|
|
|
|
|
|
by chubot
2005 days ago
|
|
|
Great info, thanks! I think derivatives could be very suited for a compact implementation of POSIX regular expressions. You need to handle unicode classes but not backreferences! Although it does seem more suited for functional languages for sure, whereas I basically only have a C runtime. Capturing might be an issue. I found this https://www.home.hs-karlsruhe.de/~suma0002/publications/posi... but I think it's actually being a bit pedantic, i.e. if "almost all POSIX implementations are buggy" then applications don't rely on that exact semantic (they probably rely on the buggy one, if anything ...) Maybe more relevant: http://www.home.hs-karlsruhe.de/~suma0002/publications/ppdp1... |
|
|