Add direct-text semantic SearchParameter MVP with Foundry - #5803
Open
Brendan Kowitz (brendankowitz) wants to merge 5 commits into
Conversation
Brendan Kowitz (brendankowitz)
requested a review
from a team
as a code owner
September 11, 2026 01:21
Brendan Kowitz (brendankowitz)
added this pull request to stack #5804
September 11, 2026 01:22
Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Anna Gornyitzki <t-annag@microsoft.com> Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Anna Gornyitzki <t-annag@microsoft.com> Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Anna Gornyitzki <t-annag@microsoft.com> Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Co-authored-by: Anna Gornyitzki <t-annag@microsoft.com> Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
Brendan Kowitz (brendankowitz)
force-pushed
the
feature/sql-semantic-search-stack/layer-2-mvp
branch
from
September 11, 2026 01:33
077c543 to
2f7c976
Compare
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
| throw new InvalidDefinitionException($"Vector SearchParameter distanceMetric must be '{VectorSearchConfiguration.SupportedDistanceMetric}'."); | ||
| } | ||
|
|
||
| configuration.DistanceMetric = distanceMetric.ToLowerInvariant(); |
| { | ||
| var bundle = context.Object as Hl7.Fhir.Model.Bundle; | ||
| resource = bundle; | ||
| bool hasExtendedSearchMetadata = bundle.Entry |
| if (isVectorSearch && sqlSearchOptions.SemanticContinuationDistance.HasValue) | ||
| { | ||
| object distanceParameter = Parameters.AddParameter(sqlSearchOptions.SemanticContinuationDistance.Value, includeInHash: false); | ||
| object resourceTypeParameter = Parameters.AddParameter(sqlSearchOptions.SemanticContinuationResourceTypeId.Value, includeInHash: false); |
| { | ||
| object distanceParameter = Parameters.AddParameter(sqlSearchOptions.SemanticContinuationDistance.Value, includeInHash: false); | ||
| object resourceTypeParameter = Parameters.AddParameter(sqlSearchOptions.SemanticContinuationResourceTypeId.Value, includeInHash: false); | ||
| object surrogateIdParameter = Parameters.AddParameter(sqlSearchOptions.SemanticContinuationResourceSurrogateId.Value, includeInHash: false); |
| await _initializationLock.WaitAsync(cancellationToken); | ||
| try | ||
| { | ||
| if (_embeddingModelId.HasValue) |
Comment on lines
+318
to
+326
| if (eTag.HasValue) | ||
| { | ||
| // You can't update a resource with a specified version if the resource does not exist | ||
| if (weakETag != null) | ||
| { | ||
| results.Add(resourceExt.GetIdentifier(), new DataStoreOperationOutcome(new ResourceNotFoundException(string.Format(Core.Resources.ResourceNotFoundByIdAndVersion, resource.ResourceTypeName, resource.ResourceId, weakETag.VersionId)))); | ||
| continue; | ||
| } | ||
| } |
Comment on lines
+533
to
+535
| catch (OperationCanceledException) when (heartbeatCancellationSource.IsCancellationRequested) | ||
| { | ||
| } |
Comment on lines
+556
to
+558
| catch (OperationCanceledException) when (cancellationToken.IsCancellationRequested) | ||
| { | ||
| } |
| { | ||
| var projectDir = GetProjectPath("src", startupType); | ||
| var testConfigPath = Path.GetFullPath("testconfiguration.json"); | ||
| var testConfigPath = Path.Combine(AppContext.BaseDirectory, "testconfiguration.json"); |
|
|
||
| configuration["ASPNETCORE_FORWARDEDHEADERS_ENABLED"] = "true"; | ||
| configuration["TestAuthEnvironment:FilePath"] = "testauthenvironment.json"; | ||
| configuration["TestAuthEnvironment:FilePath"] = Path.Combine(AppContext.BaseDirectory, "testauthenvironment.json"); |
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
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.
Description
Layer 2 of the SQL semantic-search stack. Depends on #5802.
Implements metadata-driven special SearchParameters and vector-search configuration, direct FHIRPath text extraction to synchronous Foundry embeddings, and transactional vector writes. Covers activation, reindexing, empty-text cleanup, awaited heartbeat lifetime through failures and cancellation, model-compatible tokenization with strict Unicode-safe caps, ordered Foundry input/token batching, ordinary filters and authorization, cosine ranking with deduplication, stable pagination, and standard
Bundle.entry.search.score.The feature is SQL-only and off by default.
Binary, PDF, and semantic chains are rejected before embedding. Layer 3 evidence, the Patient operation, Binary/PDF support, and refresh functionality remain unpublished. Extraction commits were authored by Anna Gornyitzki; adaptations are separate. Source: #5780.Documentation: Semantic Search and ADR 2608.
Related issues
Addresses #5780.
Testing
Recorded prior evidence: the Layer 2 solution build succeeded; Core and SQL unit suites plus targeted Shared R4, API, Azure, and schema tests passed; tokenizer Unicode/overlap/default-override tests passed; Foundry failure/cancellation/cardinality/batching tests passed; and the final isolated Layer 2 semantic E2E suite (5 tests) passed, including heartbeat coverage. No live Foundry testing was performed.
FHIR Team Checklist
Semver Change (docs)
Feature