Re: [Cocci] Checking the parsing for a nested SmPL disjunction
On Fri, 26 Jun 2020, Markus Elfring wrote: > I assumed that I may omit the semicolon in such a SmPL code. > Can the specification of a SmPL nest construct ever be sufficient here? > >>> > >>> No. <+... ...+> matches a subtree of an AST. For a variable > >>> initialization, there is no subtree of the AST that includes both the > >>> right side of an = and the ;. > >> > >> Can the abstract syntax tree be adjusted accordingly? > > > > No. > > Can such a rejection mean that software development efforts look undesirable > (while adjustments for the data model would eventually be possible)? No, it means that it makes no sense and it is not going to happen. julia ___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
Re: [Cocci] Checking the parsing for a nested SmPL disjunction
I assumed that I may omit the semicolon in such a SmPL code. Can the specification of a SmPL nest construct ever be sufficient here? >>> >>> No. <+... ...+> matches a subtree of an AST. For a variable >>> initialization, there is no subtree of the AST that includes both the >>> right side of an = and the ;. >> >> Can the abstract syntax tree be adjusted accordingly? > > No. Can such a rejection mean that software development efforts look undesirable (while adjustments for the data model would eventually be possible)? Regards, Markus ___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
Re: [Cocci] Checking the parsing for a nested SmPL disjunction
On Fri, 26 Jun 2020, Markus Elfring wrote: > >> I assumed that I may omit the semicolon in such a SmPL code. > >> Can the specification of a SmPL nest construct ever be sufficient here? > > > > No. <+... ...+> matches a subtree of an AST. For a variable > > initialization, there is no subtree of the AST that includes both the > > right side of an = and the ;. > > Can the abstract syntax tree be adjusted accordingly? No. julia ___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
Re: [Cocci] Checking the parsing for a nested SmPL disjunction
>> I assumed that I may omit the semicolon in such a SmPL code. >> Can the specification of a SmPL nest construct ever be sufficient here? > > No. <+... ...+> matches a subtree of an AST. For a variable > initialization, there is no subtree of the AST that includes both the > right side of an = and the ;. Can the abstract syntax tree be adjusted accordingly? Regards, Markus ___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
Re: [Cocci] Checking the parsing for a nested SmPL disjunction
On Fri, 26 Jun 2020, Markus Elfring wrote: > >> <+... when any > >> (t2 y = <+... > >> ( sizeof(*(resource)) > >> | > >> * *(resource) > >> ) ...+> > > > > You are missing a ; here. > > I assumed that I may omit the semicolon in such a SmPL code. > Can the specification of a SmPL nest construct ever be sufficient here? No. <+... ...+> matches a subtree of an AST. For a variable initialization, there is no subtree of the AST that includes both the right side of an = and the ;. julia ___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
Re: [Cocci] Checking the parsing for a nested SmPL disjunction
>> <+... when any >> (t2 y = <+... >> ( sizeof(*(resource)) >> | >> * *(resource) >> ) ...+> > > You are missing a ; here. I assumed that I may omit the semicolon in such a SmPL code. Can the specification of a SmPL nest construct ever be sufficient here? Regards, Markus ___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
Re: [Cocci] Checking the parsing for a nested SmPL disjunction
On Fri, 26 Jun 2020, Markus Elfring wrote: > Hello, > > I have tried another small script variant out for the semantic patch language > (according to the software combination “Coccinelle 1.0.8-00131-g675b9670”). > > @display@ > type t1, t2; > expression action; > identifier resource, y; > statement is, es; > @@ > t1* resource; > <+... when any > (t2 y = <+... > ( sizeof(*(resource)) > | > * *(resource) > ) ...+> You are missing a ; here. julia > | > *action(..., resource, ...) > ) > ...+> > > > elfring@Sonne:~/Projekte/Coccinelle/janitor> spatch --parse-cocci > show_problematic_disjunction-20200626.cocci > … > minus: parse error: > File "show_problematic_disjunction-20200626.cocci", line 14, column 0, > charpos = 196 > around = '|', > whole content = | > > > May I expect that such SmPL disjunctions should also work? > > Regards, > Markus > ___ > Cocci mailing list > Cocci@systeme.lip6.fr > https://systeme.lip6.fr/mailman/listinfo/cocci >___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
[Cocci] Checking the parsing for a nested SmPL disjunction
Hello, I have tried another small script variant out for the semantic patch language (according to the software combination “Coccinelle 1.0.8-00131-g675b9670”). @display@ type t1, t2; expression action; identifier resource, y; statement is, es; @@ t1* resource; <+... when any (t2 y = <+... ( sizeof(*(resource)) | * *(resource) ) ...+> | *action(..., resource, ...) ) ...+> elfring@Sonne:~/Projekte/Coccinelle/janitor> spatch --parse-cocci show_problematic_disjunction-20200626.cocci … minus: parse error: File "show_problematic_disjunction-20200626.cocci", line 14, column 0, charpos = 196 around = '|', whole content = | May I expect that such SmPL disjunctions should also work? Regards, Markus ___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
Re: [Cocci] Checking the display of SmPL isomorphism results (with pointers)
On Fri, 26 Jun 2020, Markus Elfring wrote: > Hello, > > I have tried another tiny script variant out for the semantic patch language > (according to the software combination “Coccinelle 1.0.8-00131-g675b9670”). > > @display@ > expression* x; > statement is, es; > @@ > *if (!x) > is > else > es > > > elfring@Sonne:~/Projekte/Coccinelle/janitor> spatch --parse-cocci > show_null_pointer_checks.cocci > … > if (*!*x *!= *NULL) > … > if (*!*NULL *!= *x) > … > if (*x *!= *NULL *== *NULL) > … > if (*NULL *!= *x *== *NULL) > … > if (*NULL *== *x *!= *NULL) > … > if (*NULL *== *NULL *!= *x) > … > > > Now I find the shown lines in such a program output questionable. > Should these presentations of elements in SmPL disjunctions be shorter? It could probably be improved, but it is difficult to do something perfect. Still, I'm confused about the double comparison to NULL, so I will check on it. julia___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
[Cocci] Checking the display of SmPL isomorphism results (with pointers)
Hello, I have tried another tiny script variant out for the semantic patch language (according to the software combination “Coccinelle 1.0.8-00131-g675b9670”). @display@ expression* x; statement is, es; @@ *if (!x) is else es elfring@Sonne:~/Projekte/Coccinelle/janitor> spatch --parse-cocci show_null_pointer_checks.cocci … if (*!*x *!= *NULL) … if (*!*NULL *!= *x) … if (*x *!= *NULL *== *NULL) … if (*NULL *!= *x *== *NULL) … if (*NULL *== *x *!= *NULL) … if (*NULL *== *NULL *!= *x) … Now I find the shown lines in such a program output questionable. Should these presentations of elements in SmPL disjunctions be shorter? Regards, Markus ___ Cocci mailing list Cocci@systeme.lip6.fr https://systeme.lip6.fr/mailman/listinfo/cocci
