Help with awk or sed.

tuxtutorials · 07-22-2009, 11:45 PM

Hi, I am stuck on the following issue with awk and need help. I currently have data that looks like this:

Directory of /home/admin/slap/test

06/09/2009 08:15 AM 954,880 bin.fla
02/09/2009 09:48 AM 173,905 bin.swf
03/23/2009 11:04 AM 2,168 bin.xml
06/09/2009 08:15 AM 830,464 dnd.fla
02/05/2009 12:47 PM 175,001 dnd.swf
02/05/2009 11:13 AM 1,835 dnd.xml
06/09/2009 08:15 AM 755,712 mcq.fla
06/09/2009 08:15 AM 46,843 mcq.swf
06/09/2009 08:15 AM 2,706 mcq.xml

Directory of /home/admin/slap/test2

06/09/2009 08:16 AM 954,880 bin.fla2
02/09/2009 09:47 AM 173,905 bin.swf2
03/23/2009 11:04 AM 2,168 bin.xml2
06/09/2009 08:17 AM 830,464 dnd.fla2
02/05/2009 12:48 PM 175,001 dnd.swf2
02/05/2009 11:10 AM 1,835 dnd.xml2
06/09/2009 08:16 AM 755,712 mcq.fla2
06/09/2009 08:14 AM 46,843 mcq.swf2
06/09/2009 08:12 AM 2,706 mcq.xml2

I need to somehow run through this file and extract out "Directory of /home/admin/slap/test" and append it to each file in the 5th field.

So the end result would be something like:

Directory of /home/admin/slap/test/bin.fla
Directory of /home/admin/slap/test/bin.swf
Directory of /home/admin/slap/test/bin.xml
etc....

for each block of files. So the next block would be

Directory of /home/admin/slap/test2/bin.fla2
Directory of /home/admin/slap/test2/bin.swf2
Directory of /home/admin/slap/test2/bin.xml2

Not sure if awk would be the right tool or if this is even possible. Any advice would be great. This was a originally a directory list of all files and we need to somehow reconstruct back to a full directory listing.

This is what I have tried so far, just confused on how to append to each file in each block.

#!bin/awk -f
#

/Directory of.*/ {
x = $3 #Set field $ into variable x
print $x
}

Thanks

colucix · 07-23-2009, 02:43 AM

You missed the part of code which prints $5 preceded by the directory line:

Code:

awk '/^Directory of/{string=($0 "/")} $5!=""{print string $5}' file

Stéphane Ascoët · 07-23-2009, 02:49 AM

Something I've made at work:

#!/usr/bin/awk -f
BEGIN {
print "Starting and searching for ",searchingfor
workingdirectory="non"
}
#If we find "Directory of", we finished the directory
/Directory of/ { workingdirectory="non" }
#We print the result if we are in a directory
{ currentline=$0
if (currentline ~ searchingfor && workingdirectory !="none") { print workingdirectory,$0 } }
#Each time we find a directory, we store its name in workingdirectory
/Directory of/ { workingdirectory=$4 }

You'll have to make a lot of changes of course.

tuxtutorials · 07-23-2009, 03:26 PM

Thanks for the reply's will muck around with some of these solutions and try to get it working. P.S sorry for the double post.