UNIX for Dummies Questions & Answers

hi

i have a file of the following format

123 abc def 2005-09-23 02:04:55
356 abc def 2005-09-23 09:04:55
123 abc def 2005-09-21 19:02:17
234 abc def 2005-09-23 09:04:55
356 abc def 2002-09-23 09:04:55
123 abc def 2005-09-23 09:04:55


i need to sort by the first column and then the combined 4th and 5th column
(4th and 5th column of date format)

i could achieve sorting by first column only i dont know how to proceed with the second requirement

123 abc def 2005-09-23 09:04:55
123 abc def 2005-09-23 02:04:55
123 abc def 2005-09-21 19:02:17

234 abc def 2005-09-23 09:04:55

356 abc def 2005-09-23 09:04:55
356 abc def 2002-09-23 09:04:55



can u please provide me with some pointers

Hi,
u can use the following command:
sort +n <filename> which sorts the file with n+1 field in a file.
ex:lets say u want to sort on second field ,then value of n has to be +1

thanks gaba

that would pretty well work for sort on single fields

and i had already obtained the result sorting on the first field

my question is regarding multiple sorts

and it is on the date field which i had provided

i am not sure how to proceed with.

Thanks Anyway

Maybe you are looking for something like this:

Code:

$ cat test.txt
1 2 3
1 1 5
1 1 4
$ sort -n -k1 -k2 -k3 test.txt
1 1 4
1 1 5
1 2 3

The '-k' option can be used to sort field by field and even by columns within a particular field. Use the '-t' option to specify a seperator if it is something other than whitespace.

Go through the man page of sort on your system anyway.

To sort date fields:

Code:

$ cat test.txt
09/10/2000
09/09/2000
08/10/1999
08/10/2001
$ sort -t'/' -k3 -k2 -k1 test.txt
08/10/1999
09/09/2000
09/10/2000
08/10/2001

Note: all dates are DD/MM/YYYY

thanks blowtorch -> that was a pointer to me

but solved it myself


awk '{print $1}' file | sort | uniq | while read line
do
grep ${line} file | awk '{ if ( $1=='${line}' ) print $4, $5 }' | \
sed -e 's/[-, ]/:/g' | sort -t":" | sed -e 's/:/-/;s/:/-/;s/:/ /' | while read sub
do
grep ${line} file | grep "${sub}" | awk '{ if ( $1=='${line}' ) print }'
done
done


any pointers on optimising it...

try this one ...

Code:

root_mybox:/tmp # cat infile
123 abc def 2005-09-23 02:04:55
356 abc def 2005-09-23 09:04:55
123 abc def 2005-09-21 19:02:17
234 abc def 2005-09-23 09:04:55
356 abc def 2002-09-23 09:04:55
123 abc def 2005-09-23 09:04:55
root_mybox:/tmp # sort -n +0 +3 +4 infile
123 abc def 2005-09-23 02:04:55
123 abc def 2005-09-23 09:04:55
123 abc def 2005-09-21 19:02:17
234 abc def 2005-09-23 09:04:55
356 abc def 2002-09-23 09:04:55
356 abc def 2005-09-23 09:04:55
root_mybox:/tmp #