Extract a block of text with sed/awk/grep from symdev list -v output?

Why not try using the -xml output and an XML parser?  Or, you could grab the XML output and just grep the fields you want?

cat symdev-v.out | grep -B 3 -w 00DA

Something like that ?

XML parser would be great, but size constraints make it difficult (i.e. the "symdev list -v -output xml" from each array is 4-5x larger than the "symdev list -v") so as these run to multiple GB per XML output, the "symdev list -v" is the most manageable approach to collecting the info.

Hi dynamox, I want to extract all information for the device, i.e. *everything* that is in the record for the device 00DA, i.e. exactly equivalent to running: symdev -sid XXXX show 00DA. Of course, symdev -sid XXXX list -v   contains every "show" record one after another. using grep -B 3 -w 00DA, will just get the line with 00DA, and the 3 lines before that.

Can you think of a way, that would extract (from a list -v), the entirety of the information for device 00DA ? (so everything in -B 3 -w 00DA, but also every line *after* 00DA right down to the start of the next device record)

play around with devices that have rdf pairing, i tested with clone and snaps ..seems to capture everything.

sed -n -e '/Device Symmetrix Name    : 00DA/,/Device Physical Name/ p' symdev-v.txt  |  sed -e '$d'

my google kung fu is not leading me to what i want. You might need to start tinkering with perl, might be easier to parse.

Thanks dynamo, my sed/perl is not good enough I guess (I don't even really see what the $d does in your note above though I see that it works!). I think your approach looks great though, so if either of these pages give an obvious clue on how to extract the multiline part that would be appreciated:

http://austinmatzko.com/2008/04/26/sed-multi-line-search-and-replace/

http://unix.stackexchange.com/questions/26284/how-can-i-use-sed-to-replace-a-multi-line-string

http://stackoverflow.com/questions/1221245/sed-multiline-replace

http://askubuntu.com/questions/347223/sed-substitute-through-a-sentence-on-multiple-lines

These are more for replacing a multiline piece of text of course, but hopefully also applicable to picking out the region of text with multiline starting point of:

"Device Physical Name     : <some value> Physical /Device Symmetrix Name    : <symdev ID>"

Please upload the output of symdev -sid xxxx list -v.

and let me know what kind of output u are looking for.

There's probably a better way to do this...

#!/usr/bin/perl

$INFILE = '/path_to_symdev_output';

open(symdev,"<$INFILE") or die "Failed opening $INFILE ($!)";

$found = 0;

while ($line = ) {

   if ($found == 1) {

        last if ($line =~ /Device Symmetrix Name :/);

        print $line;

   if ($line =~ /Device Symmetrix Name : 00DA/) {

       $found = 1;

       print $line;

   }

}