CODEWITHSUNDEEP
regexr
jkeirstead
jkeirstead

Reputation: 2961

Unexpected R regex evaluation

I'm working on some regular expressions using an online Regex tester (http://www.regexr.com) to help figure out the syntax. However when I try to evaluate the pattern in R, I get a different result from what I was expecting.

The code is:

grepl("Brent[^a-zA-Z]", c("Brent", "Brentwood", "Brent 01"))

and I'm expecting: TRUE, FALSE, TRUE which is what I get with the online tester. But R returns FALSE, FALSE, TRUE.

What am I missing here?

Upvotes: 1

Views: 62

Answers (1)

Lucas Trzesniewski
Lucas Trzesniewski

Reputation: 51430

The first string cannot match because [^a-zA-Z] has to consume exactly one character.

You can replace it with \b (word boundary): \bBrent\b.

And, if you want to allow strings like Brent01 (without the space before the 01), you can use lookaheads: \bBrent(?![a-zA-Z]) (but I don't know if this feature is available in R).

Upvotes: 2

Related Questions