Related: #4746

# Conflicts:
#	Neos.ContentRepository.Core/Classes/ContentRepository.php

# Conflicts:
#	Neos.ContentRepository.Core/Classes/ContentRepository.php
#	Neos.ContentRepository.Core/Classes/Factory/ContentRepositoryFactory.php
#	Neos.ContentRepositoryRegistry/Classes/ContentRepositoryRegistry.php
#	Neos.ContentRepositoryRegistry/Classes/Service/ProjectionService.php
#	Neos.ContentRepositoryRegistry/Classes/Service/ProjectionServiceFactory.php
#	phpstan-baseline.neon

# Conflicts:
#	Neos.ContentRepository.Core/Classes/ContentRepository.php
#	Neos.ContentRepository.Core/Classes/Factory/ContentRepositoryFactory.php
#	Neos.ContentRepositoryRegistry/Classes/ContentRepositoryRegistry.php

and implement progress callback (for progress bars)

type annotation tweaks

...probably not

to see wether other tests succeed

This reverts commit 1562435.

…tchup-mechanism-3

…le of times

…tchup-mechanism-3

Attempting to write to the content repository from two threads previously would cause likely an exception in Neos 9.0:

> Failed to acquire checkpoint lock for subscriber "Neos\ContentGraph\DoctrineDbalAdapter\DoctrineDbalContentGraphProjection" because it is acquired already

Now `forUpdate` is used in `DoctrineSubscriptionStore::findByCriteria` which solves this.

and if commented out throws as expected:

> An exception occurred while executing a query: SQLSTATE[40001]: Serialization failure: 1213 Deadlock found when trying to get lock; try restarting transaction.

Without advancing via `->next()` we will attempt to handle one event twice, and then run into the log: Which makes sense as the subscriber _was_ on that sequence number before and we are interested in the next events.

> subscription "contentGraph" is farther than the current position (5053 >= 5053)

Also simplifies `lowestPosition`

… projection (can be later reintroduced)

…jections

- retry strategy
- runmode
- pausing, finishing
- InMemorySubscriptionStore

Followup for remove other generic subscription concepts not required for projections

Detached subscriptions cannot be removed for now

…king is untested there

…SubscriberForProjection`

Simplification to make the projection use case first level
See #5375 (comment)

and extract commitExampleContentStreamEvent

…nt is only handled once and further introspection

…y applied in the same batch that did not cause errors

…al logic

No retry is simpler at first and its unlikely that a projection will fix itself.

#5321 (comment)

> Anyways, I think with the removed retry strategy we should just get rid of the automatic retry altogether right now.
  It's quite unlikely that a retry suddenly works without other changes. So I'd be fully OK if it was only possible to manually retry failed subscriptions for 9.0

> there's a mismatch between variable name "additionalProjectionsFactory" and type "ContentRepositorySubsciberFactory"
> the distinction between projection and subscription makes sense even if the only supported type of subscription target projections (for now)

#5375 (review)

which lead to it being invoked on aktive

>  Subscriber "Vendor.Package:FakeProjection" could not invoke onBeforeCatchUp: Subscriber with the subscription id "Vendor.Package:FakeProjection" not found.

…cription-pr

TASK: Radical cleanup for subscription pr to simplify and get into 9.0

Also allows the `Vendor.Package:FakeCatchupHook` to be picked up for testing

…eric `ProjectionStates`

Readded assertions as retry was removed for now ... and should probably NOT do anything if setup is called!

…rationIsRemoved

> InvalidArgumentException: Subscriber with the subscription id "Vendor.Package:FakeProjection" not found.

The subscriber is detached, so the state is not calculate-able at this point!

…terCatchUp failes

as we consider it a critical developer error

For `onBeforeCatchUp` we could probably wrap a savepoint and roll it back and skip also the projection, but errors in `onAfterCatchUp` would then analog also need to rollback only the one projection where it was registered. This is not possible and too complex.

see \Neos\ContentRepository\BehavioralTests\Tests\Functional\Subscription\CatchUpHookErrorTest::error_onBeforeCatchUp_abortsCatchup

…ailure

https://neos-project.slack.com/archives/C04PYL8H3/p1732318989845619

We dont want to rollback the main transaction, as other projections still need to be processed, the previously working events need to be applied, and we want to set the ERROR state of the projection

…uses`

…the tables are dropped

To reduce additional sql query and lock, and do it in the main transaction

We do not expect any changes during runtime. Setup and status should handle this case.

and failing test for setupIsInvokedForPreviouslyActiveSubscribers

…ark detached ones as detached.

…ction commands

…Status`

Like the `ProjectionSubscription` projections subscribers for projections will have an explicit state: `ProjectionSubscriptionStatus`. If its extended to allow other subscribers another status type should be introduced.

replayRequired todo remove we cannot figurea that out in the status after all!

…status cannot be calculated

... and when extending the system to support multiple subscribers, we cannot know their original classification but have to use a special empty placeholder like: `DetachedSubscriptionStatus`

This also makes the `$projectionStatus === null` detached case more explicit when using status.

…o CR Maintainer

Under consideration of the new `ProjectionSubscriptionStatus`

reverts partially bace8ff

see #5378 (comment)

…ught up

… event store position for debugging and status information