Claus Ibsen created CAMEL-27148:
-----------------------------------
Summary: camel-jbang - a write note when the steps of an
idempotentConsumer that skips duplicates check CamelDuplicateMessage
Key: CAMEL-27148
URL: https://issues.apache.org/jira/browse/CAMEL-27148
Project: Camel
Issue Type: Improvement
Components: camel-jbang
Reporter: Claus Ibsen
Assignee: Claus Ibsen
Fix For: 4.24.0
An idempotentConsumer skips a duplicate by default (skipDuplicate is true): the
duplicate never reaches the steps under it. It goes on with the exchange
property CamelDuplicateMessage set to true, but only to the steps after the
idempotentConsumer. A new message has no CamelDuplicateMessage property at all.
So a check of CamelDuplicateMessage in the steps under an idempotentConsumer
that skips duplicates is never true. For example, this route never logs a
duplicate:
{code:yaml}
- idempotentConsumer:
expression:
simple: "${body[orderId]}"
idempotentRepository: orderRepo
steps:
- choice:
when:
- expression:
simple: "${exchangeProperty.CamelDuplicateMessage} == true"
steps:
- log:
message: "Order ${body[orderId]} is a duplicate, skipped"
otherwise:
steps:
- log:
message: "Order ${body[orderId]} accepted"
{code}
A check with == false is never true either, because a new message has no such
property. In one run the when had == false and logged "accepted", so the
otherwise logged every new order as a duplicate.
The fix is skipDuplicate: false. Then a duplicate goes on to the steps under
the idempotentConsumer with CamelDuplicateMessage set to true.
In the local-model benchmark (r30 to r35, Nov 8-18 2026), step 1 of
route-enrich-and-dedupe asks to log the accepted orders and the skipped
duplicates. Every step run that set skipDuplicate: false passed (27). Every
step run that checked CamelDuplicateMessage without it failed (8, two of them
with == false). The catalog does not say it either. skipDuplicate says "Sets
whether to skip duplicates or not. The default behavior is to skip
duplicates.", and the CamelDuplicateMessage property says "Whether this
exchange is a duplicate detected by the Idempotent Consumer EIP".
Proposal: the write tools of the camel-jbang-mcp server (camel_write_file and
camel_edit_file) give a note when the steps under an idempotentConsumer that
skips duplicates check CamelDuplicateMessage. The note says to add
skipDuplicate: false, and to check == true for a duplicate. It is a note like
the other write notes: the write is not refused.
The descriptions of skipDuplicate and of the CamelDuplicateMessage property
could say this too. That changes the model and its generated files, so it is
for another ticket.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)