Dear all.

I want to achieve the following:

1/grep all the lines starting with id 35.64 for example, from a log file:

grep [3][5][.][6][4] file.log

result:

35.64 325323666 20050818
35.64 325325666 20050818
35.64 325327666 20050818
35.64 325328666 20050818

2/Once I get all the lines, parse them and fill values in a chart:

id user SID timestamp
-----*-----*----*----------*
35.64 325323 666 20050818
35.64 325325 666 20050818
35.64 325327 666 20050818
35.64 325328 666 20050818

Would also be nice to be able to sort the fields by timestamp or user/

Thanks for any input@

merou

Hi,

Try this:

echo "id user SID timestamp" > output.txt
echo "==== ==== ==== =======" >> output.txt
grep ^35.64 file.log | sort -t 4 >> output.txt

Thanks Vicchai,

But how do I deal with records that i want to split (space as delimiter) but belong to the same column:

Example:

from there:
35.64 325323666 20050818

to there:
35.64 325323 666 20050818

I was thinking of using the substr function in awk, but maybe there is something more simple.

Thanks.

Even substr is no good, because it is dealing with columns.

I just want a function that split a string(record), and I specify how I want to split it.

example:

0-5 : ID
6-12 : user
.......
.......

Thanks.