UNIX for Dummies Questions & Answers

ok so I have a list of names that end in either ot,om,oa.

So for example

DETOT
MANOA
DET0M
DET0A
MANOT
SEAOT

etc...

I want to be able to group this list by OT, OM, OA and have the output have some headers like this and be alphabatized and if possible be in colums instead of like below so three columns:

OT
-----
DETOT
MANOT
SEAOT

OA
----
DETOA
MAN0A

Thanks for any feedback!

Hi.

Assuming that the words can be any length, I think I would reverse the lines, sort them, then reverse them again just before printing. Check to see if you have command rev, then look at command sort ... cheers, drl

PS I think some zeros, "0", are mixed in with letter ohs, "O" in the sample data.

Thank you for the reply. Actually it will always be 5 charectors and it's all the letter O not zero's......

Is your scenerio still the best way? How do I column it out? i'm assuming some printf command but again I'm so new I'm not familiar on how to use it.

what about if I add a grep statement that greps out OT. Then I run it again to grep out OM and then again for OA....

How would I then go about printing each output into a column with a header?

Thanks!

Since you already know what groups you are expecting, you can probably skip the step of reversing the lines and just grep out the relevant bits thus:

Code:

#!/bin/sh
MYGROUPS="OT OM OA"
file_to_sort=$1
for group in $MYGROUPS
do
    echo "$group"
    echo "--"
    egrep "${group}\$" $file_to_sort | sort
    echo ""
done
# Catch any unexpected input (if you want that)
if egrep -v "`echo $MYGROUPS | sed 's/ /\|/g'`" $file_to_sort > /dev/null
then
    echo "*** UNEXPECTED ***"
    egrep -v "(`echo $MYGROUPS | sed 's/ /\|/g'`)" $file_to_sort
fi

Ok i'm really really new to this is there anyway you can break down your post?

This is what I understand.

is to echo each group and put a -- under it
then search for each group (OT OM OA) and I kinda get lost here ....

I'm so sorry that I need so much help!

by the way I did test it and it worked out perfect!!! Is there a way to make the outputs be side by side columns?
Also how can I change the headings of the groups? So I want to still group by OT OM OA but I want the headings of each to be something different?

Not a problem
First section is just to move the list of groups we are splitting into tp the top in case you want to change it later.
We then take a command line argument for the file to search over and store it.

Code:

#!/bin/sh
MYGROUPS="OT OM OA"
file_to_sort=$1

Next we perform the loop below for each group (ie once per item in $MYGROUPS)
The egrep (a version of grep that understand regular expressions better) is looking for lines that end in our group identifier (the '\$' means 'end of line') in the file. The list that comes out is then passed to sort to alphabetise it.

Code:

for group in $MYGROUPS
do
    echo "$group"
    echo "--"
    egrep "${group}\$" $file_to_sort | sort
    echo ""
done

The next part it completely optional, I just included it in case you might want to throw up a warning listing any lines that didn't match one of the groups you are looking for. You could add an 'exit 1' line just before the 'fi' at the end if you want to have it produce an error status when this happens too.

The first egrep is looking to see if there is any output left after grepping out (via the -v flag) any of the groups we are looking for.
It builds the regular expression based on the MYGROUPS variable by replacing the spaces with '|' symbols. This way it is doing an egrep -v "(OT|OM|OA)\$" over the file (ie list all output that doesn't end with Ot, OM or OA).
The first time is to see if we find any (it just throws it away), if so, it prints the header than does it again to the screen. It's a shame to do the search twice, you could store the output from the first run and just display it later but I don't like generating temporary files and was being a bit lazy

Code:

# Catch any unexpected input (if you want that)
if egrep -v "(`echo $MYGROUPS | sed 's/ /\|/g'`)\$" $file_to_sort > /dev/null
then
    echo "*** UNEXPECTED ***"
    egrep -v "(`echo $MYGROUPS | sed 's/ /\|/g'`)\$" $file_to_sort
fi

I've actually seen a couple of minor mistakes I made in that last part too - fixed in teh above version. The first grep was missing the brackets to group the regular expression and both greps were missing the \$ to indicate that it should be looking at the end of the line.