Presto sql: presto extract substring for the last occurrence of character in string

I would like to extract the substring after the last occurrence of ref_button_id value in the string, in this example, string 'ref_button_id=pivot-rows5&ref_button_id=hhh-rows&' will return hhh-rows

I am using presto sql and my try was:

select reverse(split(split(reverse('ref_button_id=pivot-rows5&ref_button_id=hhh-rows&'), '=di_nottub_fer&')[1], '&')[2]) as ref_button_id

Which can get what I need but feel it's not ideal, is there a more formal workaround of this please?

You can try using regular expressions with regexp_extract_all:

-- sample data
with dataset(str) as (
    values ('ref_button_id=pivot-rows5&ref_button_id=hhh-rows&'),
           ('ref_button_id=hhh-rows&'),
           ('non-refs=123')
)

-- query
select try(reverse(regexp_extract_all(str, '(?<=ref_button_id=)[^&#]+'))[1])
from dataset;

Output: