How to differenciate side-by-side column text in a film script?

[howiehill]

I am trying to capture all the dialogue in a film script which works fine until I hit sid-by-side like the attachment. Is there a way to determine when this happens to classify the left as the first piece of dialogue then the right as the second?

[JorjMcKie]

You seem to refer to get_text() when you say "capture".
Apparently, the text may split into two columns sometimes. How this is happening technically, decides about whether you have a chance at all (presence of \t characters in the same line). If there are just lots of spaces between left an right content, resolving this is very hard / impossible.
In any case, you must add own logic to achieve what you want.