How do I avoid greping matching lines with -

Question

I'm trying to grab only matching lines from a file using grep. The problem is that, when for example I use the following expression:

grep -w AFFX-KIT-000088 textfile

it will return:

AFFX-KIT-000088-A
AFFX-KIT-000088

However, I only want the exactly matching expression AFFX-KIT-000088. I have already trying -F -E and quoting. The textfile where I need to grep the lines is very big and I can not modify it (this also means I need an answer that doesn't involve many steps as it would take forever to process).

Note:

The file is actually very big and has many columns (not only the one I'm using for grep).

Answer 1

grep -x 'AFFX-KIT-000088' file

The -x forces a match of a complete line.

You may also want to add -F as you are matching with a string, not a regular expression. This would possibly speed up the operation.

The -w option does not work here as - is not a "word character". The substring AFFX-KIT-000088 in AFFX-KIT-000088-A is therefore a complete word.

"Word characters" are characters matching [[:alnum:]_], i.e. alphanumerical characters and underscore.

---

In multi-column data, assuming tab-delimited fields and matching in column 3:

awk -F '\t' '$3 == "AFFX-KIT-000088"' file

That is, compare the 3rd field to a specific string.

When a condition like this lacks an associated action, the default action is to print the whole input record if the condition is true. If you just want to print the field's data:

awk -F '\t' '$3 == "AFFX-KIT-000088" { print $3 }' file