feat: add Oracle AI Vector Search DocumentStore (oracle-haystack)#3096
Open
fede-kamel wants to merge 2 commits intodeepset-ai:mainfrom
Open
feat: add Oracle AI Vector Search DocumentStore (oracle-haystack)#3096fede-kamel wants to merge 2 commits intodeepset-ai:mainfrom
fede-kamel wants to merge 2 commits intodeepset-ai:mainfrom
Conversation
New integration: oracle-haystack Adds OracleDocumentStore backed by Oracle Database 23ai/26ai native VECTOR type, plus OracleEmbeddingRetriever. Key features: - VECTOR(dim, FLOAT32) column with HNSW approximate search index - Supports Oracle Autonomous Database (ADB-S) wallet connections - All three DuplicatePolicy modes via INSERT / MERGE SQL - Haystack filter grammar translated to JSON_VALUE WHERE clauses - Full async support (asyncio.to_thread wrappers) - Credentials handled via Haystack Secret — never serialised as plaintext - 34 unit tests (mocked oracledb) + 13 integration tests (live Oracle) - e2e validated: 200 SQuAD passages at 181 docs/sec write, <700ms query Closes #<issue>
|
|
github-actions
bot
added
the
type:documentation
Author
|
recheck |
- Add pydoc/config_docusaurus.yml required by CI_check_api_ref workflow - Add hatch docs/fmt scripts to [tool.hatch.envs.default] - Use tag-pattern + git_describe_command matching repo conventions - Add fallback-version = 0.1.0 for new integration with no tags yet - Add [tool.hatch.envs.test] scripts for unit/integration test runs
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #3095
What this PR adds
A new integration package
oracle-haystack(integrations/oracle/) that brings Oracle Database 23ai/26ai native vector search into the Haystack ecosystem.Components
OracleDocumentStoreDocumentStoreprotocol — write, filter, delete, count, async, serializationOracleEmbeddingRetriever@componentwith filter merging,run+arunOracleConnectionConfigSecret-wrapped credentials, ADB-S wallet supportOracle features used
VECTOR(dim, FLOAT32)column — native to Oracle 23ai/26ai, included in DB licenseCREATE VECTOR INDEX … ORGANIZATION INMEMORY NEIGHBOR GRAPH(HNSW)FETCH APPROX FIRST N ROWS ONLY— activates HNSW for approximate searchJSONcolumn +JSON_VALUE(metadata, '$.key')for metadata filteringVARCHAR2(64)primary key — stores any Haystack ID natively (UUID, SHA-256, custom)Haystack filter grammar support
AND/OR/NOT, equality, comparison,in,not in— all translated to Oracle SQLWHEREfragments via_FilterTranslator.Tests
oracledb.create_poolmockedpytest.mark.skipif(not os.getenv("ORACLE_DSN"), ...)tests/e2e_real_data.py) — 200 SQuAD passages via fastembed → Oracle 26ai ADBValidation results (Oracle AI Database 26ai, OCI Free Tier)
Checklist
ruff checkandruff format— cleanto_dict/from_dictroundtrip — passwords never serialized as plaintext_SAFE_TABLE_NAMEregex on table name