flock-core 0.5.10__py3-none-any.whl → 0.5.20__py3-none-any.whl

flock/orchestrator/tracing.py

@@ -0,0 +1,147 @@
+"""Unified tracing utilities for orchestrator workflows.
+
+Handles OpenTelemetry workflow spans and trace database management.
+Extracted from orchestrator.py to reduce complexity.
+"""
+
+from __future__ import annotations
+
+from contextlib import asynccontextmanager
+from pathlib import Path
+from typing import TYPE_CHECKING, Any
+
+from opentelemetry import trace
+from opentelemetry.trace import Status, StatusCode
+
+
+if TYPE_CHECKING:
+    from collections.abc import AsyncGenerator
+
+
+class TracingManager:
+    """Manages unified tracing for orchestrator workflows.
+
+    Provides workflow span creation and trace database cleanup utilities.
+    """
+
+    def __init__(self) -> None:
+        """Initialize tracing manager."""
+        self._workflow_span = None
+
+    @asynccontextmanager
+    async def traced_run(
+        self, name: str = "workflow", flock_id: str | None = None
+    ) -> AsyncGenerator[Any, None]:
+        """Context manager for wrapping an entire execution in a single unified trace.
+
+        This creates a parent span that encompasses all operations (publish, run_until_idle, etc.)
+        within the context, ensuring they all belong to the same trace_id for better observability.
+
+        Args:
+            name: Name for the workflow trace (default: "workflow")
+            flock_id: Optional Flock instance ID for attribution
+
+        Yields:
+            The workflow span for optional manual attribute setting
+
+        Examples:
+            # Explicit workflow tracing (recommended)
+            async with tracing_manager.traced_run("pizza_workflow"):
+                await flock.publish(pizza_idea)
+                await flock.run_until_idle()
+                # All operations now share the same trace_id!
+
+            # Custom attributes
+            async with tracing_manager.traced_run("data_pipeline") as span:
+                span.set_attribute("pipeline.version", "2.0")
+                await flock.publish(data)
+                await flock.run_until_idle()
+        """
+        tracer = trace.get_tracer(__name__)
+        with tracer.start_as_current_span(name) as span:
+            # Set workflow-level attributes
+            span.set_attribute("flock.workflow", True)
+            span.set_attribute("workflow.name", name)
+            if flock_id:
+                span.set_attribute("workflow.flock_id", flock_id)
+
+            # Store span for nested operations to use
+            prev_workflow_span = self._workflow_span
+            self._workflow_span = span
+
+            try:
+                yield span
+                span.set_status(Status(StatusCode.OK))
+            except Exception as e:
+                span.set_status(Status(StatusCode.ERROR, str(e)))
+                span.record_exception(e)
+                raise
+            finally:
+                # Restore previous workflow span
+                self._workflow_span = prev_workflow_span
+
+    @property
+    def current_workflow_span(self) -> Any:
+        """Get the current workflow span (for nested operations)."""
+        return self._workflow_span
+
+    @staticmethod
+    def clear_traces(db_path: str = ".flock/traces.duckdb") -> dict[str, Any]:
+        """Clear all traces from the DuckDB database.
+
+        Useful for resetting debug sessions or cleaning up test data.
+
+        Args:
+            db_path: Path to the DuckDB database file (default: ".flock/traces.duckdb")
+
+        Returns:
+            Dictionary with operation results:
+                - deleted_count: Number of spans deleted
+                - success: Whether operation succeeded
+                - error: Error message if failed
+
+        Examples:
+            # Clear all traces
+            result = TracingManager.clear_traces()
+            print(f"Deleted {result['deleted_count']} spans")
+
+            # Custom database path
+            result = TracingManager.clear_traces(".flock/custom_traces.duckdb")
+
+            # Check if operation succeeded
+            if result['success']:
+                print("Traces cleared successfully!")
+            else:
+                print(f"Error: {result['error']}")
+        """
+        try:
+            import duckdb
+
+            db_file = Path(db_path)
+            if not db_file.exists():
+                return {
+                    "success": False,
+                    "deleted_count": 0,
+                    "error": f"Database file not found: {db_path}",
+                }
+
+            # Connect and clear
+            conn = duckdb.connect(str(db_file))
+            try:
+                # Get count before deletion
+                count_result = conn.execute("SELECT COUNT(*) FROM spans").fetchone()
+                deleted_count = count_result[0] if count_result else 0
+
+                # Delete all spans
+                conn.execute("DELETE FROM spans")
+
+                # Vacuum to reclaim space
+                conn.execute("VACUUM")
+
+                return {"success": True, "deleted_count": deleted_count, "error": None}
+
+            finally:
+                conn.close()
+
+        except Exception as e:
+            return {"success": False, "deleted_count": 0, "error": str(e)}

