feat: add support for elasticsearch sparse inference embedding retriever

[srini047]

Related Issues

• fixes ElasticsearchInferenceSparseRetriever (ES inference for sparse) #2942

Proposed Changes:

Add support for ElasticsearchInferenceSparseRetriever which uses inference_id and query for retrieval.

How did you test it?

Add tests supporting the new ElasticsearchInferenceSparseRetriever.

Notes for the reviewer

• Should integration tests all be added?
• Use from haystack.core.serialization import component_from_dict, component_to_dict for serialization instead of to_dict and from_dict methods?

Checklist

• I have read the contributors guidelines and the code of conduct
• I have updated the related issue with new insights and changes
• I added unit tests and updated the docstrings
• I've used one of the conventional commit types for my PR title: fix:, feat:, build:, chore:, ci:, docs:, style:, refactor:, perf:, test:.

[github-actions]

Coverage report (elasticsearch)

Click to see where and how coverage changed

File	Statements	Missing	Coverage	Coverage (new stmts)	Lines missing
integrations/elasticsearch/src/haystack_integrations/components/retrievers/elasticsearch
bm25_retriever.py
embedding_retriever.py
inference_sparse_retriever.py
sparse_embedding_retriever.py
integrations/elasticsearch/src/haystack_integrations/document_stores/elasticsearch
document_store.py					562-563, 566-567
filters.py
Project Total

This report was generated by python-coverage-comment-action

[davidsbatista]

@srini047 thanks for another contribution, I did a few more adjustments, mainly testing against an Elastic Search cloud account.

inference_sparse_retriever.py

• Removed unused logger and logging import
• Removed dead SparseEmbedding import from docstring example
• Fixed :raises ValueError: docstring to cover both document_store and inference_id cases

document_store.py

• Fixed _deserialize_document to handle ELSER string token keys gracefully (try/except around int() conversion)
• Removed redundant inference_id validation from _create_sparse_retrieval_inference_body

test_inference_sparse_retriever.py

• Fixed test_run_async: assert_called_once_with → assert_awaited_once_with
• Added full integration test suite (TestElasticsearchInferenceSparseRetrieverIntegration) covering basic retrieval, top_k, filters, filter policies, async, and document shape
• Fixed _index_documents_with_inference to flatten meta keys to top level (matching Haystack's Document.to_dict() layout)
• Batched inference API calls: one round-trip per test instead of one per document
• Fixed test_retrieval_respects_top_k to assert results were actually found (0 < len <= 2)

test_document_store.py

• Added test_sparse_vector_retrieval_inference_builds_query_with_filters
• Added test_sparse_vector_retrieval_inference_async_builds_query_with_filters

conftest.py

• Added inference_sparse_document_store fixture (auto-skips when cloud credentials are absent)

[davidsbatista]

Looks good! As I mentioned earlier, I did some adjustments.