mirror of
https://github.com/Codappix/search_core.git
synced 2024-11-15 10:56:21 +01:00
25 lines
1.2 KiB
ReStructuredText
25 lines
1.2 KiB
ReStructuredText
|
Breaking Change 149 "Extract RelationResolver to a new DataProcessor"
|
||
|
|
=====================================================================
|
||
|
|
|
||
|
|
The resolving of relation, based on TCA, is no longer done by the indexer. Instead we
|
||
|
|
now provide a DataProcessor to solve this job.
|
||
|
|
|
||
|
|
As this makes it necessary to configure the DataProcessor, this is a breaking change.
|
||
|
|
Before the resolving was done out of the box.
|
||
|
|
|
||
|
|
So why did we change that? The resolving of relations was already implemented before
|
||
|
|
we added the data processors. As the concept of data processors is far more flexible,
|
||
|
|
we try to migrate hard coupled components step by step. The benefit of this change is
|
||
|
|
basically that you can now configure the resolving of relations and far more
|
||
|
|
important, the order of execution.
|
||
|
|
|
||
|
|
Now it's possible to first copy some fields, e.g. ``starttime`` and ``endtime`` to
|
||
|
|
further fields and to resolve relations afterwards. As the copied fields are not
|
||
|
|
configured in TCA, they will be skipped. This way an integrator can keep certain
|
||
|
|
information as they are.
|
||
|
|
|
||
|
|
Also the processor is now configured as all other processors. You can now optionally
|
||
|
|
configure fields to not process.
|
||
|
|
|
||
|
|
See :issue:`149` and :issue:`147`.
|