davsclaus commented on PR #27265:
URL: https://github.com/apache/camel/pull/27265#issuecomment-6467036064
@luigidemasi thanks, and sorry for the late reply. Your experience from
CAMEL-24977, CAMEL-25049 and CAMEL-25138 is the case this design is built on.
- **Define once, derive the rest:** that is the goal in the doc. The feature
writes an annotated model once. The YAML schema fragment, the XSD and the
catalog entry are generated from it in the first steps of the plan, and
generated readers come later where they are worth it.
- **Java authoring vs. parsing Java source:** the "Java DSL" section keeps
them apart. The fluent API can build the same declaration model as XML and
YAML. How the converter reads that Java code is decided separately: either a
data-only builder in the vetted `LwJavaParser` replay surface, or reading the
call chain without replaying it.
- **Your test cases:** named choice criteria vs. ordered score levels are
already the rules the model has to carry. Numeric options holding a property
placeholder are a good test of the rule that reading a declaration never
resolves a placeholder, so the model must accept `{{...}}` where it expects a
number. Declarations shared across resources are not in the doc yet. I will add
them to the open questions, and use all three as test cases for the model.
_Claude Code on behalf of davsclaus_
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
