Hello,

I know how to get a solution of a filtering problem with AWK, but i need change default field separator by a regular expression. awk -F "xxx"

What i want to do is take a stranza as a field, so the field separator must be a EMPTY line. Does anybody how to do that??

a example:

blablablablablablablablablablablablablablablabla
blablablablablablablablablablablablablablablabla
blablablablablablablablablablablablablablablabla

blablablablablablablablablablablablablablablabla
blablablablablablablablablablablablablablablabla
blablablablablablablablablablablablablablablabla

blablablablablablablablablablablablablablablabla
blablablablablablablablablablablablablablablabla
blablablablablablablablablablablablablablablabla

I want to take as a field each one of those three stranzas.

Thank you very much.
Told me I look good

I couldn't understand your second post. Here is your original post, translated into English:

Greetings,

I need to use SED to create a tool to make my job easier. The objective is to filter a log to show me only those entries that contain a certain string. The format of each entry is as follows:

a) an entry may contain several lines
b) they always begin with a date

Before now I've been able to use SED to insert a blank line between entries: cat registro.log | sed '/^2009-11-06/{x;p;x;}' (I replace the date with a variable obtained with the date command)

With this I've been able to insert a blank line before each log entry so that there is now a way to know when each log entry begins and ends, but now I need to show only those entries that contain a certain string. Can someone help me? Thank you very much.