Any good courses / exercise sets out there to get comfortable with modern vectorized instruction sets?
I did some coding for the Cell processor and CUDA programming back in ~2008, but it's been a while...
Any good courses / exercise sets out there to get comfortable with modern vectorized instruction sets?
I did some coding for the Cell processor and CUDA programming back in ~2008, but it's been a while...
@d6 how slow is "very slow"? Would it realistically be able to handle equality testing of small variations to patterns like this?
Nosey Parker is a command-line tool that finds secrets and sensitive information in textual data and Git history. - noseyparker/crates/noseyparker/data/default/builtin/rules/phpmailer.yml at 2e6e7f36ce36619852532bbe698d8cb7a26d2da7 · praetorian-inc/noseyparker
GitHub (github.com)
@pervognsen is that page discussing a slightly different problem though? That PSPACE-complete problem discussed there is to determine if the intersection of DFAs is empty. This seems like a stronger property than whether a particular input string is accepted by all the DFAs (how you might imagine implementing intersection in a DFA-based regex engine). Or am I overlooking something?
@pervognsen right! At the top level, it's straightforward to use boolean operations in the host language, possibly with multiple calls to regex operations, to approximate complement / intersection. No change in algorithmic complexity doing that
Are there any regex implementations out there that support intersection and complement operators in the regex syntax itself?
(These operators appear to have been omitted in regular expressions by historical accident, tracing back to a paper from Kleene in 1951, where he skipped them because they added no expressive power to the formalism he was working on)