compare and extract specific string from cells to new column

Hithere,

I have a sheet with a list of data in column A, and a list of data in column B.
I want to accentuate (or extract) the differences between those two rows and put those in a new column.
(and possibly have a new column with A and B added together without duplicates)

The complicating factor (which makes it impossible for me to know where to start solving this problem), is that only a specified string in these cells is to be checked for differences/similarities.

To be more specific:
I want to have cell content [ <color="black" "123" ] being considered to be a match with [ <color="black" "abc" ]
So only the specific string between [ < ] and the second [ " ] is to be checked for difference/similarity.

So something like this:

input	input	result	result	result
A	B	missing from A	missing from B	A + B
<color="black" "123"	<color="black" "abc"	<color="blue" "bcd"	<color="green" "234"	<color="black" "123"
<color="green" "234"	<color="blue" "bcd"	<color="yellow" "def"	<color="red" "345"	<color="blue" "bcd"
<color="red" "345"	<color="white" "cde"			<color="green" "234"
<color="white" "456"	<color="yellow" "def"			<color="red" "345"
				<color="white" "456"
				<color="yellow" "def"

I attached an excel sheet to hopefully make it easier to understand (and test) what I would like to achieve.

TIA!
extract column differences.xlsx

Originally Posted by Alex H

I think this basically does what you want, although the ordering may be wrong. Unfortunately I don't know how to get rid of the spaces.

Wow, thnx, that's impressive Alex.
My big challenge now will be to figure out what is happening in your formulas, and then try to translate it to the actual sheet I want to use it in.
That'll be a couple of days I'm afraid, but this is a fantastic start...

For now there is one strange hick-up I notice and already can't figure out:
When I add e.g. [ <color="pink" "111" ] in both cells A8 and B8, output in C8 and D8 is wrong? (saying pink is missing from both)
edit: got this last part.