Reverse script to extract transcripts

I managed to reverse the logic of extract_text.py so that it generates an English transcript based on the given locale.
e.g. python 0_transcripts/extract_reverse_text_to_english.py ep39_Aquah fr
would generate an English transcript using French name references as the source and translate them using the .po database.
This may be useful in the future to avoid this type of problem: !945 (merged)
I was thinking of including this as an exception to existing script when the locale is called as en, but that would make the script less reliable.

I wanted to name the script extract_text_reverse_to_english.py, but that would mess up the tab completion,
so I ended up with extract_reverse_text_to_english.py.
You can rename it if you like.

New commit: Removed all .gitkeep files no longer needed - they are the result of adding folders via the web interface. (to keep the folder in GIT)

I'm also cleaning up old branches with no content, things that have already been fixed, and changes that are no longer relevant (like covers, which are now auto-generated).
_I have seen how some people upload their work..._ 😅

I have a question for you David: I found a couple of old branches for toki pona and sitelen pona from Ret Samys. I compared the files with the master and according to my research all this was already introduced to master through other ways. I now ask if my assumptions are correct and these branches may be deleted: tp-minor-fixes | sp-new | sp-fixes