -72,6 +77,21 @@ strategies + keep-attributes-option = + Option("-K" + "--keep-attributes" + , <set-keep-attributes> "yes" + , !HelpString("-K | --keep-attributes", "Don't remove attributes when + printing ATerms (this includes positions)") + ) + + set-keep-attributes = + <set-config> ("keep-attributes", <id>) + + must-keep-attributes = + <get-config> "keep-attributes" => "yes" + +strategies + pp-prism-option = Option("-P" + "--pp-prism" , <set-pp-prism> "yes" @@ -121,8 +141,9 @@ xrm-front-usage = <tool-doc> [ Usage("xrm-front [OPTIONS]") - , Summary("Transforms an eXtended Reactive Module abstract syntax - tree in ATerm format to a PRISM-equivalant abstract syntax tree.") + , Summary("Transforms an eXtended Reactive Module source file in a + PRISM-equivalant abstract syntax tree (default) or source + code. (see option -P)") , OptionUsage() , AutoReportBugs() ] Index: src/str/prism-desugar.str --- src/str/prism-desugar.str (revision 25) +++ src/str/prism-desugar.str (working copy) @@ -64,27 +64,27 @@ compare(s) = if s then !True() else !False() end catch-div-by-zero = - ?|[ e / 0 ]|; fatal-err-msg(|"Division by zero detected") + ?|[ e / 0D ]|; debug; fatal-err-msg(|"Division by zero detected") rules AddZero: - |[ e + 0 ]| -> |[ e ]| + |[ e + 0D ]| -> |[ e ]| AddZero: - |[ 0 + e ]| -> |[ e ]| + |[ 0D + e ]| -> |[ e ]| MulOne: - |[ e * 1 ]| -> |[ e ]| + |[ e * 1D ]| -> |[ e ]| MulOne: - |[ 1 * e ]| -> |[ e ]| + |[ 1D * e ]| -> |[ e ]| MulZero: - |[ e * 0 ]| -> |[ 0 ]| + |[ e * 0D ]| -> |[ 0D ]| MulZero: - |[ 0 * e ]| -> |[ 0 ]| + |[ 0D * e ]| -> |[ 0D ]| DivOne: - |[ e / 1 ]| -> |[ e ]| + |[ e / 1D ]| -> |[ e ]| EvalPlus: |[ d1 + d2 ]| -> |[ r ]| where <addR>(d1, d2) => r Index: src/syn/xrm/XRM-Arrays.sdf --- src/syn/xrm/XRM-Arrays.sdf (revision 0) +++ src/syn/xrm/XRM-Arrays.sdf (revision 0) @@ -0,0 +1,19 @@ +module XRM-Arrays +imports + PRISM-to-XRM +exports + + %% EBNF Grammar: Arrays + %% ArrayAccess ::= Identifier ArraySubscript {ArraySubscript} + %% + %% ArraySubscript ::= "[" Identifier "]" + + sorts ArrayAccess ArraySubscript + context-free syntax + Identifier ArraySubscript+ -> ArrayAccess {cons("ArrayAccess")} + "[" Identifier "]" -> ArraySubscript {bracket} + + %% Note: why use bracket here? + %% This is a work around. In SDF, if a production contains literals, + %% a constructor is required. With brackets, we specify that literals + %% are allowed there, however this does not create a node in the AST. Index: src/syn/xrm/XRM-StaticIf.sdf --- src/syn/xrm/XRM-StaticIf.sdf (revision 25) +++ src/syn/xrm/XRM-StaticIf.sdf (working copy) @@ -4,6 +4,24 @@ XRM-Module exports + %% EBNF Grammar: Static if-then-else + %% (* if-then-else at top-level *) + %% ModulesFileSection ::= + %% "if" Expression "then" ModulesFileSection {ModulesFileSection} "end" + %% | "if" Expression "then" + %% ModulesFileSection {ModulesFileSection} + %% "else" + %% ModulesFileSection {ModulesFileSection} + %% "end" + %% + %% DeclarationOrCommand ::= + %% "if" Expression "then" DeclarationOrCommand {DeclarationOrCommand} "end" + %% | "if" Expression "then" + %% DeclarationOrCommand {DeclarationOrCommand} + %% "else" + %% DeclarationOrCommand {DeclarationOrCommand} + %% "end" + context-free syntax %% if-then at top-level "if" Expression "then" ModulesFileSection+ "end" Index: src/syn/xrm/XRM-Literals.sdf --- src/syn/xrm/XRM-Literals.sdf (revision 0) +++ src/syn/xrm/XRM-Literals.sdf (revision 0) @@ -0,0 +1,12 @@ +module XRM-Literals +imports PRISM-to-XRM +exports + + syntax + <LInt-LEX> "d" -> <LDouble-CF> {bracket} + <LInt-LEX> "D" -> <LDouble-CF> {bracket} + <LInt-LEX> "f" -> <LDouble-CF> {bracket} + <LInt-LEX> "F" -> <LDouble-CF> {bracket} + + lexical restrictions + LDouble -/- [dDfF] Index: src/syn/xrm/XRM-Main.sdf --- src/syn/xrm/XRM-Main.sdf (revision 25) +++ src/syn/xrm/XRM-Main.sdf (working copy) @@ -5,3 +5,4 @@ XRM-Expression XRM-StaticFor XRM-StaticIf + XRM-Literals Index: src/syn/xrm/XRM-Module.sdf --- src/syn/xrm/XRM-Module.sdf (revision 25) +++ src/syn/xrm/XRM-Module.sdf (working copy) @@ -1,9 +1,18 @@ module XRM-Module imports PRISM-to-XRM + XRM-Arrays exports + %% EBNF Grammar: eXtended Modules + %% DeclarationOrCommand ::= Declaration | Command + %% Module ::= "module" ArrayAccess Declaration* Command* "endmodule" + sorts DeclarationOrCommand context-free syntax Declaration -> DeclarationOrCommand Command -> DeclarationOrCommand + + context-free syntax + "module" ArrayAccess DeclarationOrCommand* "endmodule" + -> Module {cons("MetaModule")} Index: src/syn/prism/PRISM-MetaVars.sdf --- src/syn/prism/PRISM-MetaVars.sdf (revision 25) +++ src/syn/prism/PRISM-MetaVars.sdf (working copy) @@ -4,7 +4,8 @@ [ijkln][0-9]* -> LInt {prefer} [dr][0-9]* -> LDouble {prefer} [xyzfgh][0-9]* -> ID {prefer} -%% [xyzfgh][0-9]* "'" -> IdentifierPrime {prefer} + %% FIXME: The following doesn't work + [xyzfgh][0-9]* "'" -> IdentifierPrime {prefer} [e][0-9]* -> Expression {prefer} "a"[0-9]* "*" -> {Identifier ","}+ {prefer} [m][0-9]* -> Module {prefer} Index: src/syn/prism/StrategoPRISM.sdf --- src/syn/prism/StrategoPRISM.sdf (revision 25) +++ src/syn/prism/StrategoPRISM.sdf (working copy) @@ -19,7 +19,8 @@ context-free syntax "~" StrategoTerm -> Expression {prefer,cons("FromTerm")} - "~id:" StrategoTerm -> Identifier {prefer,cons("FromTerm")} + "~id:" StrategoTerm -> ID {prefer,cons("FromTerm")} + %% FIXME: The following doesn't work "~id':" StrategoTerm -> IdentifierPrime {prefer,cons("FromTerm")} - "~int:" StrategoTerm -> Int {prefer,cons("FromTerm")} - "~double:" StrategoTerm -> Double {prefer,cons("FromTerm")} + "~int:" StrategoTerm -> LInt {prefer,cons("FromTerm")} + "~double:" StrategoTerm -> LDouble {prefer,cons("FromTerm")} Index: tests/xrm/outer-static-if.xpm --- tests/xrm/outer-static-if.xpm (revision 25) +++ tests/xrm/outer-static-if.xpm (working copy) @@ -1,16 +1,13 @@ probabilistic for i from 0 to 42 do - // FIXME: the following line shall be: - //module dummy[i] - // but this is not supported ATM if i<1 & i>41 then - module dummy + module dummy[i] x : [0..1]; [] x=0 -> x'=1; endmodule else - module dummy + module dummy[i] x : [0..2]; [] x=1 -> x'=2; endmodule Index: tests/xrm/outer-static-for-step.xpm --- tests/xrm/outer-static-for-step.xpm (revision 25) +++ tests/xrm/outer-static-for-step.xpm (working copy) @@ -1,10 +1,7 @@ probabilistic for i from 0 to 42 step 2 do - // FIXME: the following line shall be: - //module dummy[i] - // but this is not supported ATM - module dummy + module dummy[i] x : [0..1]; [] x=0 -> x'=1; endmodule Index: tests/xrm/static-if.xpm --- tests/xrm/static-if.xpm (revision 25) +++ tests/xrm/static-if.xpm (working copy) @@ -1,11 +1,8 @@ probabilistic for i from 0 to 42 do - // FIXME: the following line shall be: - //module dummy[i] - // but this is not supported ATM if i<1 & i>41 then - module dummy + module dummy[i] for i from 0 to 42 do // FIXME: x here shall be replaced with x[i] // but this is not supported ATM @@ -18,7 +15,7 @@ end endmodule else - module dummyelse + module dummyelse[i] y : [0..1]; [] y=0 -> y'=1; endmodule Index: tests/xrm/static-for-step.xpm --- tests/xrm/static-for-step.xpm (revision 25) +++ tests/xrm/static-for-step.xpm (working copy) @@ -1,10 +1,7 @@ probabilistic for i from 0 to 42 step 2 do - // FIXME: the following line shall be: - //module dummy[i] - // but this is not supported ATM - module dummy + module dummy[i] for i from 0 to 42 step 2 do // FIXME: x here shall be replaced with x[i] // but this is not supported ATM Index: tests/xrm/outer-static-for.xpm --- tests/xrm/outer-static-for.xpm (revision 25) +++ tests/xrm/outer-static-for.xpm (working copy) @@ -1,10 +1,7 @@ probabilistic for i from 0 to 42 do - // FIXME: the following line shall be: - //module dummy[i] - // but this is not supported ATM - module dummy + module dummy[i] x : [0..1]; [] x=0 -> x'=1; endmodule Index: tests/xrm/static-for.xpm --- tests/xrm/static-for.xpm (revision 25) +++ tests/xrm/static-for.xpm (working copy) @@ -1,10 +1,7 @@ probabilistic for i from 0 to 42 do - // FIXME: the following line shall be: - //module dummy[i] - // but this is not supported ATM - module dummy + module dummy[i] for i from 0 to 42 do // FIXME: x here shall be replaced with x[i] // but this is not supported ATM Index: tests/test-xrm-front.sh.in --- tests/test-xrm-front.sh.in (revision 25) +++ tests/test-xrm-front.sh.in (working copy) @@ -27,17 +27,9 @@ echo ">>> Starting the test for $basefile" test_cnt=$((test_cnt + 1)) - echo @ECHO_N@ " Parsing $basefile ... " - "@top_builddir@/src/tools/parse-xrm" -i "$file" -o "$outdir/$bfile.xpm.aterm" - if [ $? -ne 0 ]; then - echo 'FAILED, continuing with the next test...' - continue - fi - echo 'OK, no ambiguities found' - echo @ECHO_N@ " Converting $basefile into standard PRISM AST ... " "@top_builddir@/src/str/xrm-front" \ - -i "$outdir/$bfile.xpm.aterm" -o "$outdir/$bfile.pm.aterm" + -i "$file" -o "$outdir/$bfile.pm.aterm" if [ $? -ne 0 ]; then echo 'FAILED, continuing with the next test...' continue @@ -46,7 +38,7 @@ echo @ECHO_N@ " Converting $basefile into standard PRISM code ... " "@top_builddir@/src/str/xrm-front" -P \ - -i "$outdir/$bfile.xpm.aterm" -o "$outdir/$bfile.pm" + -i "$file" -o "$outdir/$bfile.pm" if [ $? -ne 0 ]; then echo 'FAILED, continuing with the next test...' continue

