dashboard: bug json export changes

[a-nogikh]

Once a day, generate .jsonl.gz archives of bug infos for the selected namespaces. Refactor the dashboard code to directly reuse the public json API code.

The archives will let us more easily do more a involved syzbot stats analysis.

Cc #5439

[tarasmadan]

Let's start from the requirements part.
I'm trying to understand https://github.com/google/syzkaller/tree/master/tools/syz-db-export role here.

[a-nogikh]

Ah, didn't know we already do it regularly. Then we can probably only merge the first two commits, or just the second one.

[a-nogikh]

Dropped the third commit (and archived it).

I think the refactoring done in the first commit is still valuable as is - the handleBug function has become way too big as is doing too much.

[tarasmadan]

Could you please explain how will this additional information be used?
Why archived export is the best way to do the things?

[a-nogikh]

I plan to write a Colab python script that would download the archive, transform into a pandas dataset and do some calculations / draw some graphs for the LPC presentation(s).

If datastore were a more traditional database, some of that could be done by simple SQL queries, but in our case we can only export from Datastore and import somewhere else where we can properly filter/aggregate the data.

[a-nogikh]

@tarasmadan PTAL