[MOHAMMED WASIM KHAN (wasim-builds)]

Motivation and Context

Closes #14684

When data and vector fields define storage_name aliases (e.g. content -> body, embedding -> vec), Azure AI Search indexes are created using the storage names. However, AzureAISearchCollection.hybrid_search() was using field.name / vector_field.name instead of storage_name or name when configuring search_fields and vector_queries[0].fields. As a result, Azure AI Search rejected the hybrid search request with a 400 unknown field error because the emitted request referenced model property names not present in the index.

In contrast, SearchType.VECTOR correctly uses vector_field.storage_name or vector_field.name.

Description

• In AzureAISearchCollection._inner_search for SearchType.KEYWORD_HYBRID:
  • When options.additional_property_name is specified, resolve it to the field's storage_name or name if it matches a model field name or storage name alias.
  • When options.additional_property_name is omitted, collect field.storage_name or field.name for all full-text indexed data fields.
  • In both VectorizedQuery and VectorizableTextQuery, configure fields=vector_field.storage_name or vector_field.name if vector_field else None, matching the vector search branch.
• Added comprehensive unit tests in test_azure_ai_search.py asserting that hybrid search and vector search correctly use storage name aliases for search fields and vector query fields.

Contribution Checklist

• The code builds clean without any errors or warnings
• The PR follows the SK Contribution Guidelines and the pre-submission formatting script raises no violations
• All unit tests pass, and I have added new tests to cover the code changes
• I didn't break anyone 😄

[Copilot]

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