Hi, this is my first post so go easy...

I have two database files files called customer and vehicle. Basically they look slightly like this

CUSTOMER:

NUMBER 1
NAME John
SURNAME Stevenson
Age 28

VEHICLE:

MAKE Audi
Model TT
NUMBER 1
AGE 4

I need to write a script that searches the vehicle database for a make and model and then retrieves the customer number and searches (then lists) details from the customer file. If possible i'd like details listed on one like like this:

NAME SURNAME AGE
John Stevenson 28

What ive got so far is the customer number in a temp file, I was thinking I could somehow assign the contents as a variable using awk and then search through customer with this variable... Harder than it sounds. I imagine I'll probably need combinations of cat, grep, awk and maybe sed.

Sorry if this is a really basic question, I've searched the web a lot for awk tutorials but 80% of the stuff I've read is EXTREMLEY difficult to understand.



Thankyou SOOOOO much in advance

I'd be curious to see how multiple recs in each file are laid out, but to be honest I'd HIGHLY recommend using a DB eg MySql or it's going to be pretty complicated and slow, and you'd have to write a fair new chunk of code for each new type of 'select' or 'update/insert/delete'.
A possible alternative is to use Perl hashes.
HTH

It's skoolwork and were only allowed to use bash... BUT I managed to sort it, I'm just having some problems outputting line numbers to a variable and outputting everything to one line.

I'll probably set up a new thread or something,

If your interested i did this:

File=/media/floppy/temp
{
read CUST
} < $File

cat /media/floppy/customer | grep "$CUST" -A4 > temp

awk '/surname/ {print $2}' temp


Theres lines missing inbetween but essentially these are the crucial lines. As you can probably tell it can only deal with one record Im tryna make it count lines and then read based on the ammount of lines it counts and assign each one of them a variable.

Thanks anyway, I really appreciate it

P.S
your completelly 100% right it would of been much easier doing this in... well anything but bash basically. I got some help and someone I know managed to do the whole thing in like 5 minutes with perl.

How about coverting them to flattened one line records and using the join
command?

just a thought.

Thanks for the idea but I thought you needed 2 files for join. I've got past the comparison bit of my code I'm just working on displaying it now.

Thanks a lot anyway mate

details, details ....

I cant believe we've been asked to do this work, its a joke! The project is so complicated and we cant use switch statements. Worse of all our lecturer doesnt even teach programming, teaches cisco! Here I am writing a 100+ line code... wheres that angry smiley!!!

I'll paste my code (or at least that bit) if your really interested but its my first ever bash script as a result its EXTREMELY complex and I imagine at least 3/4 of it is unnecessary.

Thanks for all your help once again people and the speedy reply. Lemmie know if you sincerely are interested in my incredibly complex code...

By the end of this you'll be a very good bash shell guy then ...

Don't knock yerself out.
Lecturers never read the bloody
code anyway.

You do a lovely structured bit of work and you'll
get the same grade as some git who copied from his friend!