suggestions on how to go about this type of "reporting"?

[spyghost]

hi,

i am dealing with this kind of table in text (space delimited)

full UNIX Daily Backup Pool
0% Windows Daily Backup Pool
23% UNIX Monthly Backup Pool
full Windows Weekly Backup Pool
5% Test Backup Pool
...
etc

what i am planning to do is to generate this kind of format:
Name Full Partially Full Free Total
UNIX Daily Backup Pool 3 5 5 13
UNIX Monlty Backup Pool 5 2 5 12
...

the problem that i am actually encountering is that there are only actually 2 colums, but since it is space delimited, awk sees more than two columns. (i can't modify the program that generates this output, its unmodifiable). how would i go about with dealing with only 2 columns?

right now, i am looking into using 2 dimensional arrays to hold the number of counts. is that efficient enough for the problem?

suggestions are highly apprciated... thanks

 

[PHV]

A starting point:
nawk '{a=$1;$1="";b[$0]=b[$0]"\t"a}END{for(i in b)print i,b}' /path/to/input

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ222-2244

 

[futurelet]

[tt]
BEGIN { FS = "^ *full +|% +" }

$2 {
type[$2]++
if ( ""== $1 )
status = "full"
else if (0==$1)
status = "free"
else
status = "part"
counts[$2,status]++
}

END {
fmt = "%-30s%5s%13s%6s%7s\n"
printf fmt,"Name","Full","Part. full","Free","Total"
for (k in type)
printf fmt, k, counts[k,"full"], counts[k,"part"],
counts[k,"free"], type[k]
}
[/tt]
Let me know if this works.