Re: [R] A question on regular expression

2019-09-13 Thread Jeff Newmiller
Regular expressions are in much more widespread use than merely R... and there are correspondingly more resources for learning than just R-help. Please do make use of them. Here are a couple that googling "regex character set carat" found: https://www.regular-expressions.info/charclass.html htt

Re: [R] A question on regular expression

2019-09-13 Thread Christofer Bogaso
A quick question. Could you please explain the -- [^}]* -- part in finding the pattern? On Fri, Sep 13, 2019 at 12:19 AM Bert Gunter wrote: > > > You can't use the same regex for str_extract_all as I used for sub (or gsub, > which is what is required here)! If you do this sort of thing a lot, y

Re: [R] A question on regular expression

2019-09-12 Thread Christofer Bogaso
Awesome, thanks! On Fri, Sep 13, 2019 at 12:19 AM Bert Gunter wrote: > > > You can't use the same regex for str_extract_all as I used for sub (or gsub, > which is what is required here)! If you do this sort of thing a lot, you > *must* learn more about regex's. > > Anyway, this will do what you

Re: [R] A question on regular expression

2019-09-12 Thread Bert Gunter
You can't use the same regex for str_extract_all as I used for sub (or gsub, which is what is required here)! If you do this sort of thing a lot, you *must* learn more about regex's. Anyway, this will do what you want I think: z <- paste("ab{cd$ }ed", "ab{cad$ }ed", collapse = " ") ## just for r

Re: [R] A question on regular expression

2019-09-12 Thread Christofer Bogaso
Thanks Bert, This works, but if in my text there are more than one patterns then fails to generate desired result. library(stringr) str_extract_all(paste("ab{cd$ }ed", "ab{cad$ }ed", collapse = " "), ".*(\\{.*\\}).*") This generates below - [[1]] [1] "ab{cd$ }ed ab{cad$ }ed" I was expecting I

Re: [R] A question on regular expression

2019-09-12 Thread Bert Gunter
> sub(".*(\\{.*\\}).*", "\\1","ab{cd$ }ed") [1] "{cd$ }" Use ".+" instead of ".*" within the {} if you don't want to return empty {}'s. You might wish to use the stringr package for string matching and manipulation, as it provides a more user friendly and consistent interface to these tasks. Be

[R] A question on regular expression

2019-09-12 Thread Christofer Bogaso
Hi, I am wondering on what is the correct way to select a pattern which goes as - {"(any character with any length)"} The expressions " {" " and " "} " both are included in the pattern. For example, the lookup of the above pattern in the text " {"asaf455%"}57573blabla " will result in {"asaf455