Trigger information in dorothea/penthesilea

As you know, Irene is now copying the trigger information from the waveforms files. I've been asked to propagate this too to dorothea and penthesilea. But I'm not sure about the best way to do it.

It can not be done in the same way as it is done in Irene because kdsts/hdsts only have one table that can have more than one row per event, so you can not associate one row of that table to one row of the trigger table, there is no /Run/events table either in those files (event numbers and timestamps are included in the one table they create).

We have basically two options:

• Add a new column to kdsts/hdsts with the trigger type. This I think would be the most convenient way for analists, all the information would be in the same table and could be easily read as dataframe. For MC files a default value would be set (-1?)
• Create a new table for the trigger information that would include the event number. But to use this you would have to combine the two tables "manually" (a pandas from_records wouldn't be enough).

Another story is the information about the channels that have triggered. Do we want that information available at this level? If so, how do we store it? In Irene is a 2d array without event numbers.

Opinions?

Start by reviewing how Irene copies trigger information and how dorothea and penthesilea represent kdsts/hdsts, including the single table and event numbers. Resolve the issue's open choice between a trigger-type column and a separate table, and decide whether triggered-channel data belongs at this level. Done requires an agreed schema and propagation behavior for both data and MC files.