R - Using grep and gsub to return more than one match in the same (character) vector element

Question

Imagine we want to find all of the FOOs and subsequent numbers in the string below and return them as a vector (apologies for unreadability, I wanted to make the point there is no regular pattern before and after the FOOs):

xx <- "xasdrFOO1921ddjadFOO1234dakaFOO12345ndlslsFOO1643xasdf"

We can use this to find one of them (taken from 1)

gsub(".*(FOO[0-9]+).*", "\\1", xx)
[1] "FOO1643"

However, I want to return all of them, as a vector.

I've thought of a complicated way to do it using strplit() and gregexpr() - but I feel there is a better (and easier) way.

score 6 · Accepted Answer · answered Nov 30 '12 at 12:27

6

You may be interested in regmatches:

> regmatches(xx, gregexpr("FOO[0-9]+", xx))[[1]]
[1] "FOO1921"  "FOO1234"  "FOO12345" "FOO1643"

answered Nov 30 '12 at 12:27

sebastian-c

15,057
3
47
93

Perfect, I didn't realise there was anything "built in". I should have read the help page for grep more closely though, I would have found this myself! -1 for me – Jim Bo Nov 30 '12 at 15:32

score 3 · Answer 2 · answered Nov 30 '12 at 13:00

3

xx <- "xasdrFOO1921ddjadFOO1234dakaFOO12345ndlslsFOO1643xasdf"
library(stringr)
str_extract_all(xx, "(FOO[0-9]+)")[[1]]
#[1] "FOO1921"  "FOO1234"  "FOO12345" "FOO1643"

this can take vectors of strings as well, and mathces will be in list elements.

answered Nov 30 '12 at 13:00

jem77bfp

1,270
11
13

score 2 · Answer 3 · edited Nov 30 '12 at 14:49

2

Slightly shorter version.

library(gsubfn)
strapplyc(xx,"FOO[0-9]*")[[1]]

edited Nov 30 '12 at 14:49

G. Grothendieck

254,981
17
203
341

answered Nov 30 '12 at 13:47

Wojciech Sobala

7,431
2
21
27

R - Using grep and gsub to return more than one match in the same (character) vector element

3 Answers3