Shell Programming and Scripting

Hi,

I need to remove all lines from a file that do not start with numbers

For instance, if the first three characters on any line are not numbers, delete those lines

I've tried to do it with awk and it's not working, any ideas ?

Thanks

Is this a homework question?
Assuming that it's not, what is the practical business problem you are solving?

We get inventory files dumped from POS systems. They are supposed to be in this format- SKU;quantity;price;location

However, one particular store has added custom SKU's to their inventory, and the SKU's they created all contain A-Z and sometimes the ' character.

I can remove the lines with the ' character using tr and the octal code for '

However, the job that loads the inventory into the DB only accepts numerics for the first field, so it is failing due to the custom SKU's they have created.

That's why I need to do something like "if the first three characters of any line are A-Z delete that line" or even "if the first three characters of any line are not 0-9 delete that line" and I need it to do it for every line in the file...

Thanks

Code:

echo '1A23abn' | sed -n '/^[0-9]\{3\}/p'

Or with grep:

Code:

grep '^[0-9][0-9][0-9]' file > newfile

Thanks so much guys.

Doh, the grep one completely slipped my mind !