AI-Feynman VS compile-time-regular-expressions

Oh I'd love to share it, I just can't. Well I made one that did make its way to Fortran by being independently developed, the Lisp version was in 2011 and the Fortran I think...2017. The brute forcer in AI-Feynman.

https://github.com/SJ001/AI-Feynman/blob/master/aifeynman/sy...

I talk about it in my portfolio, it's Guesser 1.0 and Guesser 2.0.

https://github.com/daniel-cussen/portfolio/blob/master/portf...

CTRE (https://github.com/hanickadot/compile-time-regular-expressions) ranges::views (filter, transform, etc.) (C++20) str.find() + str.substr() freopen to stdin + cin >> extraction Parser libraries

There are suggestions that should be replaced by the high performance ctre implementation: https://github.com/hanickadot/compile-time-regular-expressions

operator"" can be used in a similar way to expression templates (DSLs), where the type of the resulting expression is dependent on the string contents. For example ctre makes use of this to build efficient regular expression parsers, and kumi uses this in conjunction with operator[] to make tuple indexing quite elegant

Check out https://github.com/hanickadot/compile-time-regular-expressions anything is possible 😂

Yes to regex, no to std::regex. Better to use CTRE. Something like "^Hello [0-9]+ how are you" should allow checking if there's a match

You could compare your implementation with https://github.com/hanickadot/compile-time-regular-expressions and see if there are any ideas you can copy.

std::compile_time_regex would be a nice addition. Something similar to ctre https://github.com/hanickadot/compile-time-regular-expressions Simply letting the compiler generate all the regex parsing machinery at compile time.... And benefitting from compiler optimizations, vectorization, etc...

ctre