techee left a comment (geany/geany#3034) > @techee do you know what is the status here? Is it acceptable after fixing > the conflicts, or the performances issues and complexity is too high? I can > dive in as well, but I'd rather trust you if you have an opinion 😊
Well, for me the problems are still the same (related to all PEG parsers): 1. We can't easily check whether there's some infinite loop in the grammar definition (happened for TOML upstream in https://github.com/universal-ctags/ctags/issues/4096) 2. or if it leads to some heavy back-tracking making performance unusable for some specific syntax (for normal recursive descent parsers we kind of know the performance characteristics which should be more or less linear, for PEG it can differ based on what rules are being used for parsing a certain source code) 3. It is hard to detect where the problem is exactly in the grammar when we e.g. get a bug report from users that e.g. there is some performance problem. Will we be able to fix such issues? In particular, the performance of the kotlin parser is really bad for some files on low-end machines (see https://github.com/geany/geany/pull/3034#issuecomment-991862837). Maybe it is fixable by changing the grammar but which way should we modify it? I don't know. So I personally don't like it much but I don't want to be the one who blocks this PR if others have a different opinion. -- Reply to this email directly or view it on GitHub: https://github.com/geany/geany/pull/3034#issuecomment-3977770205 You are receiving this because you are subscribed to this thread. Message ID: <geany/geany/pull/3034/[email protected]>
