[SOLVED] How to create a regex expression to get a substring between 2 pipes

How to create a regex expression to get a substring between 2 pipes

I have a dataset that I'm trying to work with where I need to get the text between two pipe delimiters. The length of the text is variable so I can't use length to get it. This is the string:

ENST00000000233.10|ENSG00000004059.11|OTTHUMG000

I want to get the text between the first and second pipes, that being ENSG00000004059.11. I've tried several different regex expressions, but I can't really figure out the correct syntax. What should the correct regex expression be?

Solution

Here is a regex.

x <- "ENST00000000233.10|ENSG00000004059.11|OTTHUMG000"
sub("^[^\\|]*\\|([^\\|]+)\\|.*$", "\\1", x)
#> [1] "ENSG00000004059.11"

^{Created on 2022-05-03 by the reprex package (v2.0.1)}

Explanation:

^ beginning of string;
[^\\|]* not the pipe character zero or more times;
\\| the pipe character needs to be escaped since it's a meta-character;
^[^\\|]*\\| the 3 above combined mean to match anything but the pipe character at the beginning of the string zero or more times until a pipe character is found;
([^\\|]+) group match anything but the pipe character at least once;
\\|.*$ the second pipe plus anything until the end of the string.

Then replace the 1st (and only) group with itself, "\\1", thus removing everything else.