in reply to sort csv file on "date" column

$ echo '"Add workflow, Delete workflow",resource1,Sat 04/25/09,manager
+1
"Updateworkflow, Delete workflow",resource2,Wed 04/15/09,manager2
"Retrieve workflow, Delete workflow",resource3,Sat 04/25/09,manager3
"Add workflow, Delete workflow",resource4,Wed 03/25/09,manager4
"Add workflow",resource5,Tue 04/14/09,manager5' | \
    perl -e'
        print for
            map substr( $_, 6 ),
            sort
            map join( "", ( m<,\w{3} (\d\d)/(\d\d)/(\d\d),> )[ 2,0,1 ]
+, $_ ),
            <>
'
"Add workflow, Delete workflow",resource4,Wed 03/25/09,manager4
"Add workflow",resource5,Tue 04/14/09,manager5
"Updateworkflow, Delete workflow",resource2,Wed 04/15/09,manager2
"Add workflow, Delete workflow",resource1,Sat 04/25/09,manager1
"Retrieve workflow, Delete workflow",resource3,Sat 04/25/09,manager3

[download]

Replies are listed 'Best First'.
Re^2: sort csv file on "date" column
by mbm (Novice) on Apr 25, 2009 at 16:51 UTC 
    Hi Jwkrahn,

 thanks a lot. The code you proposed looks so simple and doing the wor
+k so nicely. Thanks again.
Can you please describe the logic here if you dont mind?

Regards,
Manjunath
Alcatel-Lucent

    [download]
[reply]
[d/l]

      The logic is to rearrange the date field (for example: 03/25/09) from month-day-year to year-month-day so it can be sorted by the built-in sort function without modification.   The algorithm employed is the GRT (Guttman-Rosler Transform) which prepends the reformatted date to the current record, sorts the current record, and then removes the reformatted date from the beginning of the current record.

[reply]
[d/l]
        ok... I wanted to know what the numbers 6, [2,0,1] actually are. It he
+lps me to customize the algorithm you provided if the date field is i
+n other column.

~Manjunath

        [download]
[reply]
[d/l]

In Section Seekers of Perl Wisdom