fix(backend): fix upsert on non multiple refs (#17511)

[Jeremy Cloarec (JeremyCloarec)]

Proposed changes

This pull request enhances the upsert utilities and their test coverage, specifically focusing on the handling of reference fields in upsert operations. The main change is the export and thorough testing of the generateRefsInputsForUpsert function, ensuring correct behavior for both single and multiple reference fields under various scenarios, including synchronization mode and confidence-based updates.

Upsert Utility Improvements

• Exported the generateRefsInputsForUpsert function from upsert-utils.js for direct use in other modules and tests.
• Fixed the logic for determining when a reference field should be updated, ensuring that updates only occur when there is actual input data and the update is allowed by the current confidence and protection rules.

Test Coverage Enhancements

• Added comprehensive unit tests for generateRefsInputsForUpsert in middleware-test.js, covering:
  • Non-multiple (single) reference fields (e.g., createdBy): tests for not overwriting with null/undefined, proper filling, replacement with higher confidence, protection against flickering, and explicit removal in synchronization mode.
  • Multiple reference fields (e.g., objectLabel): tests for adding missing values, avoiding unnecessary updates, and full replacement in synchronization mode.
  • Ensuring ignored refs not present in the update patch.
• Updated test imports to include the now-exported generateRefsInputsForUpsert.

Related issues

• closes fix(backend): Upsert clears existing createdBy to null when incoming confidence is higher (createdBy ref path lacks emptiness guard) #17511

How to test this PR

Checklist

• I consider the submitted work as finished
• I tested the code for its functionality
• I wrote test cases for the relevant use cases (coverage and e2e)
• I added/updated the relevant documentation (either on GitHub or on Notion)
• Where necessary, I refactored code to improve the overall quality

Further comments

[Copilot]

Pull request overview

This PR fixes confidence-based upsert behavior for single-reference fields (e.g., createdBy) to prevent “data cleaning” (clearing an existing ref when the incoming ref is empty), and it exports generateRefsInputsForUpsert to enable direct unit testing.

Changes:

• Exported generateRefsInputsForUpsert from upsert-utils.js.
• Tightened single-ref upsert conditions so updates only occur when the input actually contains data (unless in full synchronization mode).
• Added unit tests covering single vs multiple ref upsert scenarios, including sync mode and confidence/protection cases.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
opencti-platform/opencti-graphql/src/utils/upsert-utils.js	Exports generateRefsInputsForUpsert and fixes the single-ref “updatable” condition to avoid null/empty overwrites outside sync mode.
opencti-platform/opencti-graphql/tests/01-unit/database/middleware-test.js	Adds unit tests for generateRefsInputsForUpsert across single/multiple refs and sync mode behaviors.

Suppressed comments (6)

opencti-platform/opencti-graphql/tests/01-unit/database/middleware-test.js:433

• Same as above: to reflect actual resolved store shape, resolvedElement['created-by'] should be the author internal_id string, not the full object.

        const resolvedElement = { ...baseIndicator, 'created-by': authorA };

opencti-platform/opencti-graphql/tests/01-unit/database/middleware-test.js:454

• Same as above: resolvedElement['created-by'] should be a stored internal_id string.

        const resolvedElement = { ...baseIndicator, 'created-by': authorA };

opencti-platform/opencti-graphql/tests/01-unit/database/middleware-test.js:464

• Same as above: resolvedElement['created-by'] should be a stored internal_id string.

        const resolvedElement = { ...baseIndicator, 'created-by': authorA };

opencti-platform/opencti-graphql/tests/01-unit/database/middleware-test.js:474

• Same as above: resolvedElement['created-by'] should be a stored internal_id string.

        const resolvedElement = { ...baseIndicator, 'created-by': authorA };

opencti-platform/opencti-graphql/tests/01-unit/database/middleware-test.js:485

• Same as above: resolvedElement['created-by'] should be a stored internal_id string.

        const resolvedElement = { ...baseIndicator, 'created-by': authorA };

opencti-platform/opencti-graphql/tests/01-unit/database/middleware-test.js:545

• Same as above: even in this test, resolvedElement['created-by'] should reflect the stored internal_id string.

      const resolvedElement = { ...baseIndicator, 'created-by': authorA };

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 34.07%. Comparing base (6ce2211) to head (942cb17).
⚠️ Report is 3 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master   #17518      +/-   ##
==========================================
- Coverage   34.08%   34.07%   -0.02%     
==========================================
  Files        3376     3376              
  Lines      137588   137588              
  Branches    37228    37228              
==========================================
- Hits        46891    46877      -14     
- Misses      90697    90711      +14     

Flag	Coverage Δ
opencti-client-python	48.37% <ø> (ø)
opencti-front	10.98% <ø> (ø)
opencti-graphql	69.11% <100.00%> (-0.03%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.