I ask you for the help.

I have the file tel.txt :

0523842156
0524566478
0783333333
.
.
0448888888

I have so to transform this text ( using Sed)

(052)384-21-56
(052)456-64-78
(078)333-33-33
.
.
(044)888-88-88


I will be grateful for the help

Hi,

sed 's/^\(...\)\(.*\)/(\1)\2/'infile

This sed command uses back referencing and can be broken up into 2 parts:

^\(...\) => First 3 chars at the beginning of the line
\(.*\) => the rest of the line

the \( and \) are special and point to parts that can be used into the replacement part (back referencing). In the replacement part you can use these with \1 and \2.

Hope this helps.

Maybe using the advanced regex option and a form like \(\[[:digit:]]{3}\) would be more concise, but just counting dot's seems easier to read.

Thanks a lot

Is it guaranteed that the area/city code is always three digits?


Cheers,
Tink

Tinkster brings up a good point. It is a good idea to be exact on the input set. If there are other patterns, you can have a separate sed command for each possible pattern.

You could run "egrep -v '^[[:digit:]]{10}$' tel.txt. If all of the numbers are 10 digits, there should be not output.

A phone number in the US would look like (701)555-5555. Having a simple textfile of numbers assumes that they all use the same format, and locale. Even if this is the case, what happens if one of the entries has an error. You might want to add a rule to reject it, or to write the entry to a separate file.