Grep template for extracting lines where second word has only two vowels

Question

For example, i have file with content

hello world
it's nice to see you
amazing night
what a wonderful day
my name is Robert
still breathing
speaking bottom soul
something wrong

I need to match those lines, in which second word has exactly two vowels. So the output should be:

it's nice to see you
my name is Robert
speaking bottom soul

How can I do this using grep?

@thrig, i don't have much experience in working with regular expressions, so the last thing i tried grep -e ".*\s[aueioy]\{2\}" test.txt. BTW, it's still wrong. — Maksym Musiienko, Dec 10 '16 at 17:07
Does it have to be grep? it would be simpler and clearer IMHO using a tool such as awk or perl that handles whitespace delimited fields more naturally. For example in awk you could use the return value of a gsub on $2 (or in perl use the return value of a tr on $F[1]). — steeldriver, Dec 10 '16 at 18:34

glenn jackman · Accepted Answer · 2016-12-10 18:49:42Z

grep with extended regular expressions:

grep -iE '^[^[:blank:]]+[[:blank:]]+([^aeiou]*[aeiou]){2}[^aeiou]*\>' file

grep with pcre

grep -iP '^\S+\s+([^aeiou]*[aeiou]){2}[^aeiou]*\b' file

perl (honestly, I did this independent of steeldriver's comment)

perl -ane 'print if (lc($F[1]) =~ tr/aeiou/aeiou/) == 2' file

awk

awk '{col2 = tolower($2); gsub(/[aeiou]/,"",col2)} length($2) - length(col2) == 2' file

1 Answer 1