How to split strings separated by many semicolons in R?

Question

My desire is to know the length of a certain text separated by ; which comes after any number. In the text named txt below, I don't want to consider the first two semicolons. To get the length, the ; comes after 6, 5 should be considered. I mean the code should lookbehind some number(s) to consider the appropriate ;.

library(stringr)
txt <- "A;B; dd (2020) text  pp. 805-806; Mining; exercise (1999), ee, p-123-125; F;G;H text, (2017) kk"

lenghths(strsplit(txt,";")) gives me 8. In my case, however, it should be 3. Any help is highly appreciated.

akrun · Answer 1 · 2020-05-29T19:43:02.610

1

We can use a regex lookaround to match a ; that succeeds a digit ((?<=[0-9])) and get the lengths

lengths(strsplit(txt, "(?<=[5-6]);", perl = TRUE))
#[1] 3

Or using str_count

library(stringr)
str_count(txt, '[5-6];') + 1
#[1] 3

edited May 29 '20 at 19:43

answered May 29 '20 at 19:26

akrun

874,273
37
540
662

Can it be possible to look behind the numbers or closing bracket `)`? – iGada May 29 '20 at 19:41
@Gadaa theere is no closing brakcet before the `;`in your example – akrun May 29 '20 at 19:42
Sure! I'm thinking in the presence of `)`. – iGada May 29 '20 at 19:43
@Gada e.g. in here `dd (2020) text pp. 805-806`. Are you looking for any `)` – akrun May 29 '20 at 19:44
@Gadaa Do you need `lengths(lapply(strsplit(txt, "(?<=[5-6]);", perl = TRUE), function(x) grep('\\)', x)))# [1] 3` – akrun May 29 '20 at 19:45
In some of my text, there is a case like `dd (2020) text 8(33)`. – iGada May 29 '20 at 19:47
@Gadaa which one should be matched – akrun May 29 '20 at 19:47
@Gadaa If it is to match the four digits, `lengths(lapply(strsplit(txt, "(?<=[5-6]);", perl = TRUE), function(x) grep('\\(\\d{4}\\)', x)))` – akrun May 29 '20 at 19:48

How to split strings separated by many semicolons in R?

1 Answers1