1 0

XRM 25: Extend real-eq.
by SIGOURE Benoit 18 May '06

18 May '06

https://svn.lrde.epita.fr/svn/xrm/trunk Index: ChangeLog from SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> Extend real-eq. * src/str/xrm-to-prism.str: Optimize traversals. * src/str/reals.str: Extend real-eq. reals.str | 11 ++++++----- xrm-to-prism.str | 9 ++------- 2 files changed, 8 insertions(+), 12 deletions(-) Index: src/str/xrm-to-prism.str --- src/str/xrm-to-prism.str (revision 24) +++ src/str/xrm-to-prism.str (working copy) @@ -4,7 +4,7 @@ strategies xrm-to-prism = - topdown(repeat(xrm-to-prism-desugar)) + topdown(try(xrm-to-prism-desugar)) ; id // unroll static for loops here ; topdown(try(reorder-module-contents)) @@ -33,7 +33,7 @@ reorder-module-contents = ?Module(name, _) ; {| DeclarationList, CommandList: - Module(id, map(try(fetch-module-content))) + Module(id, map(add-command <+ add-declaration)) ; where( bagof-DeclarationList ; ?dec-list @@ -43,11 +43,6 @@ |} ; !Module(name, dec-list, cmd-list) - // we're inside a Module(_, content) - // each element is either a Command(_, _, _) or a declaration - fetch-module-content = - add-command <+ add-declaration - // if the current term is a Command, add it in the CommandList DR add-command = ?Command(_, _, _) Index: src/str/reals.str --- src/str/reals.str (revision 24) +++ src/str/reals.str (working copy) @@ -3,7 +3,7 @@ strategies // for some reason the following rules are only defined for integers - // in the stratego-lib. Here is their equivalant for reals. + // in the stratego-lib. Here is their equivalent for reals. addR = (string-to-real, string-to-real); addr; real-to-string subtR = (string-to-real, string-to-real); subtr; real-to-string mulR = (string-to-real, string-to-real); mulr; real-to-string @@ -15,13 +15,14 @@ /** ** tests whether two reals are equal - ** => tests if there difference is less than 10^-7 + ** => tests if there difference is less than the precision given in parameter ** operates on the input term (a, b) - ** eg: <real-eq> ("0.42", "0.43") -> fail + ** eg: <real-eq(|0.00001)> ("0.42", "0.43") -> fail */ - real-eq = + real-eq(|precision) = (string-to-real, string-to-real) ; subtr ; abs => difference - ; <not(gtr)>(difference, 0.0000001) + ; <not(gtr)>(difference, precision) + real-eq = real-eq(|0.0000001)

