Those super optimized solutions are fun to read about. However, in real life I would never assume a millions rows of text all have all valid data in a specific

Those super optimized solutions are fun to read about. However, in real life I would never assume a millions rows of text all have all valid data in a specific format, much less a billion rows. Thus a slower but more robust solution would be more realistic.