Separate vs. RegExpr

Hello,

RegExpr (String) seems to be considerably faster than Separate (String) when parsing lager amounts of data (see attached patch). The only thing that´s keeping me from using it: it´s “missing” a Former Index output. Would be really nice to have.

SeparateVsRegExpr.v4p (10.3 kB)

Blimey thats quite a difference, maybe i should use regexpr more!

hmm. i wrote the seperate function as a quite long procedure running through each single charater. that was long before the plugin interface was available.

i bet there are more clever c# implementations out there.
e.g http://www.codeproject.com/Articles/9258/A-Fast-CSV-Reader
maybe it got the right properties to copy the Seperate (String) node.
anyone?

hm but keep in mind that the marshalling costs for strings are quite high ( the transition from unmanaged to managed and vice versa), so going the plugin route might not be the best way if you’re dealing with a large amount of string data which changes every other frame. if it stays constant it should be ok.

there is a minor fix in place now that at least stops separate from using energy while the inputs don’t change. see: https://discourse.vvvv.org/