Am 17.11.19 um 19:36 schrieb Markus Elfring: >>> It was chosen to transform source code fragments (pointer expressions) >>> by two SmPL rules so that the search pattern “sizeof(T)” would work >>> in the third rule. >> >> Ah, right, it would be nice to get rid of those normalization rules, >> especially the second one. > > Thanks for such positive feedback. > > >> I don't see how, > > Where is your view too limited at the moment? I don't know. Or perhaps it's rather too wide and I worry about irrelevant details? >> though, without either causing a combinatorial explosion > > Growing combinations can become more interesting, can't they? Perhaps, but not if they blow up the size of the semantic patch or the runtime of Coccinelle. René
