Testing new functionality for error reporting from Swedeb

[rebeckahw]

Info from Swedeb regarding speech:
{"id":"prot-202223--135_032","gender":"Kvinna","party":"M","year":2022,"speaker":"Maria Malmer Stenergard"}

There is probably a segmentation error for this speech, it contains only the following text:
Tryck: Elanders Sverige AB, Vällingby 2023

[MansMeg]

This is excellent. Although, I guess this might be better as an issue? Or what does @BobBorges thinks?

[fredrik1984]

OK. We can discuss how to best create issues of these error reports.

Yes, the error report will essentially always be connected to a specific speech. @rebeckahw can we add xml:ids for each speech too and have that as metadata for each speech?

[rebeckahw]

Anything is possible :)
Looking in the source xml file, it seems like the xml:id is the id minus the order number (_032 in this case), so that would be an easy fix. (More detailed info about where exactly in the source data this segment comes from would probably be a future feature.)

[MansMeg]

Great! The id will help a lot. The rest would probably be solved in a discussion with the person reporting the error.

[fredrik1984]

Ok, good! Excellent if you @rebeckahw could add that to the speech metadata in SweDeb. We only need it when the user copies the speech metadata to the error report.

[rebeckahw]

Sure!

[BobBorges]

It would be best to report the file and the xml:id attribute of the offending element. We don't (yet) have speech IDs -- these are a sewdeb invention.