Shell Programming and Scripting

File contents:
1. 4 33 45 21 1 01:25:25
2. 4 33 45 22 1 01:25:25
3. 4 33 45 23 1 01:25:25
4. 4 33 45 24 1 01:25:25
5. 4 33 45 25 1 01:25:25
6. 77 35 70 17 14 02:21:40
7. 77 35 70 18 14 02:21:40
8. 77 35 70 19 14 02:21:40
9. 77 35 70 20 14 02:21:40
10. 77 35 70 27 14 02:21:40
11. 77 35 70 17 1 02:2210
12. 77 35 70 18 14 02:2210
13. 77 35 70 19 1 02:2210
14. 77 35 70 20 1 02:2210
15. 77 35 70 27 1 02:2210
16. 87 31 109 27 99 03:35:34
17. 87 31 109 18 1 03:35:34
18. 87 31 109 19 14 03:35:34
19. 87 31 109 20 98 03:35:34
20. 87 31 109 27 1 03:35:34
21. 77 35 70 17 1 03:36:15
22. 77 35 70 18 1 03:36:15
23. 77 35 70 19 1 03:36:15
24. 77 35 70 20 1 03:36:15
25. 77 35 70 27 1 03:36:15
26. 500 29 577 11 1 03:37:59
27. 500 29 577 12 1 03:37:59
28. 500 29 577 13 1 03:37:59
29. 500 29 577 14 1 03:37:59
30. 500 29 577 15 1 03:37:59
31. 701 33 924 11 1 05:22:44
32. 701 33 924 12 1 05:22:44
33. 701 33 924 13 1 05:22:44
34. 701 33 924 14 1 05:22:44
35. 701 33 924 15 1 05:22:44
To explain the lines above, Column 1 numbered from 1. to 35. is just a column numbering to explain the rows in question. Therefore, please ignore column 1 numbered from 1 to 35 when processing the file.

Rows 1-5 is considered one point, rows 6-10 is another point, rows 11-15 is the same point as rows 6-10, rows 16-20 is another point, rows 21-25 is the same point as rows 6-10 & rows 11-15 and so on.

What I want to do is to print the output of the Point if its Column 5 shows 14, 98 or 99 and only if the same Point does not show 14 or 98 again.

To explain further, Rows 6-10 & 11-15 have 14 in Column 5 but since they are the same Point and it's repeated in rows 21-25 with no more 14 in Column 5, it is not printed.
Rows 16-20 has 14, 98 & 99 in Column 5 but since it was not repeated within the file, the expected output would be:

87 31 109 03:35:34 (without the 4th & 5th Columns).

I tried the following commands but it didn't work:

i=1
awk '
for (i=1;i<=NR;++1) {
if ($5=14 | $5=98 | $5=99);\
print NF
}' file

Any help?

If I'm not missing something:

Code:

awk '{ K[$1,$2,$3]++ }
$5 ~ /^(14|9[89])$/ { 
R[$1,$2,$3] = $1 FS $2 FS $3 FS $6
} END {
for (k in K)
 if ((K[k]<6) && (k in R))
   print R[k]
}' filename

Use nawk or /usr/xpg4/bin/awk on Solaris.

radoulov,

Could you please explain your script further?

Thanks!

Two arrays - the first counts the number of logical units:
given your sample it seems it's a five record set with
unique key based on the first three fields, the second is populated
with the first three fileds (as keys) and the output formated
as you want as elements but only for a records that match your criteria
(fifth field in 14,98 or 99). After reading the input it outputs
only the records that are part of a unique section and whose
fifth field matches the pattern.

Thank you for your explanation but I got the following error messages after running the script:

awk: syntax error near line 1
awk: illegal statement near line 1
awk: syntax error near line 3
awk: illegal statement near line 3
awk: syntax error near line 6
awk: illegal statement near line 6

read the radoulov's reply once again!

Opss, sorry about that. Yup, nawk in the script worked. Thanks!

One more question, I would like to show all lines with 14, 98, and 99, too. How would I do that?