About 'grep -f': match pattern with file

Question

I am using the grep -f function to extract lines from a file which match a particular pattern. Let's say my pattern file is pattern.txt, as follows.

1
2
3
4
5

And the file against which I am matching this pattern is file.txt,

1::anv
2::tyr
3::yui
4::fng
5::gdg
6::ere
7::rer
8::3rr
9::gty

Now when I do a grep -f pattern.txt file.txt, I am getting this ->

1::anv
2::tyr
3::yui
4::fng
5::gdg
8::3rr

The last line in the output above, is causing my problem. How do I modify this grep command so as to get the output (showing correct correspondences) as follows?

1::anv
2::tyr
3::yui
4::fng
5::gdg

Answer 1

Add ^ before your patterns so that grep will match a line start with your pattern. If your pattern is actually set of numbers, You don't need a list of patterns. Just use ^[1-5]:.

Answer 2

Well, there are lots of possible solutions. You are telling grep to find all lines containing a 3, and the line '8::3rr' does in fact contain a 3. so you need to be more specific in what you're looking for.

For example, you can change the patterns to '1:', '2:', etc. to match only numbers followed by a colon. Or you could change them to '^1', '^2', etc. to match only numbers at the beginning of the line. It depends on your data, but probably you want both, so that your search for '1:' doesn't match '21:' and your search for '^5' doesn't match '53:'. In which case your patterns file would like like this:

^1:
^2:
^3:
^4:
^5: