Using regex to selectively extract substrings in R

Question

Suppose I have the following strings:

string <- c(
  "DATE_OF_BIRTH_B1",
  "HEIGHT_BABY2",
  "WEIGHT_BABY_3",
  "OTHER_CONDITION_4",
  "OTHER_OPERATION_5"
)

How can I use regex in gsub() to extract:

Everything except the trailing underscore up until the number suffixes in the first three strings;
Nothing from the last two strings.

In other words, my expected gsub() output is:

"DATE_OF_BIRTH_B", "HEIGHT_BABY", "WEIGHT_BABY"

I managed to use gsub("(.+_B[A-Z]*)_?[0-9]", "\\1", string) to extract the desired substrings from the first three strings, but it failed to excluded the last two strings.

Could anyone help to correct and improve my regex, with a bit of explanation? Many thanks!

Do the strings you want to exclude have some kind of communality? Otherwise I don't see how you could exclude them generally. In your example you could e.g. just filter for OTHER. But something similar to that would need to be present — Julian_Hn
– Julian_Hn, Commented Oct 21, 2020 at 13:59
It is done using alternation with .+: sub("(.+_B[A-Z]*)_?[0-9]|.+", "\\1", string) — Wiktor Stribiżew
– Wiktor Stribiżew, Commented Oct 21, 2020 at 17:40

Paul · Accepted Answer · 2020-10-21 14:08:39Z

3

Remove OTHER or the suffix.

gsub("^OTHER.*|_?[0-9]+$", "", string)
#> [1] "DATE_OF_BIRTH_B"
#> [2] "HEIGHT_BABY"    
#> [3] "WEIGHT_BABY"    
#> [4] ""               
#> [5] ""

Or, if you specifically want capture groups, use a non-greedy capture.

gsub("(OTHER.*)?(.*?)_?[0-9]", "\\2", string)
#> [1] "DATE_OF_BIRTH_B"
#> [2] "HEIGHT_BABY"    
#> [3] "WEIGHT_BABY"    
#> [4] ""               
#> [5] ""

edited Oct 21, 2020 at 14:08

answered Oct 21, 2020 at 14:00

Paul

9,1571 gold badge31 silver badges38 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

elarry Over a year ago

Thanks very much for your solutions! Filtering for "OTHER" is a good tip, but @Wiktor Stribiżew's solution may be more universal?

Wiktor Stribiżew · Accepted Answer · 2020-10-21 17:49:06Z

1

If you expect gsub (or sub, usually, in this case, you really should use a sub since you only expect a single replacement operation) to return a result of the replacement or an empty string, you need to follow this technique:

sub("...(<what_you_want_to_extract>)...|.+", "\\1", x)

That is, your regex is before | alternation operator that is followed with .+ that matches any one or more chars as many as possible.

So, in your case, assuming your regex is just what you need and meets all your requirements, you can use

> res <- sub("(.+_B[A-Z]*)_?[0-9]|.+", "\\1", string)
> res
[1] "DATE_OF_BIRTH_B" "HEIGHT_BABY"     "WEIGHT_BABY"     ""                ""

If you need to remove empty items, just use

> res[nzchar(res)]
[1] "DATE_OF_BIRTH_B" "HEIGHT_BABY"     "WEIGHT_BABY"

answered Oct 21, 2020 at 17:49

Wiktor Stribiżew

631k41 gold badges502 silver badges632 bronze badges

1 Comment

elarry Over a year ago

Excellent! Thanks very much for the tips on not to capture anything in alternative strings!

Collectives™ on Stack Overflow

Using regex to selectively extract substrings in R

2 Answers 2

1 Comment

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related