oftware.com
> To: programm...@jsoftware.com
> Subject: Re: [Jprogramming] Regex vs I./E. for pattern matching
>
> note that there is an even faster version
>
> 10 timespacex '( I. ''CTAG'' E. DNA)'
> 2.656e_6 6016
> 10 timespacex
note that there is an even faster version
10 timespacex '( I. ''CTAG'' E. DNA)'
2.656e_6 6016
10 timespacex '( ''CTAG'' I.@:E. DNA)'
2.144e_6 6400
for Henry's example of
'CTAG*ACTA'
there can be some enhanced flexibility for using E.
('CTAG';'ACTA') I.@:E.each < DNA
one way to use t
If you something more than a simple match, say CTAG*ATTA, regex will be
easier and probably faster. As I understand, regex builds a state
machine for your pattern and then executes it.
Do note that you can separate the building of the machine from the
execution by using rxcomp:
pat =. rxcom
Yes?
E. is faster than rxmatches precisely because it is less powerful.
And note that you could also use ;: (but the trace operation is 5x
larger than you need and most people don't like making state
machines).
--
Raul
On Sun, Aug 16, 2015 at 2:09 AM, Jon Hough wrote:
> I recently went throu
Sorry I don't know regex, but I like to solve problems in simple J. So here'
are some steps to answer your problem. There are lots of ways to condense
this and many ways to present your result.
However, answering the question itself always seems like a start to me.
DNA=:
'CGATTGACTAGTCGATTGCTG
