Bash: Extract cells from output formatted as table

Question

I am using TreeTagger (http://www.cis.uni-muenchen.de/~schmid/tools/TreeTagger/) to extract nouns from a text. My problem is that the output is formatted as such:

word    pos     lemma

The     DT      the 
TreeTagger      NP      TreeTagger 
is      VBZ     be 
easy    JJ      easy 
to      TO      to 
use     VB      use

with apparently no option to get nouns only ("NP" and "NN"). With bash, how could I get the cells in the first column that have "NP" or "NN" in the second column?

anubhava · Accepted Answer · 2017-02-07T15:40:04.733

2

You can use an awk for this:

awk '$2 ~ /^N[PN]$/{print $1}' file

TreeTagger

Regex /^N[PN]$/ will match either NP or NN

As @Cyrus rightly commented below, you can use alternation in your regex as:

awk '$2 ~ /^(NP|NN)$/ {print $1}' file

edited Feb 07 '17 at 15:40

answered Feb 07 '17 at 15:35

anubhava

761,203
64
569
643

1

or `awk '$2 ~ /^(NP|NN)$/ {print $1}' file` – Cyrus Feb 07 '17 at 15:38
That's great how do I redirect the nouns to an array instead of a file? – Sulli Feb 07 '17 at 15:48
For array population use: `arr=($(awk '$2 ~ /^N[PN]$/{print $1}' file))` – anubhava Feb 07 '17 at 15:50
1

yes thanks! I had a problem with your code but it's only because I had changed the IFS separator – Sulli Feb 07 '17 at 16:43

Bash: Extract cells from output formatted as table

1 Answers1