alloy-runtime-types 0.2.76__tar.gz → 0.2.78__tar.gz

{alloy_runtime_types-0.2.76 → alloy_runtime_types-0.2.78}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alloy-runtime-types
-Version: 0.2.76
+Version: 0.2.78
 Summary: Type definitions and DTOs for Alloy Runtime
 Keywords: alloy,api,dto,python,types
 Classifier: Development Status :: 3 - Alpha

{alloy_runtime_types-0.2.76 → alloy_runtime_types-0.2.78}/alloy_runtime_types/dtos/knowledge.py

@@ -343,6 +343,10 @@ class CollectionDocumentStatusSummary(BaseModel):
     failed_documents: int
     duplicate_documents: int
     filtered_documents: int
+    staged_documents: int = Field(
+        default=0,
+        description="Documents accepted into bulk ingest intake but not yet promoted to the processing queue.",
+    )
 
 
 class ListCollectionDocumentStatusesResponse(BaseModel):
@@ -533,8 +537,8 @@ class SpeakerConfig(BaseModel):
         return self
 
 
-class IngestDocumentRequest(BaseModel):
-    """Request to ingest a document.
+class IngestDocumentItem(BaseModel):
+    """One document item in a bulk ingestion request.
 
     Preprocessing pipeline: Raw content → NormalizationStage → ContentTypeProcessor
     (transcript/markdown/document) → StructuralNormalizationStage → QualityFilterStage
@@ -545,9 +549,8 @@ class IngestDocumentRequest(BaseModel):
     removes page numbers, headers/footers, and fixes hyphenation.
     """
 
-    collection_id: str | UUID = Field(
-        ..., description="Target collection identifier - can be UUID or collection name"
-    )
+    model_config = ConfigDict(extra="forbid")
+
     title: str = Field(..., min_length=1, max_length=500, description="Document title")
     content: str = Field(..., min_length=1, description="Document text content")
     question_gen_agent_id: str | UUID | None = Field(
@@ -630,11 +633,6 @@ class IngestDocumentRequest(BaseModel):
         "meeting date, publication date). If omitted, falls back to metadata['source_date'] if present "
         "and parseable. Stored as UTC timestamptz.",
     )
-    dry_run: bool = Field(
-        default=False,
-        description="If True, only preprocess the document and return the result without persisting. "
-        "Useful for testing preprocessing pipeline output.",
-    )
 
     @model_validator(mode="after")
     def validate_speaker_config_for_transcript(self) -> Self:
@@ -670,32 +668,111 @@ class IngestDocumentRequest(BaseModel):
         return self
 
 
-class IngestDocumentResponse(BaseModel):
-    """Response from ingesting a document.
+class IngestDocumentsRequest(BaseModel):
+    """Request to stage one or more documents for ingestion."""
 
-    New documents are usually queued for background processing.
-    Exact duplicate replays may instead return the existing document directly.
+    collection_id: str | UUID = Field(
+        ..., description="Target collection identifier - can be UUID or collection name"
+    )
+    documents: list[IngestDocumentItem] = Field(
+        ...,
+        min_length=1,
+        max_length=500,
+        description="Documents to ingest in request order.",
+    )
 
-    When dry_run=True, includes processed_content and diff fields.
-    """
 
-    document_id: UUID | None = None  # None when dry_run=True
-    title: str
-    status: IngestResponseStatus
-    message: str
+class IngestDocumentRequest(IngestDocumentItem):
+    """Internal single-document promotion request used by ingestion workers."""
+
+    collection_id: str | UUID = Field(
+        ..., description="Target collection identifier - can be UUID or collection name"
+    )
+    dry_run: bool = Field(
+        default=False,
+        description="Internal single-document preprocessing preview flag.",
+    )
+
+
+BulkIngestItemStatus = Literal[
+    "staged",
+    "duplicate_external_id",
+    "duplicate_content",
+    "duplicate_within_batch",
+    "queued",
+    "filtered",
+    "failed",
+]
+
+BulkIngestOperationStatus = Literal[
+    "accepted",
+    "dispatching",
+    "waiting_for_capacity",
+    "completed",
+    "failed",
+]
+
+
+class IngestDocumentItemResult(BaseModel):
+    ordinal: int
+    item_id: UUID
     external_id: str | None = None
-    # Dry run fields
-    dry_run: bool = False
-    processed_content: str | None = None
-    preprocessing_metadata: dict[str, Any] | None = None
-    rejected: bool = False
-    rejection_reason: str | None = None
-    # Diff fields (only populated when dry_run=True)
-    diff_raw: list[dict[str, Any]] | None = None
-    diff_unified: str | None = None
-    diff_html: str | None = None
-    diff_markers: str | None = None
-    diff_stats: dict[str, Any] | None = None
+    status: BulkIngestItemStatus
+    result_reason: str | None = None
+    document_id: UUID | None = None
+
+
+class IngestDocumentsResponse(BaseModel):
+    """Response from staging a bulk document ingestion operation."""
+
+    operation_id: UUID
+    status: BulkIngestOperationStatus
+    submitted_count: int
+    staged_count: int
+    duplicate_external_id_count: int
+    duplicate_content_count: int
+    duplicate_within_batch_count: int
+    queued_count: int = 0
+    failed_count: int = 0
+    items: list[IngestDocumentItemResult]
+
+
+IngestDocumentResponse = IngestDocumentsResponse
+
+
+class BulkIngestOperationItemStatus(BaseModel):
+    item_id: UUID
+    ordinal: int
+    external_id: str | None = None
+    status: BulkIngestItemStatus
+    result_reason: str | None = None
+    document_id: UUID | None = None
+    created_at: datetime
+    updated_at: datetime
+    queued_at: datetime | None = None
+    completed_at: datetime | None = None
+
+
+class GetBulkIngestOperationResponse(BaseModel):
+    operation_id: UUID
+    collection_id: UUID
+    status: BulkIngestOperationStatus
+    submitted_count: int
+    staged_count: int
+    duplicate_external_id_count: int
+    duplicate_content_count: int
+    duplicate_within_batch_count: int
+    queued_count: int
+    failed_count: int
+    queued_at: datetime
+    started_at: datetime | None = None
+    last_heartbeat_at: datetime | None = None
+    completed_at: datetime | None = None
+    error_message: str | None = None
+    items: list[BulkIngestOperationItemStatus]
+    limit: int
+    offset: int
+    total_items: int
 
 
 class DocumentSummary(BaseModel):

alloy_runtime_types-0.2.78/alloy_runtime_types/dtos/retrieval_diagnostics.py

@@ -0,0 +1,72 @@
+"""Public retrieval diagnostics DTO consumed by SDK, CLI, and server code."""
+
+from __future__ import annotations
+
+from typing import Literal
+
+from pydantic import BaseModel, ConfigDict, Field, computed_field, model_validator
+
+DiagnosticSeverity = Literal["info", "warning", "error"]
+
+DiagnosticCode = Literal[
+    "decomposition_timeout",
+    "decomposition_failed",
+    "hyde_timeout",
+    "hyde_fallback",
+    "voyage_degraded",
+    "semantic_no_results",
+    "sparse_overfetch_exhausted",
+    "question_low_headroom",
+    "subquery_fanout_timeout",
+    "subquery_fanout_failed",
+    "tool_cancelled",
+    "tool_errored",
+    "no_collections_configured",
+    "configured_collections_not_found",
+    "collection_filter_not_found",
+]
+
+DiagnosticStatus = Literal["ok", "degraded", "failed"]
+
+
+def _empty_warnings() -> list["RetrievalDiagnosticWarning"]:
+    return []
+
+
+class RetrievalDiagnosticWarning(BaseModel):
+    model_config = ConfigDict(extra="forbid")
+
+    code: DiagnosticCode
+    severity: DiagnosticSeverity
+    message: str
+    fields: dict[str, object] = Field(default_factory=dict)
+
+
+class RetrievalDiagnostics(BaseModel):
+    model_config = ConfigDict(extra="forbid")
+
+    warnings: list[RetrievalDiagnosticWarning] = Field(default_factory=_empty_warnings)
+
+    @model_validator(mode="before")
+    @classmethod
+    def _discard_computed_round_trip_fields(
+        cls, data: dict[str, object]
+    ) -> dict[str, object]:
+        cleaned = dict(data)
+        cleaned.pop("status", None)
+        cleaned.pop("degraded", None)
+        return cleaned
+
+    @computed_field
+    @property
+    def status(self) -> DiagnosticStatus:
+        if any(w.severity == "error" for w in self.warnings):
+            return "failed"
+        if self.warnings:
+            return "degraded"
+        return "ok"
+
+    @computed_field
+    @property
+    def degraded(self) -> bool:
+        return self.status != "ok"

{alloy_runtime_types-0.2.76 → alloy_runtime_types-0.2.78}/alloy_runtime_types.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alloy-runtime-types
-Version: 0.2.76
+Version: 0.2.78
 Summary: Type definitions and DTOs for Alloy Runtime
 Keywords: alloy,api,dto,python,types
 Classifier: Development Status :: 3 - Alpha

{alloy_runtime_types-0.2.76 → alloy_runtime_types-0.2.78}/alloy_runtime_types.egg-info/SOURCES.txt

@@ -30,6 +30,7 @@ alloy_runtime_types/dtos/organizations.py
 alloy_runtime_types/dtos/pipeline.py
 alloy_runtime_types/dtos/pipeline_costs.py
 alloy_runtime_types/dtos/render.py
+alloy_runtime_types/dtos/retrieval_diagnostics.py
 alloy_runtime_types/dtos/schedule.py
 alloy_runtime_types/dtos/schemas.py
 alloy_runtime_types/dtos/sessions.py

{alloy_runtime_types-0.2.76 → alloy_runtime_types-0.2.78}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "alloy-runtime-types"
-version = "0.2.76"
+version = "0.2.78"
 description = "Type definitions and DTOs for Alloy Runtime"
 readme = "README.md"
 requires-python = ">=3.13"