I am using regex to split a string <book name> by <author name> into book and author names.

re.split(r'\bby\b', text, 0, re.I)

But problem arises when the book name contains the word "by" (e.g. Death by Blackhole by Tyson => ['Death', 'by Black...'])

How do I split the string by the last occurrence of the search pattern?

I have a hunch -/+ve look-ahead/behind could be useful here, but currently splitting hairs trying to construct the proper syntax.

You could use findall with a greedy .* before the by:

re.findall(r'(.*)\s+by\s+(.*)', text, re.I)

See it on repl.it