1 0

XRM 24: Improve desugarisation, fix concrete syntax.
by SIGOURE Benoit 18 May '06

18 May '06

https://svn.lrde.epita.fr/svn/xrm/trunk Index: ChangeLog from SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> Improve desugarisation, fix concrete syntax. * src/str/xrm-front.str: Remove empty line. * src/str/prism-desugar.meta: New. * src/str/Makefile.am: Update header. * src/str/prism-desugar.str: Use concrete syntax. Improve and simplify desugarisation. Extend to doubles. * src/str/reals.str: New. * src/syn/xrm/Makefile.am: Add parse table generation for concrete syntax. * src/syn/prism/Makefile.am: Likewise. * src/syn/prism/PRISM-MetaVars.sdf: Fix meta-vars. * src/syn/prism/PRISM-Identifier.sdf: Fix restrictions. str/Makefile.am | 4 - str/prism-desugar.meta | 1 str/prism-desugar.str | 111 ++++++++++++++++++++--------------------- str/reals.str | 27 +++++++++ str/xrm-front.str | 1 syn/prism/Makefile.am | 5 + syn/prism/PRISM-Identifier.sdf | 4 + syn/prism/PRISM-MetaVars.sdf | 7 +- syn/xrm/Makefile.am | 5 + 9 files changed, 98 insertions(+), 67 deletions(-) Index: src/str/xrm-front.str --- src/str/xrm-front.str (revision 23) +++ src/str/xrm-front.str (working copy) @@ -113,7 +113,6 @@ must-desugar = <get-config> "desugar" => "yes" - /** * Documentation */ Index: src/str/prism-desugar.meta --- src/str/prism-desugar.meta (revision 0) +++ src/str/prism-desugar.meta (revision 0) @@ -0,0 +1 @@ +Meta([Syntax("StrategoXRM")]) Index: src/str/Makefile.am --- src/str/Makefile.am (revision 23) +++ src/str/Makefile.am (working copy) @@ -1,11 +1,11 @@ ## -## Makefile.am for xrm in /home/tsuna/work/xrm/trunk/src +## Makefile.am for xrm in /home/tsuna/work/xrm/trunk/src/str ## ## Made by SIGOURE Benoit ## Mail <sigoure.benoit(a)lrde.epita.fr> ## ## Started on Thu Apr 27 17:59:35 2006 SIGOURE Benoit -## Last update Fri May 12 16:55:22 2006 SIGOURE Benoit +## Last update Thu May 18 16:49:50 2006 SIGOURE Benoit ## include $(top_srcdir)/config/Transformers.mk Index: src/str/prism-desugar.str --- src/str/prism-desugar.str (revision 23) +++ src/str/prism-desugar.str (working copy) @@ -1,15 +1,48 @@ module prism-desugar -imports PRISM +imports + PRISM + reals strategies + /** + ** 1. First convert all literal ints to doubles. This makes the + ** simplifications easier to write since there is less cases to + ** address (eg: int + int, int + double, double + int, + ** double + double etc.). + ** 2. Desugar the AST + ** 3. Convert the doubles which are round numbers back to ints. + ** Simplify doubles (round them up and remove trailing zeros) + */ prism-desugar = - innermost( + topdown(try(IntToDouble)) + ; innermost( RemoveUnconditionnalUpdates - <+ AddZero <+ MulOne <+ MulZero <+ DivOne <+ warn-div-by-zero + <+ AddZero <+ MulOne <+ MulZero <+ DivOne <+ catch-div-by-zero <+ EvalPlus <+ EvalMinus <+ EvalMul <+ EvalDiv <+ EvalLt <+ EvalLtEq <+ EvalGt <+ EvalGtEq <+ EvalEq <+ EvalNeq ) + ; topdown(try(SimplifyDoubles <+ TruncateDouble)) + +rules + + IntToDouble: Int(i) -> Double(i) + + // Transform a round (or almost round) Double into an Int. + SimplifyDoubles: + Double(d) -> Int(i) // eg: Double("41.9999999999999") -> Int("42") + where !d // "41.9999999999999" + ; string-to-real // 4.199999999999990e+01 + ; real-to-string(|6) // "42.000000" + ; split-at-dot // ("42", "000000") + ; (?i, ?decimals) // i="42", decimals="000000" + // strcmp(decimals, "000000") == 0 ? id : fail + ; <strcmp> (<explode-string> decimals, <explode-string> "000000") => 0 + + // Round real numbers up to 6 digits and remove trailing zeros + TruncateDouble: + Double(d) -> Double(d') + where <rtrim-chars(?'0')> (<real-to-string(|6)> (<string-to-real> d)) => d' rules @@ -30,89 +63,55 @@ // used by the rules below to evaluate constant comparisons compare(s) = if s then !True() else !False() end - // for some reason the following rules are only defined for integers - // in the stratego-lib. Here is their equivalant for reals. - // NOTE: We only keep 6 digits after the period - addR = (string-to-real, string-to-real); addr; real-to-string(|6) - subtR = (string-to-real, string-to-real); subtr; real-to-string(|6) - mulR = (string-to-real, string-to-real); mulr; real-to-string(|6) - divR = (string-to-real, string-to-real); divr; real-to-string(|6) - - warn-div-by-zero = - ?Div(e, Int("0")); fatal-err-msg(|"Division by zero") + catch-div-by-zero = + ?|[ e / 0 ]|; fatal-err-msg(|"Division by zero detected") rules AddZero: - Plus(e, Int("0")) -> e + |[ e + 0 ]| -> |[ e ]| AddZero: - Plus(Int("0"), e) -> e + |[ 0 + e ]| -> |[ e ]| MulOne: - Mul(e, Int("1")) -> e + |[ e * 1 ]| -> |[ e ]| MulOne: - Mul(Int("1"), e) -> e + |[ 1 * e ]| -> |[ e ]| MulZero: - Mul(e, Int("0")) -> Int("0") + |[ e * 0 ]| -> |[ 0 ]| MulZero: - Mul(Int("0"), e) -> Int("0") + |[ 0 * e ]| -> |[ 0 ]| DivOne: - Div(e, Int("1")) -> e + |[ e / 1 ]| -> |[ e ]| EvalPlus: - Plus(Int(i), Int(j)) -> Int(<addS>(i, j)) - EvalPlus: - Plus(Double(i), Int(j)) -> Double(<addR>(i, j)) - EvalPlus: - Plus(Int(i), Double(j)) -> Double(<addR>(i, j)) - EvalPlus: - Plus(Double(i), Double(j)) -> Double(<addR>(i, j)) + |[ d1 + d2 ]| -> |[ r ]| where <addR>(d1, d2) => r EvalMinus: - Minus(Int(i), Int(j)) -> Int(<subtS>(i, j)) - EvalMinus: - Minus(Double(i), Int(j)) -> Double(<subtR>(i, j)) - EvalMinus: - Minus(Int(i), Double(j)) -> Double(<subtR>(i, j)) - EvalMinus: - Minus(Double(i), Double(j)) -> Double(<subtR>(i, j)) + |[ d1 - d2 ]| -> |[ r ]| where <subtR>(d1, d2) => r EvalMul: - Mul(Int(i), Int(j)) -> Int(<mulS>(i, j)) - EvalMul: - Mul(Double(i), Int(j)) -> Double(<mulR>(i, j)) - EvalMul: - Mul(Int(i), Double(j)) -> Double(<mulR>(i, j)) - EvalMul: - Mul(Double(i), Double(j)) -> Double(<mulR>(i, j)) + |[ d1 * d2 ]| -> |[ r ]| where <mulR>(d1, d2) => r EvalDiv: - Div(Int(i), Int(j)) -> Double(<divR>(i, j)) - EvalDiv: - Div(Double(i), Int(j)) -> Double(<divR>(i, j)) - EvalDiv: - Div(Int(i), Double(j)) -> Double(<divR>(i, j)) - EvalDiv: - Div(Double(i), Double(j)) -> Double(<divR>(i, j)) - - // FIXME: considere adding the Double's in the following: + |[ d1 / d2 ]| -> |[ r ]| where <divR>(d1, d2) => r EvalLt: - Lt(Int(i), Int(j)) -> <compare(ltS)>(i, j) + |[ d1 < d2 ]| -> <compare(ltR)>(d1, d2) EvalLtEq: - LtEq(Int(i), Int(j)) -> <compare(leqS)>(i, j) + |[ d1 <= d2 ]| -> <compare(leqR)>(d1, d2) EvalGt: - Gt(Int(i), Int(j)) -> <compare(gtS)>(i, j) + |[ d1 > d2 ]| -> <compare(gtR)>(d1, d2) EvalGtEq: - GtEq(Int(i), Int(j)) -> <compare(geqS)>(i, j) + |[ d1 >= d2 ]| -> <compare(geqR)>(d1, d2) EvalEq: - Eq(Int(i), Int(j)) -> <compare(eq)>(i, j) + |[ d1 = d2 ]| -> <compare(real-eq)>(d1, d2) EvalNeq: - NotEq(Int(i), Int(j)) -> <compare(not(eq))>(i, j) + |[ d1 != d2 ]| -> <compare(not(real-eq))>(d1, d2) Index: src/str/reals.str --- src/str/reals.str (revision 0) +++ src/str/reals.str (revision 0) @@ -0,0 +1,27 @@ +module reals + +strategies + + // for some reason the following rules are only defined for integers + // in the stratego-lib. Here is their equivalant for reals. + addR = (string-to-real, string-to-real); addr; real-to-string + subtR = (string-to-real, string-to-real); subtr; real-to-string + mulR = (string-to-real, string-to-real); mulr; real-to-string + divR = (string-to-real, string-to-real); divr; real-to-string + gtR = where((string-to-real, string-to-real); gtr) + geqR = where((string-to-real, string-to-real); ?(x,x) <+ gtr) + ltR = where((string-to-real, string-to-real); not(?(x,x) <+ gtr)) + leqR = where((string-to-real, string-to-real); not(gtr)) + + /** + ** tests whether two reals are equal + ** => tests if there difference is less than 10^-7 + ** operates on the input term (a, b) + ** eg: <real-eq> ("0.42", "0.43") -> fail + */ + real-eq = + (string-to-real, string-to-real) + ; subtr + ; abs => difference + ; <not(gtr)>(difference, 0.0000001) + Index: src/syn/xrm/Makefile.am --- src/syn/xrm/Makefile.am (revision 23) +++ src/syn/xrm/Makefile.am (working copy) @@ -5,7 +5,7 @@ ## Mail <sigoure.benoit(a)lrde.epita.fr> ## ## Started on Thu Apr 27 17:40:41 2006 SIGOURE Benoit -## Last update Fri May 12 16:57:54 2006 SIGOURE Benoit +## Last update Thu May 18 16:53:05 2006 SIGOURE Benoit ## include $(top_srcdir)/config/Transformers.mk @@ -28,7 +28,8 @@ XRM.str \ XRM-StartSymbols.tbl \ XRM-Prefixed.def \ - StrategoXRM.def + StrategoXRM.def \ + StrategoXRM.tbl pkgdata_DATA = $(pkgdata_DATA_built) sdfdata_DATA = $(pkgdata_DATA_built) Index: src/syn/prism/PRISM-MetaVars.sdf --- src/syn/prism/PRISM-MetaVars.sdf (revision 23) +++ src/syn/prism/PRISM-MetaVars.sdf (working copy) @@ -1,9 +1,10 @@ module PRISM-MetaVars exports variables - [ijkln][0-9]* -> Int {prefer} - [xyzfgh][0-9]* -> Identifier {prefer} - [xyzfgh][0-9]* "'" -> IdentifierPrime {prefer} + [ijkln][0-9]* -> LInt {prefer} + [dr][0-9]* -> LDouble {prefer} + [xyzfgh][0-9]* -> ID {prefer} +%% [xyzfgh][0-9]* "'" -> IdentifierPrime {prefer} [e][0-9]* -> Expression {prefer} "a"[0-9]* "*" -> {Identifier ","}+ {prefer} [m][0-9]* -> Module {prefer} Index: src/syn/prism/PRISM-Identifier.sdf --- src/syn/prism/PRISM-Identifier.sdf (revision 23) +++ src/syn/prism/PRISM-Identifier.sdf (working copy) @@ -31,7 +31,9 @@ Keyword -> ID {reject} lexical restrictions - Identifier -/- [A-Za-z0-9\_] + ID -/- [A-Za-z0-9\_] + + context-free restrictions IdentifierPrime -/- [A-Za-z0-9\_]%% "'" %% FIXME ^^^ %% causes an error oO Index: src/syn/prism/Makefile.am --- src/syn/prism/Makefile.am (revision 23) +++ src/syn/prism/Makefile.am (working copy) @@ -5,7 +5,7 @@ ## Mail <sigoure.benoit(a)lrde.epita.fr> ## ## Started on Thu Apr 27 17:40:41 2006 SIGOURE Benoit -## Last update Fri May 12 16:52:04 2006 SIGOURE Benoit +## Last update Thu May 18 16:52:52 2006 SIGOURE Benoit ## include $(top_srcdir)/config/Transformers.mk @@ -26,7 +26,8 @@ PRISM.str \ PRISM-StartSymbols.tbl \ PRISM-Prefixed.def \ - StrategoPRISM.def + StrategoPRISM.def \ + StrategoPRISM.tbl pkgdata_DATA = $(pkgdata_DATA_built) sdfdata_DATA = $(pkgdata_DATA_built)