flock/storage/__init__.py

@@ -0,0 +1,10 @@
+"""Storage backends for Flock blackboard."""
+
+from flock.storage.sqlite.query_builder import SQLiteQueryBuilder
+from flock.storage.sqlite.schema_manager import SQLiteSchemaManager
+
+
+__all__ = [
+    "SQLiteQueryBuilder",
+    "SQLiteSchemaManager",
+]

flock/storage/artifact_aggregator.py

@@ -0,0 +1,158 @@
+"""Artifact aggregation utilities for summary statistics.
+
+Handles aggregation logic for artifact collections, computing statistics
+like type distribution, producer counts, and time ranges.
+"""
+
+from __future__ import annotations
+
+from datetime import datetime
+from typing import Any
+
+from flock.core.artifacts import Artifact
+from flock.utils.time_utils import format_time_span
+
+
+class ArtifactAggregator:
+    """
+    Aggregates artifact statistics for summary reports.
+
+    Provides clean separation of aggregation logic from storage implementations.
+    Each aggregation method is simple and focused.
+    """
+
+    def aggregate_by_type(self, artifacts: list[Artifact]) -> dict[str, int]:
+        """
+        Count artifacts by type.
+
+        Args:
+            artifacts: List of artifacts to aggregate
+
+        Returns:
+            Dict mapping type names to counts
+        """
+        by_type: dict[str, int] = {}
+        for artifact in artifacts:
+            by_type[artifact.type] = by_type.get(artifact.type, 0) + 1
+        return by_type
+
+    def aggregate_by_producer(self, artifacts: list[Artifact]) -> dict[str, int]:
+        """
+        Count artifacts by producer.
+
+        Args:
+            artifacts: List of artifacts to aggregate
+
+        Returns:
+            Dict mapping producer names to counts
+        """
+        by_producer: dict[str, int] = {}
+        for artifact in artifacts:
+            by_producer[artifact.produced_by] = (
+                by_producer.get(artifact.produced_by, 0) + 1
+            )
+        return by_producer
+
+    def aggregate_by_visibility(self, artifacts: list[Artifact]) -> dict[str, int]:
+        """
+        Count artifacts by visibility kind.
+
+        Args:
+            artifacts: List of artifacts to aggregate
+
+        Returns:
+            Dict mapping visibility kinds to counts
+        """
+        by_visibility: dict[str, int] = {}
+        for artifact in artifacts:
+            kind = getattr(artifact.visibility, "kind", "Unknown")
+            by_visibility[kind] = by_visibility.get(kind, 0) + 1
+        return by_visibility
+
+    def aggregate_tags(self, artifacts: list[Artifact]) -> dict[str, int]:
+        """
+        Count tag occurrences across artifacts.
+
+        Args:
+            artifacts: List of artifacts to aggregate
+
+        Returns:
+            Dict mapping tag names to occurrence counts
+        """
+        tag_counts: dict[str, int] = {}
+        for artifact in artifacts:
+            for tag in artifact.tags:
+                tag_counts[tag] = tag_counts.get(tag, 0) + 1
+        return tag_counts
+
+    def get_date_range(
+        self, artifacts: list[Artifact]
+    ) -> tuple[datetime | None, datetime | None]:
+        """
+        Find earliest and latest creation times.
+
+        Args:
+            artifacts: List of artifacts to analyze
+
+        Returns:
+            Tuple of (earliest, latest) datetimes, or (None, None) if empty
+        """
+        if not artifacts:
+            return None, None
+
+        earliest: datetime | None = None
+        latest: datetime | None = None
+
+        for artifact in artifacts:
+            if earliest is None or artifact.created_at < earliest:
+                earliest = artifact.created_at
+            if latest is None or artifact.created_at > latest:
+                latest = artifact.created_at
+
+        return earliest, latest
+
+    def build_summary(
+        self,
+        artifacts: list[Artifact],
+        total: int,
+        is_full_window: bool,
+    ) -> dict[str, Any]:
+        """
+        Build complete summary statistics for artifacts.
+
+        Args:
+            artifacts: List of artifacts to summarize
+            total: Total count (may differ from len(artifacts) if paginated)
+            is_full_window: Whether this represents all artifacts (no filters)
+
+        Returns:
+            Dictionary with complete summary statistics:
+            - total: Total artifact count
+            - by_type: Type distribution
+            - by_producer: Producer distribution
+            - by_visibility: Visibility distribution
+            - tag_counts: Tag occurrence counts
+            - earliest_created_at: ISO string of earliest artifact
+            - latest_created_at: ISO string of latest artifact
+            - is_full_window: Whether all artifacts included
+            - window_span_label: Human-readable time span
+        """
+        by_type = self.aggregate_by_type(artifacts)
+        by_producer = self.aggregate_by_producer(artifacts)
+        by_visibility = self.aggregate_by_visibility(artifacts)
+        tag_counts = self.aggregate_tags(artifacts)
+        earliest, latest = self.get_date_range(artifacts)
+
+        window_span_label = format_time_span(earliest, latest)
+
+        return {
+            "total": total,
+            "by_type": by_type,
+            "by_producer": by_producer,
+            "by_visibility": by_visibility,
+            "tag_counts": tag_counts,
+            "earliest_created_at": earliest.isoformat() if earliest else None,
+            "latest_created_at": latest.isoformat() if latest else None,
+            "is_full_window": is_full_window,
+            "window_span_label": window_span_label,
+        }

