To expand a bit, Cayenne basically provides a way to map DB-specific JSON columns to "org.apache.cayenne.value.Json". This gives a (hopefully) smooth experience reading and writing such values. How to use it:
1. Ensure your DB supports native JSON type (e.g. Postgres) 2. Map your DbAttribute as type "OTHER" 3. Map your ObjAttribute as "org.apache.cayenne.value.Json" 2 & 3 would happen automatically on reverse-engineering if you are using DB-first flow. "org.apache.cayenne.value.Json" is simply a String wrapper. It intentionally doesn't have any parsing logic. Its purpose is to be a performant dependency-free container to pass JSON payload between the app and DB. If you need JSON as a tree, you may take a few more steps: 4. Pick a JSON parser such as Jackson. 5. Map ObjAttribute as a type supported by that parser (e.g. JsonNode in Jackson) 6. Write and register a custom "ValueObjectType<JsonNode, Json>" [1] 7. There is even a way to use DB-specific functions to query JSON [2] @Nikita: is it possible to apply custom functions in the WHERE clause as well? Andrus [1] https://cayenne.apache.org/docs/4.2/cayenne-guide/#value-object-type <https://cayenne.apache.org/docs/4.2/cayenne-guide/#value-object-type> [2] https://github.com/apache/cayenne-examples/blob/master/cayenne-jdbc-type-other/src/main/java/org/apache/cayenne/demo/JsonTestCommand.java#L59 > On Oct 13, 2020, at 11:07 AM, Nikita Timofeev <ntimof...@objectstyle.com> > wrote: > > Hi Markus, > > No docs for that yet, but here's an example of new types usage: > > https://github.com/apache/cayenne-examples/tree/master/cayenne-jdbc-type-other > > On Tue, Oct 13, 2020 at 10:41 AM Markus Reich <reich.mar...@gmail.com> wrote: >> >> Hi Andrus, >> >> is there an example or doc for the support of JSON types? >> >> regards >> Meex >> >> Am Di., 13. Okt. 2020 um 08:48 Uhr schrieb Andrus Adamchik < >> and...@objectstyle.org>: >> >>> 4.2.M2 release is out [1]. A few cool things in this release: >>> >>> * Support for JSON and geospatial value types. >>> * Ordering on aggregate functions and in-memory evaluation of aggregate >>> expressions. >>> * Modeler support for downloading JDBC drivers from maven central. >>> * Runtime changes that may be less obvious to the end user, but allow to >>> handle a variety of edge cases. >>> * Bug fixes >>> >>> Also looks like we are pretty close to beta freeze. The main scope of 4.2 >>> is done, though of course improvement ideas keep popping up all the time, >>> especially now that the new stack significantly expanded our SQL vocabulary >>> and JDBC capabilities. E.g. one such idea is using EXISTS instead of JOIN + >>> DISTINCT for to-many conditions [2]. >>> >>> Enjoy M2! >>> >>> Andrus >>> >>> [1] https://cayenne.apache.org/2020/10/cayenne-42m2-released/ >>> [2] https://issues.apache.org/jira/browse/CAY-2684 >> >> >> >> -- >> *Markus Reich* >> Waldweg 62 >> 6393 St. Ulrich am Pillersee >> www.markusreich.at / www.meeximum.at >> markus.re...@markusreich.at > > > > -- > Best regards, > Nikita Timofeev
