How to extract values from a command with multi-line output?

[Bo Berglund]

I am trying to script working with a file system image file after I have found
out how I can do this manually.

When scripting the procedure I have one big problem and that is how I can
extract data from the output of an fdisk command...
I am not versed in the various tools on Linux I have seen mentioned when
searching. Something simple to just extract the numbers is what I need.

I do this to get the info (the lines may be wrapped by the newsreader):
----
$ fdisk -lu filename.img | tail -n 2
filename.img1        8192  532479  524288  256M  c W95 FAT32 (LBA)
filename.img2      532480 3923967 3391488  1.6G 83 Linux
----

The response from this command is two lines (one for each partition in the
image) and I am only interested in the 3 numbers on the two lines:
8192 524288 532480 

These are the 1st and 3rd number of the first line and the first number on the
second line, following the file name.

How can I get these into 3 variables inside the script?

I use them in order to mount the images:

sudo mount -t ext4 -o loop,offset=$((532480*512)) filename.img /mnt/disk
sudo mount -t vfat -o loop,offset=$((8192*512)),sizelimit=$((524288*512))
filename.img /mnt/disk/boot

I want to automate this mount operation in a "mountimage" script, so how to
extract the numbers from the response shown?


-- 
Bo Berglund
Developer in Sweden