CODEWITHSUNDEEP
rregexstringstringr
guiotan
guiotan

Reputation: 170

How to keep only occurrences that end with a specific letter?

I would like to keep all occurrences which end only with a specific letter (lets say "a") from a string. I am not dealing with a vector composed of different elements but rather a string whose occurrences are separated by spaces.

Here are the data:

 have="5a 4a 8a 10a 3a 5m 10m 7a 8p 11s 5s 4h 24h"
 want="5a 4a 8a 10a 3a 7a"

Here are some codes:

gsub("([A-Z]|[0-9])([m|p|h|s])","", have)
gsub("\\w+m|p|h|s *", "", have)

After applying one of these gsub codes, I get 2 types of info (alphanumeric followed by "a" and purely figures).

But I still need to clean so I get only the occurrences ending with "a". Whould you have any idea?

Upvotes: 1

Views: 114

Answers (4)

The fourth bird
The fourth bird

Reputation: 163517

You might use this pattern and replace with an empty string:

[ ]?[a-z0-9]+[mphs]
  • [ ]? Optional space (the square brackets are only for clarity)
  • [a-z0-9]+ Character class, match 1+ times a-z 0-9
  • [mphs] Character class, match m, p h or s

For example

have="5a 4a 8a 10a 3a 5m 10m 7a 8p 11s 5s 4h 24h"
gsub(" ?[a-z0-9]+[mphs]","", have)

Result

[1] "5a 4a 8a 10a 3a 7a"

Regex demo | R demo

Perhaps you could match them instead:

\b\da\b
  • \b Word boundary
  • \d a digit
  • a match a
  • \b Word boundary

Regex demo

Note that in the character class [m|p|h|s] the | does not mean or but a | char and can also be written as [mphs|].

Upvotes: 2

NelsonGon
NelsonGon

Reputation: 13319

You can do:

  trimws(gsub("([A-Z]|[0-9]{1,})([b-z])","",have))
[1] "5a 4a 8a 10a 3a   7a"

To remove the extra space:

gsub("\\s{2,}"," ",
      trimws(gsub("([A-Z]|[0-9]{1,})([b-z])","",have)))
#[1] "5a 4a 8a 10a 3a 7a"

Upvotes: 3

Sharon
Sharon

Reputation: 3786

Or, much longer but easier regexp, turn it into a vector and then turn it back to a string.

have_string <- "5a 4a 8a 10a 3a 5m 10m 7a 8p 11s 5s 4h 24h"
have_vector <- unlist(strsplit(have_string," "))
library(stringr)
want_vector <- have_vector[str_detect(have_vector, ".*?a$")]
want_string <- paste(want_vector, sep = " ", collapse = " ")

Upvotes: 2

G5W
G5W

Reputation: 37661

You can split it into words, use grep to identify the words ending in a, then paste them back together. 

Words = strsplit(have, "\\W+")[[1]]
paste(grep("a$", Words, value=T), collapse=" ")
[1] "5a 4a 8a 10a 3a 7a"

Upvotes: 2

Related Questions