[Cocci] Using the same replacement for different source code

[Cocci] Using the same replacement for different source code

[Markus Elfring]

Hello,

The following change specification gets accepted by the Coccinelle software.

@replacement@
expression x, y;
@@
(
-x + y
+compute(x, y)
|
-x * y
+compute(x, y)
)


I would appreciate if the specification of duplicate SmPL code
can be avoided also for such an use case.
Thus I have tried further code variants out for the semantic patch language.

@replacement@
expression x, y;
@@
(
-x + y
|
-x * y
)
+compute(x, y)


An error message is reported then.

elfring@Sonne:~/Projekte/Coccinelle/Probe> spatch --parse-cocci replacement_for_two_cases2.cocci
…
9: no available token to attach to


@replacement@
expression x, y;
@@
- \( x + y \| x * y \)
+compute(x, y)

elfring@Sonne:~/Projekte/Coccinelle/Probe> spatch --parse-cocci replacement_for_two_cases3.cocci
…
5: no available token to attach to


Can such a transformation approach ever work for similar source code?

Regards,
Markus
_______________________________________________
Cocci mailing list
Cocci@systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci

Re: [Cocci] Using the same replacement for different source code

[Julia Lawall]

[-- Attachment #1: Type: text/plain, Size: 1240 bytes --]



On Tue, 12 Nov 2019, Markus Elfring wrote:

> Hello,
>
> The following change specification gets accepted by the Coccinelle software.
>
> @replacement@
> expression x, y;
> @@
> (
> -x + y
> +compute(x, y)
> |
> -x * y
> +compute(x, y)
> )
>
>
> I would appreciate if the specification of duplicate SmPL code

It can't.  You have asked this many times before.

julia

> can be avoided also for such an use case.
> Thus I have tried further code variants out for the semantic patch language.
>
> @replacement@
> expression x, y;
> @@
> (
> -x + y
> |
> -x * y
> )
> +compute(x, y)
>
>
> An error message is reported then.
>
> elfring@Sonne:~/Projekte/Coccinelle/Probe> spatch --parse-cocci replacement_for_two_cases2.cocci
> …
> 9: no available token to attach to
>
>
> @replacement@
> expression x, y;
> @@
> - \( x + y \| x * y \)
> +compute(x, y)
>
> elfring@Sonne:~/Projekte/Coccinelle/Probe> spatch --parse-cocci replacement_for_two_cases3.cocci
> …
> 5: no available token to attach to
>
>
> Can such a transformation approach ever work for similar source code?
>
> Regards,
> Markus
> _______________________________________________
> Cocci mailing list
> Cocci@systeme.lip6.fr
> https://systeme.lip6.fr/mailman/listinfo/cocci
>

[-- Attachment #2: Type: text/plain, Size: 136 bytes --]

_______________________________________________
Cocci mailing list
Cocci@systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci

Re: [Cocci] Using the same replacement for different source code

[Markus Elfring]

I am looking for possibilities to reduce undesirable code duplication
also for applications of the semantic patch language.
How much common code can be put into SmPL constraints for this purpose
(when the usage of SmPL disjunctions shows limitations so far)?

Example:
@replacement@
binary operator bo = { +, * };
expression x, y;
@@
+compute(
         x
-        bo
+        ,
         y
+       )


Regards,
Markus
_______________________________________________
Cocci mailing list
Cocci@systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci

Re: [Cocci] Using the same replacement for different source code

[Julia Lawall]

On Thu, 14 Nov 2019, Markus Elfring wrote:

> I am looking for possibilities to reduce undesirable code duplication
> also for applications of the semantic patch language.
> How much common code can be put into SmPL constraints for this purpose
> (when the usage of SmPL disjunctions shows limitations so far)?
>
> Example:
> @replacement@
> binary operator bo = { +, * };
> expression x, y;
> @@
> +compute(
>          x
> -        bo
> +        ,
>          y
> +       )

I can't figure out what this is an example of?  What do you want to do?
Does this work or not?

julia

>
>
> 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

Re: [Cocci] Using the same replacement for different source code

[Markus Elfring]

>> @replacement@
>> binary operator bo = { +, * };
>> expression x, y;
>> @@
>> +compute(
>>          x
>> -        bo
>> +        ,
>>          y
>> +       )
>
> I can't figure out what this is an example of?

It shows just a possible transformation specification where a SmPL constraint
is applied instead of a related SmPL disjunction.

See also:
https://lore.kernel.org/cocci/59c2f820-af2b-030f-3f8e-1dc0fd23f4ba@web.de/
https://systeme.lip6.fr/pipermail/cocci/2019-November/006582.html


> Does this work or not?

This bit of SmPL code gets parsed by the Coccinelle software as expected.

Does it demonstrate similar development opportunities?

Regards,
Markus
_______________________________________________
Cocci mailing list
Cocci@systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci

Re: [Cocci] Using the same replacement for different source code

[Julia Lawall]

On Thu, 14 Nov 2019, Markus Elfring wrote:

> >> @replacement@
> >> binary operator bo = { +, * };
> >> expression x, y;
> >> @@
> >> +compute(
> >>          x
> >> -        bo
> >> +        ,
> >>          y
> >> +       )
> >
> > I can't figure out what this is an example of?
>
> It shows just a possible transformation specification where a SmPL constraint
> is applied instead of a related SmPL disjunction.
>
> See also:
> https://lore.kernel.org/cocci/59c2f820-af2b-030f-3f8e-1dc0fd23f4ba@web.de/
> https://systeme.lip6.fr/pipermail/cocci/2019-November/006582.html
>
>
> > Does this work or not?
>
> This bit of SmPL code gets parsed by the Coccinelle software as expected.
>
> Does it demonstrate similar development opportunities?

OK, so we have an example that works.  What doesn't work that you are
asking about?

julia
_______________________________________________
Cocci mailing list
Cocci@systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci

Re: [Cocci] Using the same replacement for different source code

[Markus Elfring]

> OK, so we have an example that works.

Probably not only for a simple parsing test.


> What doesn't work that you are asking about?

We got different development views around applications of SmPL disjunctions
for source code replacements.

* Which open issues will still be reconsidered here?

* Will it become more interesting then to move special stuff into SmPL constraints?

Regards,
Markus
_______________________________________________
Cocci mailing list
Cocci@systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci

[Cocci] Using the same replacement for different source code

[Markus Elfring]

Hello,

The following change specification gets accepted by the Coccinelle software.

@replacement@
expression x, y;
@@
(
-x + y
+compute(x, y)
|
-x * y
+compute(x, y)
)


I would appreciate if the specification of duplicate SmPL code
can be avoided also for such an use case.
Thus I have tried further code variants out for the semantic patch language.

@replacement@
expression x, y;
@@
(
-x + y
|
-x * y
)
+compute(x, y)


An error message is reported then.

elfring@Sonne:~/Projekte/Coccinelle/Probe> spatch --parse-cocci replacement_for_two_cases2.cocci
…
9: no available token to attach to


@replacement@
expression x, y;
@@
- \( x + y \| x * y \)
+compute(x, y)

elfring@Sonne:~/Projekte/Coccinelle/Probe> spatch --parse-cocci replacement_for_two_cases3.cocci
…
5: no available token to attach to


Can such a transformation approach ever work for similar source code?

Regards,
Markus
_______________________________________________
Cocci mailing list
Cocci@systeme.lip6.fr
https://systeme.lip6.fr/mailman/listinfo/cocci