bug(securityEvent): Add missing security event names to db

[nshirley]

Because:

• We're missing some security event names from some events in the db
• And we're missing some security event names in the model
• And some names have mismatching id

This commit:

• Updates all security event names in 'fxa.securityEventNames' that are missing
• Aligns all values on the model to match the db as source of truth

Checklist

Put an x in the boxes that apply

• My commit is GPG signed.
• If applicable, I have modified or added tests which pass locally.
• I have added necessary documentation (if appropriate).
• I have verified that my changes render correctly in RTL (if appropriate).

Screenshots (Optional)

Please attach the screenshots of the changes made in case of change in user interface.

Other information (Optional)

Any other information that is important to this pull request.

[nshirley]

This was not found in code, but the new recovery_phone_reset_password_complete is.

[nshirley]

here, _failed is what was in code, but the db had it as _failure. I'm unsure if we should keep the db as the source of truth or update it to match the model here so I went with the former given the comment at the top

[nshirley]

I was just digging into how events are written and, since they're written with whatever name is passed to recordSecurityEvent then this is probably not a great idea to change. We'd have some events with the old _failed and then all new events would have the _failure name.

Maybe it makes more sense to update the event name in the db instead of code/model?

[nshirley]

I also noticed that several of the recent values are missing from the PageRecentActivity/SecurityEvent.tsx mapping to display them. I see there's L10n values in there - how do I go about getting those that are missing added so we don't just show Other account activity on them?

[nshirley]

I also found that the securityEvents table records the event name as the ID so changing the ID's in the model is probably not a good idea 🤔 I'll see if I can dig in further to find the best solution that won't require updating the securityEvents table

[nshirley]

Okay, after a bit more review of how securityEvents are recorded, I think this is a safe change. Since we look up the nameId from EVENT_NAMES with bracket notation, we'll get undefined if the name doesn't exist on the object. That will fail to write since there's a FK constraint on the table (this also prevents writing of IDs beyond what is in the table)

So, we've just dropped these events that don't exist/map correctly. This should fix that!

[vbudhram]