1 0

XRM 23: Add constant folding / constant expression evaluation.
by SIGOURE Benoit 16 May '06

16 May '06

https://svn.lrde.epita.fr/svn/xrm/trunk Index: ChangeLog from SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> Add constant folding / constant expression evaluation. * src/str/prism-desugar.str: Here. prism-desugar.str | 112 +++++++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 107 insertions(+), 5 deletions(-) Index: src/str/prism-desugar.str --- src/str/prism-desugar.str (revision 22) +++ src/str/prism-desugar.str (working copy) @@ -4,13 +4,115 @@ strategies prism-desugar = - topdown(repeat(PrismDesugar)) + innermost( + RemoveUnconditionnalUpdates + <+ AddZero <+ MulOne <+ MulZero <+ DivOne <+ warn-div-by-zero + <+ EvalPlus <+ EvalMinus <+ EvalMul <+ EvalDiv + <+ EvalLt <+ EvalLtEq <+ EvalGt <+ EvalGtEq <+ EvalEq <+ EvalNeq + ) rules - PrismDesugar: - // [] guard -> unconditionnal-update-list; - // is rewritten as: - // [] guard -> 1:(unconditionnal-update-list); + /** + ** [] guard -> unconditionnal-update-list; + ** is rewritten as: + ** [] guard -> 1:(unconditionnal-update-list); + */ + RemoveUnconditionnalUpdates: AlwaysUpdate(UpdateList(update-elements)) -> ProbUpdateList([ProbUpdate(Int("1"), UpdateList(update-elements))]) + +/** +** Constant folding / Constant expression evaluation +*/ +strategies + + // used by the rules below to evaluate constant comparisons + compare(s) = if s then !True() else !False() end + + // for some reason the following rules are only defined for integers + // in the stratego-lib. Here is their equivalant for reals. + // NOTE: We only keep 6 digits after the period + addR = (string-to-real, string-to-real); addr; real-to-string(|6) + subtR = (string-to-real, string-to-real); subtr; real-to-string(|6) + mulR = (string-to-real, string-to-real); mulr; real-to-string(|6) + divR = (string-to-real, string-to-real); divr; real-to-string(|6) + + warn-div-by-zero = + ?Div(e, Int("0")); fatal-err-msg(|"Division by zero") + +rules + + AddZero: + Plus(e, Int("0")) -> e + AddZero: + Plus(Int("0"), e) -> e + + MulOne: + Mul(e, Int("1")) -> e + MulOne: + Mul(Int("1"), e) -> e + + MulZero: + Mul(e, Int("0")) -> Int("0") + MulZero: + Mul(Int("0"), e) -> Int("0") + + DivOne: + Div(e, Int("1")) -> e + + EvalPlus: + Plus(Int(i), Int(j)) -> Int(<addS>(i, j)) + EvalPlus: + Plus(Double(i), Int(j)) -> Double(<addR>(i, j)) + EvalPlus: + Plus(Int(i), Double(j)) -> Double(<addR>(i, j)) + EvalPlus: + Plus(Double(i), Double(j)) -> Double(<addR>(i, j)) + + EvalMinus: + Minus(Int(i), Int(j)) -> Int(<subtS>(i, j)) + EvalMinus: + Minus(Double(i), Int(j)) -> Double(<subtR>(i, j)) + EvalMinus: + Minus(Int(i), Double(j)) -> Double(<subtR>(i, j)) + EvalMinus: + Minus(Double(i), Double(j)) -> Double(<subtR>(i, j)) + + EvalMul: + Mul(Int(i), Int(j)) -> Int(<mulS>(i, j)) + EvalMul: + Mul(Double(i), Int(j)) -> Double(<mulR>(i, j)) + EvalMul: + Mul(Int(i), Double(j)) -> Double(<mulR>(i, j)) + EvalMul: + Mul(Double(i), Double(j)) -> Double(<mulR>(i, j)) + + EvalDiv: + Div(Int(i), Int(j)) -> Double(<divR>(i, j)) + EvalDiv: + Div(Double(i), Int(j)) -> Double(<divR>(i, j)) + EvalDiv: + Div(Int(i), Double(j)) -> Double(<divR>(i, j)) + EvalDiv: + Div(Double(i), Double(j)) -> Double(<divR>(i, j)) + + // FIXME: considere adding the Double's in the following: + + EvalLt: + Lt(Int(i), Int(j)) -> <compare(ltS)>(i, j) + + EvalLtEq: + LtEq(Int(i), Int(j)) -> <compare(leqS)>(i, j) + + EvalGt: + Gt(Int(i), Int(j)) -> <compare(gtS)>(i, j) + + EvalGtEq: + GtEq(Int(i), Int(j)) -> <compare(geqS)>(i, j) + + EvalEq: + Eq(Int(i), Int(j)) -> <compare(eq)>(i, j) + + EvalNeq: + NotEq(Int(i), Int(j)) -> <compare(not(eq))>(i, j)

1 0

XRM 22: Fix literals' and identifiers' constructions.
by SIGOURE Benoit 16 May '06

16 May '06

https://svn.lrde.epita.fr/svn/xrm/trunk Index: ChangeLog from SIGOURE Benoit <sigoure.benoit(a)lrde.epita.fr> Fix literals' and identifiers' constructions. The Int, Double and Identifier nodes were constructed only when they appeared within Expression nodes. This commit fixes this. * src/syn/prism/PRISM-Literals.sdf: Always construct AST nodes for literals (Int and Double)... * src/syn/prism/PRISM-Identifier.sdf: ... and for Identifiers. * src/syn/prism/PRISM-Expression.sdf: Remove constructions. * src/lib/xrm/pp/xrm-module.str: Change the boxing rules accordingly. * src/lib/xrm/pp/xrm-staticfor.str, * src/lib/prism/pp/prism-update.str, * src/lib/prism/pp/prism-reward.str, * src/lib/prism/pp/prism-module.str, * src/lib/prism/pp/prism-declaration.str, * src/lib/prism/pp/prism-constant.str, * src/lib/prism/pp/prism-expression.str, * src/lib/prism/pp/prism-formula.str, * src/lib/prism/pp/prism-systemcomposition.str, * src/lib/prism/pp/prism-command.str: Likewise. * vcs/xrm.rb: Change the mailing list to projects(a)lrde.epita.fr. src/lib/prism/pp/prism-command.str | 4 +-- src/lib/prism/pp/prism-constant.str | 24 +++++++++--------- src/lib/prism/pp/prism-declaration.str | 14 +++++----- src/lib/prism/pp/prism-expression.str | 4 +-- src/lib/prism/pp/prism-formula.str | 2 - src/lib/prism/pp/prism-module.str | 10 +++---- src/lib/prism/pp/prism-reward.str | 4 +-- src/lib/prism/pp/prism-systemcomposition.str | 4 +-- src/lib/prism/pp/prism-update.str | 4 +-- src/lib/xrm/pp/xrm-module.str | 4 +-- src/lib/xrm/pp/xrm-staticfor.str | 36 +++++++++++++-------------- src/syn/prism/PRISM-Expression.sdf | 6 ++-- src/syn/prism/PRISM-Identifier.sdf | 2 - src/syn/prism/PRISM-Literals.sdf | 15 +++++++---- vcs/xrm.rb | 2 - 15 files changed, 70 insertions(+), 65 deletions(-) Index: src/lib/xrm/pp/xrm-module.str --- src/lib/xrm/pp/xrm-module.str (revision 21) +++ src/lib/xrm/pp/xrm-module.str (working copy) @@ -4,8 +4,8 @@ // "module" Identifier DeclarationOrCommand* "endmodule" -> Module prism-to-box: - Module(s, dec-or-cmd-list) - -> box |[ V[ V is=2 [ H hs=1 [KW["module"] s ] + Module(idf, dec-or-cmd-list) + -> box |[ V[ V is=2 [ H hs=1 [KW["module"] ~idf ] ~*dec-or-cmd-list ] KW["endmodule"] Index: src/lib/xrm/pp/xrm-staticfor.str --- src/lib/xrm/pp/xrm-staticfor.str (revision 21) +++ src/lib/xrm/pp/xrm-staticfor.str (working copy) @@ -5,11 +5,11 @@ // "for" Identifier "from" Int "to" Int "do" ModulesFileSection+ "end" // -> OuterStaticFor prism-to-box: - OuterStaticFor(s, for-from, for-to, mfs-list) + OuterStaticFor(idf, for-from, for-to, mfs-list) -> box |[ V[ V is=2 [ - H hs=1 [ KW["for"] s - KW["from"] NUM[~lit:for-from] - KW["to"] NUM[~lit:for-to] KW["do"] ] + H hs=1 [ KW["for"] ~idf + KW["from"] ~for-from + KW["to"] ~for-to KW["do"] ] ~*mfs-list // in the innermost V box ] KW["end"] // in the outermost V box ] ]| @@ -17,12 +17,12 @@ // "for" Identifier "from" Int "to" Int "step" Int "do" // ModulesFileSection+ "end" -> OuterStaticFor prism-to-box: - OuterStaticFor(s, for-from, for-to, for-step, mfs-list) + OuterStaticFor(idf, for-from, for-to, for-step, mfs-list) -> box |[ V[ V is=2 [ - H hs=1 [ KW["for"] s - KW["from"] NUM[~lit:for-from] - KW["to"] NUM[~lit:for-to] - KW["step"] NUM[~lit:for-step] KW["do"] ] + H hs=1 [ KW["for"] ~idf + KW["from"] ~for-from + KW["to"] ~for-to + KW["step"] ~for-step KW["do"] ] ~*mfs-list // in the innermost V box ] KW["end"] // in the outermost V box ] ]| @@ -30,11 +30,11 @@ // "for" Identifier "from" Int "to" Int "do" DeclarationOrCommand+ "end" // -> InnerStaticFor prism-to-box: - InnerStaticFor(s, for-from, for-to, dec-or-cmd-list) + InnerStaticFor(idf, for-from, for-to, dec-or-cmd-list) -> box |[ V[ V is=2 [ - H hs=1 [ KW["for"] s - KW["from"] NUM[~lit:for-from] - KW["to"] NUM[~lit:for-to] KW["do"] ] + H hs=1 [ KW["for"] ~idf + KW["from"] ~for-from + KW["to"] ~for-to KW["do"] ] ~*dec-or-cmd-list // in the innermost V box ] KW["end"] // in the outermost V box ] ]| @@ -42,12 +42,12 @@ // "for" Identifier "from" Int "to" Int "step" Int "do" // DeclarationOrCommand+ "end" -> InnerStaticFor prism-to-box: - InnerStaticFor(s, for-from, for-to, for-step, dec-or-cmd-list) + InnerStaticFor(idf, for-from, for-to, for-step, dec-or-cmd-list) -> box |[ V[ V is=2 [ - H hs=1 [ KW["for"] s - KW["from"] NUM[~lit:for-from] - KW["to"] NUM[~lit:for-to] - KW["step"] NUM[~lit:for-step] KW["do"] ] + H hs=1 [ KW["for"] ~idf + KW["from"] ~for-from + KW["to"] ~for-to + KW["step"] ~for-step KW["do"] ] ~*dec-or-cmd-list // in the innermost V box ] KW["end"] // in the outermost V box ] ]| Index: src/lib/prism/pp/prism-update.str --- src/lib/prism/pp/prism-update.str (revision 21) +++ src/lib/prism/pp/prism-update.str (working copy) @@ -28,5 +28,5 @@ -> H hs=0 [ "(" ~idf-prime "=" ~exp ")" ] prism-to-box: - IdentifierPrime(s) -> box |[ s2 ]| - where <conc-strings> (s, "'") => s2 + IdentifierPrime(idf) -> box |[ VAR[~lit:idf-primed] ]| + where <conc-strings> (idf, "'") => idf-primed Index: src/lib/prism/pp/prism-reward.str --- src/lib/prism/pp/prism-reward.str (revision 21) +++ src/lib/prism/pp/prism-reward.str (working copy) @@ -10,8 +10,8 @@ // "[" Identifier? "]" Expression ":" Expression ";" -> RewardStructItem prism-to-box: - RewardStructItem(Some(s), exp1, exp2) - -> H hs=1 [ H hs=0 ["[" s "]"] ~exp1 ":" H hs=0 [~exp2 ";"] ] + RewardStructItem(Some(idf), exp1, exp2) + -> H hs=1 [ H hs=0 ["[" ~idf "]"] ~exp1 ":" H hs=0 [~exp2 ";"] ] prism-to-box: RewardStructItem(None(), exp1, exp2) Index: src/lib/prism/pp/prism-module.str --- src/lib/prism/pp/prism-module.str (revision 21) +++ src/lib/prism/pp/prism-module.str (working copy) @@ -5,8 +5,8 @@ // "module" Identifier Declaration* Command* "endmodule" -> Module prism-to-box: - Module(s, dec-list, cmd-list) - -> box |[ V[ V is=2 [ H hs=1 [KW["module"] s ] + Module(idf, dec-list, cmd-list) + -> box |[ V[ V is=2 [ H hs=1 [KW["module"] ~idf ] ~*dec-list ~*cmd-list ] @@ -16,8 +16,8 @@ // "module" Identifier "=" Identifier "[" Renames "]" "endmodule" // -> RenamedModule prism-to-box: - RenamedModule(s1, s2, renames) - -> H hs=1 [KW["module"] s1 "=" s2 "[" ~renames "]" KW["endmodule"]] + RenamedModule(idf1, idf2, renames) + -> H hs=1 [KW["module"] ~idf1 "=" ~idf2 "[" ~renames "]" KW["endmodule"]] // {Rename ","}+ -> Renames prism-to-box: @@ -26,4 +26,4 @@ // Identifier "=" Identifier -> Rename prism-to-box: - Rename(s1, s2) -> H hs=0 [ s1 "=" s2 ] + Rename(idf1, idf2) -> H hs=0 [ ~idf1 "=" ~idf2 ] Index: src/lib/prism/pp/prism-declaration.str --- src/lib/prism/pp/prism-declaration.str (revision 21) +++ src/lib/prism/pp/prism-declaration.str (working copy) @@ -4,22 +4,22 @@ // Identifier ":" "[" Expression ".." Expression "]" ";" -> IntDecNoInit prism-to-box: - IntDecNoInit(s, range-start, range-end) - -> H hs=1 [ s ":" H hs=0["[" ~range-start ".." ~range-end "]" ";"] ] + IntDecNoInit(idf, range-start, range-end) + -> H hs=1 [ ~idf ":" H hs=0["[" ~range-start ".." ~range-end "]" ";"] ] // Identifier ":" "[" Expression ".." Expression "]" "init" Expression ";" // -> IntDec prism-to-box: - IntDec(s, range-start, range-end, exp-init) - -> H hs=1 [ s ":" H hs=0 ["[" ~range-start ".." ~range-end "]"] + IntDec(idf, range-start, range-end, exp-init) + -> H hs=1 [ ~idf ":" H hs=0 ["[" ~range-start ".." ~range-end "]"] KW["init"] H hs=0[~exp-init ";"] ] // Identifier ":" "bool" ";" -> BoolDecNoInit prism-to-box: - BoolDecNoInit(s) -> H hs=1 [ s ":" H hs=0 [KW["bool"] ";"] ] + BoolDecNoInit(idf) -> H hs=1 [ ~idf ":" H hs=0 [KW["bool"] ";"] ] // Identifier ":" "bool" "init" Expression ";" -> BoolDec prism-to-box: - BoolDec(s, exp-init) - -> H hs=1 [ s ":" KW["bool"] KW["init"] H hs=0 [~exp-init ";"] ] + BoolDec(idf, exp-init) + -> H hs=1 [ ~idf ":" KW["bool"] KW["init"] H hs=0 [~exp-init ";"] ] Index: src/lib/prism/pp/prism-constant.str --- src/lib/prism/pp/prism-constant.str (revision 21) +++ src/lib/prism/pp/prism-constant.str (working copy) @@ -4,30 +4,30 @@ // "const" "int" Identifier "=" Expression ";" -> ConstInt prism-to-box: - ConstInt(s, exp) - -> H hs=1 [ KW["const"] KW["int"] s "=" H hs=0 [~exp ";"] ] + ConstInt(idf, exp) + -> H hs=1 [ KW["const"] KW["int"] ~idf "=" H hs=0 [~exp ";"] ] // "const" "int" Identifier ";" -> ConstIntNoInit prism-to-box: - ConstIntNoInit(s) - -> H hs=1 [ KW["const"] KW["int"] H hs=0 [s ";"] ] + ConstIntNoInit(idf) + -> H hs=1 [ KW["const"] KW["int"] H hs=0 [~idf ";"] ] // "const" "double" Identifier "=" Expression ";" -> ConstDouble prism-to-box: - ConstDouble(s, exp) - -> H hs=1 [ KW["const"] KW["double"] s "=" H hs=0[~exp ";"] ] + ConstDouble(idf, exp) + -> H hs=1 [ KW["const"] KW["double"] ~idf "=" H hs=0 [~exp ";"] ] // "const" "double" Identifier ";" -> ConstDoubleNoInit prism-to-box: - ConstDoubleNoInit(s) - -> H hs=1 [ KW["const"] KW["double"] H hs=0[s ";"] ] + ConstDoubleNoInit(idf) + -> H hs=1 [ KW["const"] KW["double"] H hs=0 [~idf ";"] ] // "const" "bool" Identifier "=" Expression ";" -> ConstBool prism-to-box: - ConstBool(s, exp) - -> H hs=1 [ KW["const"] KW["bool"] s "=" H hs=0[~exp ";"] ] + ConstBool(idf, exp) + -> H hs=1 [ KW["const"] KW["bool"] ~idf "=" H hs=0 [~exp ";"] ] // "const" "bool" Identifier ";" -> ConstBoolNoInit prism-to-box: - ConstBoolNoInit(s) - -> H hs=1 [ KW["const"] KW["bool"] H hs=0[s ";"] ] + ConstBoolNoInit(idf) + -> H hs=1 [ KW["const"] KW["bool"] H hs=0 [~idf ";"] ] Index: src/lib/prism/pp/prism-expression.str --- src/lib/prism/pp/prism-expression.str (revision 21) +++ src/lib/prism/pp/prism-expression.str (working copy) @@ -95,8 +95,8 @@ where <implode-list(|",")> args => args-imploded prism-to-box: - Call(s, args) - -> box |[ H hs=0 [ "func" "(" H hs=1 [ H hs=0[s ","] + Call(idf, args) + -> box |[ H hs=0 [ "func" "(" H hs=1 [ H hs=0[~idf ","] ~args-imploded ] ")" ] ]| where <implode-list(|",")> args => args-imploded Index: src/lib/prism/pp/prism-formula.str --- src/lib/prism/pp/prism-formula.str (revision 21) +++ src/lib/prism/pp/prism-formula.str (working copy) @@ -4,4 +4,4 @@ // "formula" Identifier "=" Expression ";" -> FormulaDef prism-to-box: - FormulaDef(s, exp) -> H hs=1 [ KW["formula"] s "=" H hs=0 [~exp ";"] ] + FormulaDef(idf, exp) -> H hs=1 [ KW["formula"] ~idf "=" H hs=0 [~exp ";"] ] Index: src/lib/prism/pp/prism-systemcomposition.str --- src/lib/prism/pp/prism-systemcomposition.str (revision 21) +++ src/lib/prism/pp/prism-systemcomposition.str (working copy) @@ -50,11 +50,11 @@ // Identifier -> SystemAction prism-to-box: - SystemAction(s) -> s + SystemAction(idf) -> idf // Identifier -> SystemModule prism-to-box: - SystemModule(s) -> s + SystemModule(idf) -> idf // "(" SystemParallels ")" -> SystemBrackets prism-to-box: Index: src/lib/prism/pp/prism-command.str --- src/lib/prism/pp/prism-command.str (revision 21) +++ src/lib/prism/pp/prism-command.str (working copy) @@ -4,8 +4,8 @@ // "[" Identifier? "]" Expression "->" Updates ";" -> Command prism-to-box: - Command(Some(s), exp, updates) - -> H hs=0 [ "[" s H hs=1 [ "]" ~exp "->" ~updates ] ";" ] + Command(Some(idf), exp, updates) + -> H hs=0 [ "[" ~idf H hs=1 [ "]" ~exp "->" ~updates ] ";" ] prism-to-box: Command(None(), exp, updates) Index: src/syn/prism/PRISM-Literals.sdf --- src/syn/prism/PRISM-Literals.sdf (revision 21) +++ src/syn/prism/PRISM-Literals.sdf (working copy) @@ -9,10 +9,15 @@ %% NOTE: PRISM's grammar doesn't allow things such as "01" - sorts Int Double + sorts LInt LDouble lexical syntax - [1-9][0-9]* -> Int - "0" -> Int + [1-9][0-9]* -> LInt + "0" -> LInt + + [1-9][0-9]* "." [0-9]+ -> LDouble + "0" "." [0-9]+ -> LDouble - [1-9][0-9]* "." [0-9]+ -> Double - "0" "." [0-9]+ -> Double + sorts Int Double + context-free syntax + LInt -> Int {cons("Int")} + LDouble -> Double {cons("Double")} Index: src/syn/prism/PRISM-Identifier.sdf --- src/syn/prism/PRISM-Identifier.sdf (revision 21) +++ src/syn/prism/PRISM-Identifier.sdf (working copy) @@ -23,7 +23,7 @@ sorts Identifier IdentifierPrime ID context-free syntax - ID -> Identifier + ID -> Identifier {cons("Identifier")} ID "'" -> IdentifierPrime {cons("IdentifierPrime")} lexical syntax Index: src/syn/prism/PRISM-Expression.sdf --- src/syn/prism/PRISM-Expression.sdf (revision 21) +++ src/syn/prism/PRISM-Expression.sdf (working copy) @@ -116,11 +116,11 @@ %% Literals "true" -> Expression {cons("True")} "false" -> Expression {cons("False")} - Int -> Expression {cons("Int")} - Double -> Expression {cons("Double")} + Int -> Expression + Double -> Expression %% Variables - Identifier -> Expression {cons("Identifier")} + Identifier -> Expression %% Arithmetic and logical operations %% The order of the list defines the operator precedence Index: vcs/xrm.rb --- vcs/xrm.rb (revision 21) +++ vcs/xrm.rb (working copy) @@ -7,7 +7,7 @@ def xrm_commit! ( *args ) common_commit!("XRM 22: <%= title %>", *args) do |subject| - mail!(:to => %w[lrde-patches(a)lrde.epita.fr], :subject => subject) + mail!(:to => %w[projects(a)lrde.epita.fr], :subject => subject) end end alias_command :xrmci, :xrm_commit

