Parsing String Variable in Bash Script

Question

I am working on a bash script that will parse a full string name to provide me with a firmware value. For example the original string name will be something like:

Example_v0p1

Desired Output:

0.1

The first part of my function deliminates the string based on the underscore value and sets the second element of the array equal to a variable called version:

arr=(`echo $1| tr '_' ' '`)
version=${arr[1]}

Using the example I provided, $version now equals "v0p1". I am searching for a way to parse this string further so that it will look like the desired output.

If it is possible I would like the solution to be able to handle multiple variations of the same format so that I can use the same script for all of the version codes that could be generated. The syntax however will always be the same. v#p#{p#p#....} Examples:

v0p1p1
v3p2p0p1
v1p3

Desired output:

0.1.1 
3.2.0.1
1.3 

I am unsure of the best way to approach this problem. I am not experienced enough with REGEX's to accomplish this and I cannot come up with an appropriate way to parse the strings because they are going to be different lengths.

Answer 1

You can do something like this:

s='v3p2p0p1'
ver=$(sed 's/p/./g' <<< "${s:1}")

echo "$ver"
3.2.0.1

s='v0p1p1'
ver=$(sed 's/p/./g' <<< "${s:1}")

echo "$ver"
0.1.1

Answer 2

Use parameter expansion:

#!/bin/bash
for v in v0p1p1 v3p2p0p1 v1p3 ; do
    v=${v//[vp]/.}             # Replace v's and p's with dots.
    v=${v#.}                   # Remove the leading dot.
    echo "$v"
done

Answer 3

This will work:

var="$(echo "v0p1p1" | tr -s 'vp' " .")"

It replaces v with space though.

If you don't want that you can do:

var=$(echo "v0p1p1" | tr -d 'v' | tr -s 'p' '.')