Re: reducing redundant lines, partially

On Sun, Nov 20, 2016 at 8:05 AM, Thierry Blanc [email protected]
[sed-users] <[email protected]> wrote:
> I have the output of a script in the form
>
> file1/full/path:<number>
> file1/full/path:<number>
> ....
> file2/full/path:<number>
>
> I want the output
>
> file1/full/path:<number>:<number> ...file2/full/path:<number>:<number> ...
>
> there might be one or many lines with the same file name.

#!/bin/sed -Ef
:a
N
s/(([^:\n]*)[^\n]*)\n\2(:.*)/\1\3/
ta
P
D

gets you all the `:number`s for a sequence appended on a single line,
did you want just the endpoints?  I'd do that with a pair of subs
before the `P`.