Search: Use default semantic_text inference

[reakaleek]

What

Remove explicit ELSER and Jina inference IDs from semantic_text field mappings, and drop the redundant .jina sub-fields entirely.

Why

Elasticsearch Serverless now defaults semantic_text to Jina, making the explicit Jina sub-fields redundant and the hardcoded ELSER inference endpoint unnecessary.

How

• Removed ElserInferenceId and JinaInferenceId constants from SemanticConfig
• Removed all 5 .jina field mappings
• Changed .semantic_text fields to use f.SemanticText() with no inference ID (platform default)

Test plan

• dotnet build — 0 errors
• dotnet test tests/Elastic.Markdown.Tests — 1591 tests pass
• Verify new index mapping no longer references ELSER/Jina inference IDs on a Serverless deployment

🤖 Generated with Claude Code

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

Run ID: 810668c8-e454-4263-8850-8d29db456a38

📥 Commits

Reviewing files that changed from the base of the PR and between e578fac and 8db9e90.

📒 Files selected for processing (1)

• src/Elastic.Documentation/Search/DocumentationMappingConfig.cs

---

📝 Walkthrough

Walkthrough

The change removes hard-coded inference ID constants for ELSER and Jina models from the semantic configuration. Semantic field mappings for title.semantic_text, abstract.semantic_text, ai_rag_optimized_summary.semantic_text, ai_questions.semantic_text, and ai_use_cases.semantic_text are updated to rely on platform defaults instead of explicitly assigning inference IDs. This eliminates the need for separate Jina-specific field variants under the *.jina namespaces.

Suggested labels

enhancement

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the main change: removing explicit inference IDs and using default semantic_text inference.
Description check	✅ Passed	The description clearly explains what was changed, why it was necessary, and how the changes were implemented, directly relating to the changeset.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

✨ Simplify code

• Create PR with simplified code
• Commit simplified code in branch maddening-badge

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}