Separating Awk input in Unix

Question

I am trying to write an Awk program that takes two dates separated by / so 3/22/2013 for example and breaks them into the three separate numbers so that I could work with the 3 the 22 and the 2013 separately.

I would like the program to be called like

awk -f program_file 2/23/2013 4/15/2013

so far I have:

BEGIN {
d1 = ARGV[1]
d2 = ARGV[2]
}

This will accept both dates, but I am not sure how to break them up. Additionally, the above program must be called with nawk, with awk says it cannot open 2/23/2013.

Thanks in advance.

score 2 · Answer 1 · answered Mar 22 '13 at 13:52

you cannot do it in your way. since awk thinks you have two files as input. that is, your date strings were looked as filenames. That's why you got that error message.

if the two dates are stored in shell variables, you could:

awk -vd1="$d1" -vd2="$d2" BEGIN{split(d1,one,"/");split(d2,two,"/");...}{...}'

the ... part is your logic, in the line above, the splitted parts are stored in array one and two. for example, you just want to print the elements of one:

kent$  d1=2/23/2013
kent$  d2=4/15/2013
kent$  awk -vd1="$d1" -vd2="$d2"  'BEGIN{split(d1,one,"/");split(d2,two,"/"); for(x in one)print one[x]}'
2
23
2013

or as other suggested, you could use FS of awk, but you have to do in this way:

kent$  echo $d1|awk -F/ '{print $1,$2,$3}'         
2 23 2013

if you pass the two vars in one short, the -F/ won't work, unless they(the two dates) are in different lines

hope it helps

Basile Starynkevitch · Answer 2 · 2013-03-22T13:49:18.023

0

You could decide to use / as a field separator, and pass -F / to GNU awk (or to nawk)

edited Mar 22 '13 at 13:49

answered Mar 22 '13 at 13:43

Basile Starynkevitch

223,805
18
296
547

score 0 · Answer 3 · answered Mar 22 '13 at 13:46

0

How about it?

[root@01 opt]# echo 2/23/2013 | awk -F[/] '{print $1}'
2
[root@01 opt]# echo 2/23/2013 | awk -F[/] '{print $2}'
23
[root@01 opt]# echo 2/23/2013 | awk -F[/] '{print $3}'
2013

answered Mar 22 '13 at 13:46

Satish

16,544
29
93
149

How would this work when the dates are being read in like `awk -f prog_file 1/2/3 4/5/6` on the command line? – user2177896 Mar 22 '13 at 13:51
no comment for your awk answer, but I just want to remind better don't do those test with `#(root)`, even if you are 200% sure it is safe. :D, ignore it if it doesn't suit you.. I have learned lesson from it, so when I saw this I remind people... – Kent Mar 22 '13 at 14:01
prog_file as of now contains the BEGIN segment in the original question. – user2177896 Mar 22 '13 at 14:14

score 0 · Answer 4 · answered Mar 23 '13 at 02:45

If you're on a machine with nawk and awk, there's a chance you're on Solaris and using /bin/awk or /usr/bin/awk, both of which are old, broken awk which must never be used. Use /usr/xpg4/bin/awk on Solaris instead.

Anyway, to your question:

$ cat program_file
BEGIN {
   d1 = ARGV[1]
   d2 = ARGV[2]

   split(d1,array,/\//)
   print array[1]
   print array[2]
   print array[3]

   exit
}

$ awk -f program_file 2/23/2013 4/15/2013
2
23
2013

There may be better approaches though. Post some more info about what you're trying to do if you'd like help.

Separating Awk input in Unix

4 Answers4