Shell Script - Find Drive by Size

[hernandito]

This may or may not be a newbie question, but a newbie I am.

I need guidance defining a command variable that will return the
disk#

The trick is that I need the script to look at the disks (not partitions) that are between 2GB and 8 GB in size, and return the disk#.

In plain english, I want for the command to:
"Tell me the disk number of drives that are between 2Gb and 8Gb. If there is more that one drive that meets the criteria, or if there are none, then just yell "Error"."

I am going to be doing a dd command to restore an image file to a removable thumb drive and I want to avoid restoring the image to one of the system's hard drives.

Any help or guidance is greatly appreciated.

Thank you.

[Guiyon]

No, there is no single command to do so (that I know of, at least). The diskutil tool will be able to provide you with both a list of all known attached disks and the total size of each of those disks but you will need to write a small script to parse out that information. Below is a quick & dirty Perl script that should at least give you a general idea of what you'll need to do.

Edit: You could also parse out the information from the 'df' command but then you would still have to lookup the device->mountpoint mappings or do some addition.

#!/usr/bin/env perl
#

@lines = split /\s+/, qx( /usr/sbin/diskutil list );

# Filter out any lines that are not of the form /dev/diskN, where
# 'N' is a 1 or more digit number, ie: /dev/disk6.
@disks = grep { $_ =~ /^\/dev\/disk[0-9]{1,}/ } @lines;
%disk_size = ();

foreach (@disks) {
my $info_str = qx( /usr/sbin/diskutil info $_ );

# Search for the line that begins with "Total Size:" and strip
# out the size of the disk in bytes and store it in a hash.
# Capturing "\d+\.\d+\s+\w+" instead of the current "\d+"
# will give you the size in a human-readable value
if( $info_str =~ /Total Size:\s+\d+\.\d+\s+\w+\s+\((\d+)\s+/ ) {
print "Disk '$_' => $1 Bytes\n";
$disk_size{ $_ } = $1;
}
}

[chown33]

The 'df' command provides the sizes of mounted partitions, so you could add them up by identifying the drive. Piping df's output to awk would work. It's more than capable of matching patterns by line, splitting args, and doing arithmetic. It's probably less than a dozen lines of awk.

Sample data:
df -k

Filesystem 1K-blocks Used Avail Capacity Mounted on
/dev/disk0s3 26083328 16249064 9578264 63% /
devfs 104 104 0 100% /dev
fdesc 1 1 0 100% /dev
<volfs> 512 512 0 100% /.vol
/dev/disk0s2 26083328 16187656 9895672 62% /Volumes/Foow
/dev/disk0s4 26083328 16147632 9935696 62% /Volumes/Barw2
/dev/disk0s5 233592112 38475312 195116800 16% /Volumes/Work
automount -nsl [207] 0 0 0 100% /Network
automount -fstab [212] 0 0 0 100% /automount/Servers
automount -static [212] 0 0 0 100% /automount/static
/dev/disk1s3 1853912 557772 1296140 30% /Volumes/SD1-HFS
/dev/disk2s2 1000904 106232 894672 11% /Volumes/Xyzzy-HFS
/dev/disk3s2 20840448 16696212 4144236 80% /Volumes/Plugh
/dev/disk3s4 348432432 140594912 207837520 40% /Volumes/Plover
/dev/disk3s3 20840448 4040760 16799688 19% /Volumes/Duh
afp_2YPUGc00000H0000oM0000VU-1.2c000007 156883120 74910092 81973028 48% /Volumes/WTF


My main drive is partitioned, and so is /dev/disk3. The others are not.

Notice the naming pattern (/dev/diskNsM), where 's' stands for "slice", an archaic synonym for "partition". The sizes under the "1K-blocks" heading are the overall size of the slice, measured in 1K blocks. Read the man page for df for other options (GB, MB, 512-byte blocks).

[hernandito]

The 'df' command provides the sizes of mounted partitions, so you could add them up by identifying the drive. Piping df's output to awk would work. It's more than capable of matching patterns by line, splitting args, and doing arithmetic. It's probably less than a dozen lines of awk.

Sample data:
df -k

Filesystem 1K-blocks Used Avail Capacity Mounted on
/dev/disk0s3 26083328 16249064 9578264 63% /
devfs 104 104 0 100% /dev
fdesc 1 1 0 100% /dev
<volfs> 512 512 0 100% /.vol
/dev/disk0s2 26083328 16187656 9895672 62% /Volumes/Foow
/dev/disk0s4 26083328 16147632 9935696 62% /Volumes/Barw2
/dev/disk0s5 233592112 38475312 195116800 16% /Volumes/Work
automount -nsl [207] 0 0 0 100% /Network
automount -fstab [212] 0 0 0 100% /automount/Servers
automount -static [212] 0 0 0 100% /automount/static
/dev/disk1s3 1853912 557772 1296140 30% /Volumes/SD1-HFS
/dev/disk2s2 1000904 106232 894672 11% /Volumes/Xyzzy-HFS
/dev/disk3s2 20840448 16696212 4144236 80% /Volumes/Plugh
/dev/disk3s4 348432432 140594912 207837520 40% /Volumes/Plover
/dev/disk3s3 20840448 4040760 16799688 19% /Volumes/Duh
afp_2YPUGc00000H0000oM0000VU-1.2c000007 156883120 74910092 81973028 48% /Volumes/WTF


My main drive is partitioned, and so is /dev/disk3. The others are not.

Notice the naming pattern (/dev/diskNsM), where 's' stands for "slice", an archaic synonym for "partition". The sizes under the "1K-blocks" heading are the overall size of the slice, measured in 1K blocks. Read the man page for df for other options (GB, MB, 512-byte blocks).

Thank you guys.... Sadly this is all way above my head... I get the idea of what to do, but I have no idea how to go about it. I am not sure if this is proper here, but I would be happy to pay an expert to assist me with this. I could easily post this in GetAfreelancer.com, or Paypal someone directly. It needs to run inside a shell script.

Please email me directly if you are interested... hernando(at)econoco.com

Thanks again!

[Doctor Q]

Offers to buy are limited to the Marketplace forum. See the forum rules and Marketplace rules.