NagyDonat wrote:

> If this is an educational checker, should we rename this to 
> NaiveStreamChecker?

I think `Simple` is better at describing our intentions: we want this checker 
to be simple, so it also ends up being naive. 

Actually I wouldn't say that "naive" is a good word for it: IIRC 
`SimpleStreamChecker` is reasonably accurate at modeling a small set of the 
most common stream functions, and the advantage of `StreamChecker` is that it 
is also modeling many other less common functions. (And maybe has some 
inelegant, but practical heuristics? I'm only speaking from old memories, we 
should definitely review the contents of the two source files before making a 
final decision.)

If you think that `Simple` is not clear enough, I would suggest `Simplified`, 
`Abbreviated`, `Condensed` or similar adjectives (instead of `Naive`).

Also note that a name like `AdjectiveStreamChecker` would be parsed as 
(_adjective_ + "stream") + "checker" by most readers (because most checkers are 
named as (_something_ + _something_) + "checker"), while we want to state a 
property of the _checker_. To avoid this misunderstanding, perhaps we should 
choose a name like `StreamCheckerSimplified.cpp` or 
`StreamCheckerPrototype.cpp` – which is unfortunately a bit awkward.

By the way, I'm not sure whether the educational nature of this checker was 
actually useful for anyone. Perhaps we should survey the CSA contributors and 
if nobody found this useful, then we should just delete it. (I, for one, 
vaguely remember that this was mentioned as a code example when I first came 
into contact with CSA in 2015 as an intern, but even then I didn't actually 
follow it as an example. If I had to show a checker as an example to an intern, 
I would probably pick another checker that is more modern [was refactored 
recently] and that I'm more familiar with.)

https://github.com/llvm/llvm-project/pull/230288
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to