TestKit for Persistence Query

[patriknw]

First step would be to add query support to the new memory journal #26825

Not sure we need much more since ordinary Streams testkit can be used, but we can discuss ideas.

[manonthegithub]

@patriknw I am a bit confused of what is expected.
Do you mean that the testkit should allow users to test if persisted events contain needed information and can be queried by particular type of PersistenceQuery?

Then I think that just implementing support of PersistenceQuery for the new memory journal should be enough. Then users would be able to do their checks obtaining ReadJournal instance.

[patriknw]

The idea is that the new memory journal would implement the queries, such as eventsByPersistenceId and eventsByTag. You can look at the leveldb journal for inspiration.

[manonthegithub]

Ok great, I already started, looks like its almost working from the box :)

[manonthegithub]

There is no way of showing that the event was removed when you use for example eventsByPersistenceId which return infinite Source, the same with persistence ids. Is that as expected?

[manonthegithub]

@patriknw one more thing... in current implementation of the testkit the highest sequence number is preserved separately by each persistence id, so for each persistence id they start from 0. When we do persistence query by tag (which may get events with different persistence ids), the ordering by sequence number may give several events with the same seq number. Is that also expected?

[patriknw]

The ordering for eventsByTag is only promised per persistenceId so they arrive in seqNr order, but best if you sort them by timestamp so that events for different persistenceId are interleaved.

If you keep PersistentRepr around in the journal you can use the PersistentRepr.timestamp field. You have to set the timestamp when writing. That is also the timestamp that should be included in EventEnvelope.