On Thu, 17 May 2018, SF Markus Elfring wrote:
> >> I would suggest to proceed carefully...
> > In each rule that contains S1, you first could replace the S1 by e;
> > where e is an expression metavariable.
> Please reconsider this suggestion.
> * I guess that there can statements occur which are not expressions.
> * Should return values trigger any more software development concerns?
I don't think you understood the idea at all. The rule will likely make
hundreds or throusands of changes. There is little risk in the case of
e;. Getting all of those cases out of the way first will reduce the size
of the result that has to be checked carefully.
> > Then you could commit that, and then run the original rule,
> > thus being able to focus on the more complex results.
> The application of additional SmPL script variants can help.
> I am curious on how the fine-tuning of corresponding implementation details
> will evolve further.
Cocci mailing list