Dear experts,

A rather simple issue I cannot manage to solve.

My infile:
alpha
alphabet
alpha-beta
beta
bebeta
gamma
ga-gamma

I want to retrieve exact match only.

Output:
alpha
beta
gamma

I'm using the following loop, but unsuccessfully.

for word in alpha beta gamma;
do
grep -w ${word} infile
done

Ideally,I'm looking for something like /^${word}$/
but I don't know exactly how to define spaces, given a variable.

Any help would be highly appreciated.

Udiubu

double quote and curly brackets removal do the job

do
grep "^$word$" infile
done

nevermind

That works in situations where the word you are searching for takes up the whole line and there are not even any leading or trailing spaces.

If by 'exact match' you mean the string is not a substring but is a standalone word, then you need a more general approach for finding the whole word then you can also search for the null-string word boundaries. Most systems will refer to those as \< and \> for the start of a word boundary and the end of a word boundary respectively. So the following would find the word if it is whole, regardless of where on the line it is or how much of the line it takes up:

Some systems might require it to be written a little differently,

but that won't work with GNU grep

For plain string search use fgrep (or grep -F)!
The -x option is for full line match.
The standard grep would match gamma for both gamma and gam.a because the dot is an "any character".