Fix/attribute loss when updating a device

[edwinvautier]

related to #1660

It is the continuation of this PR #1667 from https://github.com/GuillaumeOd as he won't work on it anymore.

[Copilot]

Pull Request Overview

This PR addresses an issue where attributes could be lost during device updates when both name and object_id are defined by adjusting the filtering logic and adding a new functional test.

• Added a new test case to cover attributes with special object_id formats and ensure correct type preservation.
• Updated the NGSI-v2 entity update logic to only filter out measurements by object_id.
• Updated CHANGES_NEXT_RELEASE with the fix note.

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

File	Description
test/functional/testCases.js	Added test 0021c for special-named object_id handling
lib/services/ngsi/entities-NGSI-v2.js	Changed filter logic to only drop measurements matching object_id
CHANGES_NEXT_RELEASE	Added entry for attribute-loss fix

Comments suppressed due to low confidence (3)

test/functional/testCases.js:595

• [nitpick] The test ID '0021c' deviates from the established numeric sequence; consider using a consistent numbering scheme for clarity.

        describeName: '0021c - Simple group with active attributes with special names in object_id',

CHANGES_NEXT_RELEASE:1

• [nitpick] Please ensure this changelog entry follows the project's standard format (e.g., include version heading or consistent bullet formatting).

- Fix attribute loss during device update when both name and object_id are defined by filtering only object_id (#1660)

lib/services/ngsi/entities-NGSI-v2.js:470

• By dropping only object_id, entries whose name equals the alias may remain and cause duplicate measures; verify if alias duplicates need handling.

            measures = measures.filter((item) => item.name !== currentAttr.object_id);

[Copilot]

[nitpick] This comment is now outdated since the filter only removes by object_id; please update or remove it to match the new logic.

Suggested change

	//remove measures that has been shadowed by an alias (some may be left and managed later)
	//Maybe we must filter object_id if there is name == object_id
	//remove measures that have been shadowed by an alias (some may be left and managed later)
	//Filter measures by removing items where name matches object_id

[fgalan]

@edwinvautier so PR #1667 can be closed so we can continue in this PR?

[edwinvautier]

@edwinvautier so PR #1667 can be closed so we can continue in this PR?

yes #1667 can be closed

[fgalan]

@edwinvautier so PR #1667 can be closed so we can continue in this PR?

yes #1667 can be closed

Closed

[fgalan]

Let's see how the same test works with current master branch (to fully understand which current behaviour this PR is fixing)

[fgalan]

Let's see how the same test works with current master branch (to fully understand which current behaviour this PR is fixing)

We have create temporal PR #1726 to check that.

In that PR we have copi-paste the same test in this PR in a fresh branch from master without applying the change done in this PR in lib/services/ngsi/entities-NGSI-v2.js.

Surprisingly, no test is failing in that PR. In other words, the change done in this PR in lib/services/ngsi/entities-NGSI-v2.js is unneeded. Or maybe the test is not being run for some reason?

@edwinvautier what do you think?

[fgalan]

Or maybe the test is not being run for some reason?

This confirms that the test is run https://github.com/telefonicaid/iotagent-node-lib/actions/runs/16668949855/job/47180736642#step:5:963

[edwinvautier]

Sorry for the delay, my test case did not cover the fix, should be ok now, i have set a test where object id and other attribute names conflict.

@fgalan should be ok now

[edwinvautier]

Sorry for the delay, my test case did not cover the fix, should be ok now, i have set a test where object id and other attribute names conflict.

@fgalan should be ok now

just made a new change i missed something in my test

[tzzed]

@fgalan can you check please the PR of @edwinvautier who took over from guillaume plz. Thanks to all of you

[AlvaroVega]

Sorry for the delay, my test case did not cover the fix, should be ok now, i have set a test where object id and other attribute names conflict.
@fgalan should be ok now

just made a new change i missed something in my test

Could you please @edwinvautier update the following PR #1726 just to update your new test to show that really tests is failing currently in master branch?