1 0

build-farm: 285: Modify caracas CXXFLAGS.
by michael.cadilhac＠lrde.org 13 May '06

13 May '06

Index: buildfarm_worker/caracas.fns =================================================================== --- buildfarm_worker/caracas.fns (revision 283) +++ buildfarm_worker/caracas.fns (working copy) @@ -20,7 +20,7 @@ CC=icpc # fixme: to remove ccache CXX=icpc -CXXFLAGS="-D_LINUX -cxxlib-icc" +CXXFLAGS="-D_LINUX -cxxlib-icc -I/home/lrde/admin/lrde/usr/include/c++/" PATH=$PATH:/home/lrde/admin/lrde/usr/bin export LD_LIBRARY_PATH=/home/lrde/admin/lrde/usr/lib/

1 0

build-farm: 283: Clean-up before up and add some projects.
by michael.cadilhac＠lrde.org 11 May '06

11 May '06

https://svn/svn/lrde-tools/trunk/build-farm Index: maintainer/update_unpacked --- maintainer/update_unpacked (revision 282) +++ maintainer/update_unpacked (working copy) @@ -5,14 +5,15 @@ # SVN_PATH="/work/master/unpacked" -SVN_PACKAGES="c-grammar contract-c cxx-grammar esdf boxedsdf sdf-astgen sdf-attribute sdf-detgen sdf-option str-lazy cxx-basic specs-grammar transformers olena vaucanson tc-maintainer clsp havm nolimips monoburg" +SVN_PACKAGES="c-grammar contract-c cxx-grammar esdf boxedsdf sdf-astgen sdf-attribute sdf-detgen sdf-option str-lazy cxx-basic specs-grammar transformers olena vaucanson tc-maintainer clsp havm nolimips monoburg ranch ddd" # these packages need a svn-revision file containing the lastest svn revision. # historical dep from nix. SVN_REVISIONS="c-grammar contract-c cxx-grammar esdf boxedsdf sdf-astgen sdf-attribute sdf-detgen sdf-option str-lazy cxx-basic specs-grammar transformers" for i in $SVN_PACKAGES; do - ( cd "$SVN_PATH/$i" && svn -q update ) + echo $i + ( cd "$SVN_PATH/$i" && svn cleanup && svn -q update ) done for i in $SVN_REVISIONS; do

1 0

Re: [Projects] support of -N option in VCS
by Tsuna 11 May '06

11 May '06

On 2006-05-11, Nicolas Pouillard <nicolas.pouillard(a)gmail.com> wrote: > On 5/11/06, Tsuna <tsuna(a)warszawa.lrde.epita.fr> wrote: >> On 2006-05-10, Nicolas Pouillard <nicolas.pouillard(a)gmail.com> wrote: >> > On 5/11/06, Tsuna <tsuna(a)warszawa.lrde.epita.fr> wrote: >> >> On 2006-05-10, Nicolas Pouillard <nicolas.pouillard(a)gmail.com> wrote: >> >> > On 5/10/06, Tsuna <tsuna(a)warszawa.lrde.epita.fr> wrote: >> >> >> >> >> >> Cool thanks for the reply :) >> >> >> Is it possible to exclude ``.*.sw?'' in ~/.vcs ? >> >> >> >> >> > >> >> > Yes but if you want talk about vim files toto.swp toto.swo this regexp >> >> > will works better: \.sw.$ >> >> > >> >> >> >> ok if it's a perl-like (or ruby-like) regexp, shouldn't it be \..*\.sw. >> >> ? >> >> >> > >> > It's a ruby regexp (so almost like a perl one). >> > >> > Don't forget the $ to anchor the end of the filename: \..*\.sw.$ >> > >> >> For some reason it won't work :/ >> $ cat ~/.vcs >> exclude: >> - Makefile.in >> - \..*\.sw.$ >> $ svn st >> + +committed >> ? src/tools/.pp-prism.str.swp >> >> it's weird since it works for the Makefile.in >> moreover, if it's a real ruby regexp, shouldn't I have: >> exclude: >> - Makefile\.in >> instead? >> >> any idea? >> > > By default the type is a plain string not a regexp, one use "!re" to > specify the type in Yaml: > > exclude: > - Makefile.in # and not Makefile\.in > - !re ^\..*\.sw.$ > > BTW please fix your from address (Tsuna <tsuna(a)**warszawa**.lrde.epita.fr>). > Done :) Thank you very much for your support! -- SIGOURE Benoit aka Tsuna _____ /EPITA\ Promo 2008.CSI Rock & tRoll

2 1

Re: [Projects] support of -N option in VCS
by Tsuna 11 May '06

11 May '06

On 2006-05-10, Nicolas Pouillard <nicolas.pouillard(a)gmail.com> wrote: > On 5/11/06, Tsuna <tsuna(a)warszawa.lrde.epita.fr> wrote: >> On 2006-05-10, Nicolas Pouillard <nicolas.pouillard(a)gmail.com> wrote: >> > On 5/10/06, Tsuna <tsuna(a)warszawa.lrde.epita.fr> wrote: >> >> >> >> Cool thanks for the reply :) >> >> Is it possible to exclude ``.*.sw?'' in ~/.vcs ? >> >> >> > >> > Yes but if you want talk about vim files toto.swp toto.swo this regexp >> > will works better: \.sw.$ >> > >> >> ok if it's a perl-like (or ruby-like) regexp, shouldn't it be \..*\.sw. >> ? >> > > It's a ruby regexp (so almost like a perl one). > > Don't forget the $ to anchor the end of the filename: \..*\.sw.$ > For some reason it won't work :/ $ cat ~/.vcs exclude: - Makefile.in - \..*\.sw.$ $ svn st + +committed ? src/tools/.pp-prism.str.swp it's weird since it works for the Makefile.in moreover, if it's a real ruby regexp, shouldn't I have: exclude: - Makefile\.in instead? any idea? -- SIGOURE Benoit aka Tsuna _____ /EPITA\ Promo 2008.CSI Rock & tRoll

2 1

Re: [Projects] support of -N option in VCS
by Tsuna 10 May '06

10 May '06

On 2006-05-10, Nicolas Pouillard <nicolas.pouillard(a)gmail.com> wrote: > On 5/10/06, Tsuna <tsuna(a)warszawa.lrde.epita.fr> wrote: >> >> Cool thanks for the reply :) >> Is it possible to exclude ``.*.sw?'' in ~/.vcs ? >> > > Yes but if you want talk about vim files toto.swp toto.swo this regexp > will works better: \.sw.$ > ok if it's a perl-like (or ruby-like) regexp, shouldn't it be \..*\.sw. ? -- SIGOURE Benoit aka Tsuna _____ /EPITA\ Promo 2008.CSI Rock & tRoll

2 1