Am Dienstag, 21. August 2007 schrieb Leen de Braal:

First is to replace the first (only the first) space in the line with a <TAB>. After that I want to delete a given number of characters that always begin with the same sequence, but may have different contents.

sed 's/^ /\t/;s/beginning.\{n\}//' works for me, where "beginning" ist the beginning sequence and n is the given number of characters minus the length of "beginning".

I used sed 's/^ /\t/' txtfile, but I do not get a tab.

That works for me.

Besides, this might do more then just the first space, I guess.

No, 's/ /\t/g' would do so. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org

On Tuesday 21 August 2007 01:51, Leen de Braal wrote:

Hi all,

Need some help with sed:

...

I used sed 's/^ /\t/' txtfile, but I do not get a tab. Besides, this might do more then just the first space, I guess.

Assuming you're using bash, you need to use the $'string literal with \t escapes' notation. The dollar sign and single quote marks enable the backslash escapes. If you want to use a newline, it needs to be preceded by a backslash, which will then need to be double, so a newline in a $'...' string will look like $'... \\\n ...'

Could someone shed some more light here?

-- L. de Braal

Randall Schulz -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org