[
https://issues.apache.org/jira/browse/NIFI-15856?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
David Handermann updated NIFI-15856:
------------------------------------
Fix Version/s: 2.10.0
Resolution: Fixed
Status: Resolved (was: Patch Available)
> Add Serialized JSON Input Handling property to JsonRecordSetWriter
> ------------------------------------------------------------------
>
> Key: NIFI-15856
> URL: https://issues.apache.org/jira/browse/NIFI-15856
> Project: Apache NiFi
> Issue Type: Improvement
> Reporter: Pierre Villard
> Assignee: Pierre Villard
> Priority: Major
> Fix For: 2.10.0
>
> Time Spent: 1h 20m
> Remaining Estimate: 0h
>
> When {{JsonTreeReader}} feeds {{JsonRecordSetWriter}} (in my case with a
> {{{}ConvertRecord{}}}), the writer's configured {{{}Timestamp Format{}}},
> {{{}Date Format{}}}, {{{}Time Format{}}}, and {{Suppress Null Values}}
> properties are effectively ignored (the writer emits the reader's raw input
> bytes verbatim). This is an internal optimization: when the reader attaches
> the source JSON to the record, the writer detects a matching MIME type and
> short-circuits to the cached string instead of re-serializing from the typed
> values it already has.
>
> The consequence is that, for example, an input timestamp
> {{2025-03-20T17:33:11.000+0000}} passes through unchanged even when the
> writer is explicitly configured with {{yyyy-MM-dd'T'HH:mm:ss.SSSX}} (which
> would emit {{...Z}} for UTC)
>
> The existing optimisation is there for very specific use cases and it should
> be possible to turn it off if needed.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
