mjhjmtu

I have a table_1

A D G
B E H
C F I

I print the column 1 of the above table using awk 'print $1' table_1 to get list_1

A
B
C

I want to use list_1 above to grep -f from table_2

A n m
B m n
C n m
D m n
E n m

to get table_3

A n m
B m n
C n m

but I want to do it using a one-liner, not needing to save an intermediate file of list_1.

How can I structure the command?

A trivial (but not bash-specific) variation on Cyrus’s comment:

awk 'print $1' table_1 | grep -f - table_2

which uses the widespread convention that a filename of -
means “read from the standard input”.

Use awk:

awk 'NR==FNRa[$1];next ($1 in a)' table_1 table_2

Here’s another way to use join. 
If the files are sorted, and table_2 really has only three columns,
you can use

join -o "2.1 2.2 2.3" table_1 table_2 > table_3

As Kusalananda says,
join is a program that exists specifically to combine two files
by matching values from one column of the first file
against values from one column of the second file. 
By default, if uses the first column of each file (you can override this). 
By default, it combines the matching lines, like this:

$ join table_1 table_2
A D G n m
B E H m n
C F I n m

The -o "2.1 2.2 2.3" says “output the first field of the second file,
and then the second field of the second file,
and then the first third of the second file
(but nothing from the first file¹)”. 
Unfortunately, there don’t seem to be any shortcuts, shorthands,
accelerators, or wildcards here;
there’s no way to say “output the entire line from the second file”. 
If the second file has many fields,
then the -o format will have to be very long.

This differs from the grep-based solutions (yours and mine)
in that it specifically matches the values from column 1 of table_1
against the values from column 1 of table_2,
and will not give you lines from table_2
that have an A in column 2 or column 3, or an AZ in column 1. 
join, like grep, recognizes a -i (--ignore-case) option,
so, if the first line of table_2 is a n m (with a lower-case a),
you will get that line in your table_3 output
if (and only if) you specify -i.

________
¹ Except for the first field of the first file —
which is the same as first field of the second file.

Assuming that both tables are sorted on the first column, and using a shell like bash or ksh93 that understands process substitution with <(...):

$ join <( awk ' print $1 ' table_1 ) table_2
A n m
B m n
C n m

If the tables are not sorted, then we need to sort them:

$ join <( awk ' print $1 ' table_1 | sort ) <( sort table_2 )
A n m
B m n
C n m

The join utility performs a relational INNER JOIN operation on the first column (by default) of the two files that you give it.

Redirect the output to a new file to create table_3:

$ join <( awk ' print $1 ' table_1 | sort ) <( sort table_2 ) >table_3