caucus | Realtime Collaborate Editor with Embedded Compiler | Collaboration library

I re-implemented your steps into 3 blocks for more maintainability and clarity. The steps are explained as follows.

1. Loader-discarder: As one can easily seen, the previously-loaded "substrings" are not useful at all. They can be processed as the inevitable and costly line-by-line loading occurs within open(). As a result, every line retained is informational after this step.
2. Repeating-substring finder: A simple brute-force searcher for the length of repeated substring. If not found, 0 is returned. This takes 2 strings as arguments.
3. Repeating-substring remover: Apply the function of 2. onto your data. By separating 2. and 3., the responsibility of the string-searching logic and the data-chopping logic were made more clear and traceable.

The rest should be self-explanatory. Try to divide your logic into small and independently testable units, each serving a single logical purpose instead of mixing them altogether. I didn't reuse your code not because I feel the logic is fairly complex and is unlikely to be well-maintained without a proper refactoring.

Code

You can prettify the html using Beautifulsoup

You may combine the two solutions: