turn a string of form
'[/str1XYZ.html, /str2XYZ.html, /str3XYZ.html, ...]' into a a list of strings '/str1XYZ.html', '/str2XYZ.html', ...
Each occurence of XYZ represents a potential different 1-3 digit number, and 'strN's might be different in length, example of the string originally returned by query:
I started with substr functions, and increasingly more complex, nested locate functions, to move the starting point for a pattern to the right for each new element, but quickly gets unwieldy. Looking for a cleaner construct, like, pseudo-codish:
find(find_me, look_here, occurence n), rather than find(find_me, look_here, start_at).