On Thu, Nov 08, 2007 at 03:42:33PM +0100, Rodolfo Medina wrote: > Rodolfo Medina <rodolfo.medina@gmail.com> [071106 06:42]: > > >> > In the documentation of `sort' I couldn't find the possibility of ignoring > >> > digits. I have a document that contains names and phone numbers and I > >> > wish to sort it by names. The `-k' option doesn't seem to help because > >> > names and numbers are not in definite fields. > > > > "Douglas A. Tutty" <dtutty@porchlight.ca> writes: > > > Rodolfo, you'll need to send us a sample couple of lines EXACTLY as they > > appear in your file so that someone who knows sort can help. > > > > You'll also need to tell us EXACTLY what you want the file to look like > > when all is said and done. What do you want to do with this? > > > > Here are a few lines from the file: > > 3938269241 320192481 Stan Laurel > 3939424701 0815605311 Oliver Hardy > 800151611 800822051 Harpo Marx > ABC Travel 0111111111 33811111111 > Acade;Paolo;;; 335111111 Paolo Acade > Acade;Pasquale;;; 335122211 Pasquale Acade > > I wish it to be sorted by letters (i.e. by names) - and also eliminate those > ugly semicolons! well the semicolons are easy: sed -i -e 's/^.*\; //' filename this (untested) should match any string at the beginning of a line endind with semi-colon/space and replace it with nothing. It will do it in place on the existing filename, so make a copy first, so you can confirm its behavior. You need, in my opinion, to get this data into a more uniform format to do much with it. And you need to decide a couple of things. Do you want the numbers first or second. Do you want to sort on The whole name? the last name? how do you want to handle things like ABC Travel? Also, are these lines representative of all the data? can you rely on the lines looking only like this and not having any other format? A
Attachment:
signature.asc
Description: Digital signature