0

I have many strings that all have the following format:

mystrings <- c(
  "(ABFUHIASH)THISISAVERYLONGSTRINGWITHOUTANYSPACES(ENDING)",
  "(SECONDSTR)YETANOTHERBORINGSTRINGWITHOUTSPACES(RANDOMENDING)", 
  "(JOWERIC)THISPARTSHOULDNOTBEEXTRACTED(GETTHIS)", 
  "(CAPTURETHIS)IOJSDOIOIADSNCXZZCX(IJFAI)"
)

I need to capture the strings that are inside parentheses both at the start and the end of the original mystrings.

Therefore, variable start will store the starting characters for each of the above strings with the same index. The result will be this:

start[1]
ABFUHIASH

start[2]
SECONDSTR

start[3]
JOWERIC

start[4]
CAPTURETHIS

And similarly, the ending for each string in mystrings will be saved into end:

end[1]
ENDING

end[2]
RANDOMENDING

end[3]
GETTHIS

end[4]
IJFAI

Parentheses themselves should NOT be captured.

Is there a way/function to do this quickly in R?

I have tried stringr::word and stringi::stri_extract, but I am getting very strange results.

motapekog
  • 13
  • 2

2 Answers2

2

We can use the stringr library for this. For example

library(stringr)
mm <- str_match(mystrings, "^\\(([^)]+)\\).*\\(([^)]+)\\)$")
mm

The match finds the stuff between the parenthesis at the beginning and end of the string in capture groups so they can be easily extracted.

It returns a character matrix, and you seem to just want the 2nd and 3rd column. mm[,2:3]

     [,1]          [,2]          
[1,] "ABFUHIASH"   "ENDING"      
[2,] "SECONDSTR"   "RANDOMENDING"
[3,] "JOWERIC"     "GETTHIS"     
[4,] "CAPTURETHIS" "IJFAI"
MrFlick
  • 195,160
  • 17
  • 277
  • 295
  • Wow, brilliant, some real regex witchcraft there. I would have never thought such a complicated regex would be required for this task. No wonder my rudimentary regexes (such as `\\(*\\)` were failing all the way to even capture the start! – motapekog Jan 31 '18 at 19:32
0

Something like this might work for you:

> regmatches(mystrings,gregexpr("\\(.+?\\)",mystrings))
[[1]]
[1] "(ABFUHIASH)" "(ENDING)"   

[[2]]
[1] "(SECONDSTR)"    "(RANDOMENDING)"

[[3]]
[1] "(JOWERIC)" "(GETTHIS)"

[[4]]
[1] "(CAPTURETHIS)" "(IJFAI)"

E.g., to extract endings you could:

lapply(x,tail,1)
Brian Davis
  • 990
  • 5
  • 11