[ShousenZHANG]

Summary

Part of #36288. Following #44775, #44853 and #44898, type the MyScale and Elasticsearch JA unit tests and replace their MagicMock objects with real values and small recording fakes.

• test_myscale_vector.py imports the real provider module instead of reloading it under a stub clickhouse_connect module; clickhouse-connect is a dependency of dify-vdb-myscale. Only get_client is replaced: FakeClickHouseClient records the connection arguments and each command and query with its bound parameters. create, _create_collection, add_texts and _search now run their real code instead of being stubbed, so the tests check the generated SQL and parameters end to end.
• test_elasticsearch_ja_vector.py imports the real modules and injects FakeElasticsearch through the real constructor by replacing the Elasticsearch class, so ping() and the version check run too. The fake index API records each creation, redis_client is a small recording fake, and the factory test uses a real Flask app context instead of a SimpleNamespace for current_app.
• The factory tests compare the whole MyScale or Elasticsearch config instead of only the collection name.

Every test parameter and return is annotated. This changes test code only. Importing the real clickhouse_connect emits a DeprecationWarning from the library itself on Python 3.13 (the 'u' array type code).

Validation on Python 3.13:

• All 18 cases pass, with the same test IDs and order as on main. The Elasticsearch JA tests also pass when test_elasticsearch_vector.py, which reloads these modules, runs first.
• Focused pyrefly check with implicit-any-parameter, unannotated-return and implicit-any-empty-container set to error: 50 diagnostics on main under the unit-test configuration, 0 after under both the production and unit-test configurations. CI does not enable these three rules, so they were checked locally.
• 17 deliberate source changes were applied one at a time, such as a dropped SET allow_experimental_object_type, a swapped connection user, a hard-coded LIMIT, a wrong sort order, a non-Japanese analyzer or a shorter cache TTL. The new tests fail for 16; the tests on main fail for 11, all of which the new tests also catch. The remaining change only affects documents with empty metadata, which these tests do not cover.
• Ruff format/check and all six repository guards pass. The diff adds no MagicMock, cast, SimpleNamespace, spec, Any or type ignore.

Tests ran through the local Windows pytest wrapper; the lint and type-check commands were run individually rather than through make.

Screenshots

Before	After
Focused Pyrefly diagnostics: 50	0
Seeded source changes caught: 11 of 17	16 of 17
Lines using MagicMock: 24	0

Checklist

• This change requires a documentation update, included: Dify Document
• I understand that this PR may be closed in case there was no previous discussion or issues. (This doesn't apply to typos!)
• I've verified the change and added or updated tests where meaningful regression risk justifies coverage.
• I've updated the documentation accordingly.
• I ran make lint && make type-check (backend) and vp staged (frontend) to appease the lint gods

From Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota.

An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing. Review the Copilot code review billing settings.

[github-actions]

Pyrefly Type Coverage

Metric	Base	PR	Delta
Type coverage	80.30%	80.34%	+0.05%
Strict coverage	79.98%	80.03%	+0.05%
Typed symbols	77,347	77,454	+107
Untyped symbols	19,054	19,022	-32
Modules	3989	3989	0

[codecov]

Codecov Report

❌ Patch coverage is 96.69421% with 8 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.54%. Comparing base (1f500e5) to head (41f233e).
⚠️ Report is 43 commits behind head on main.

Files with missing lines	Patch %	Lines
..._vdb_elasticsearch/test_elasticsearch_ja_vector.py	95.91%	4 Missing ⚠️
..._tests/src/dify_vdb_myscale/test_myscale_vector.py	97.22%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #44913      +/-   ##
==========================================
+ Coverage   89.52%   89.54%   +0.01%     
==========================================
  Files        6009     6012       +3     
  Lines      376843   377210     +367     
  Branches    76238    76278      +40     
==========================================
+ Hits       337380   337758     +378     
+ Misses      33465    33459       -6     
+ Partials     5998     5993       -5     

Flag	Coverage Δ
api	89.41% <96.69%> (+0.02%)	⬆️

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:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.