flock/storage/in_memory/__init__.py

@@ -0,0 +1,6 @@
+"""In-memory storage implementation utilities."""
+
+from __future__ import annotations
+
+
+__all__ = ["ArtifactFilter", "HistoryAggregator"]

flock/storage/in_memory/artifact_filter.py

@@ -0,0 +1,114 @@
+"""Artifact filtering utilities for in-memory storage.
+
+Provides focused filtering logic for InMemoryBlackboardStore.query_artifacts.
+Extracted from store.py to reduce complexity from B (10) to A (4).
+"""
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+
+
+if TYPE_CHECKING:
+    from flock.core.artifacts import Artifact
+    from flock.core.store import FilterConfig
+
+
+class ArtifactFilter:
+    """
+    Filter artifacts based on FilterConfig criteria.
+
+    Separates filtering logic from query orchestration for better
+    testability and reduced complexity.
+    """
+
+    def __init__(self, filters: FilterConfig):
+        """
+        Initialize filter with configuration.
+
+        Args:
+            filters: Filter configuration with optional criteria
+        """
+        from flock.registry import type_registry
+
+        # Pre-resolve canonical types once
+        self.canonical_types: set[str] | None = None
+        if filters.type_names:
+            self.canonical_types = {
+                type_registry.resolve_name(name) for name in filters.type_names
+            }
+
+        self.produced_by = filters.produced_by or set()
+        self.correlation_id = filters.correlation_id
+        self.tags = filters.tags or set()
+        self.visibility_kinds = filters.visibility or set()
+        self.start = filters.start
+        self.end = filters.end
+
+    def matches(self, artifact: Artifact) -> bool:
+        """
+        Check if artifact matches all filter criteria.
+
+        Uses focused helper methods to keep complexity low (A-rated).
+        Each criterion is evaluated independently for clarity.
+
+        Args:
+            artifact: Artifact to check against filters
+
+        Returns:
+            True if artifact matches all criteria, False otherwise
+
+        Examples:
+            >>> filter = ArtifactFilter(FilterConfig(produced_by={"agent1"}))
+            >>> artifact = Artifact(type="Result", produced_by="agent1", ...)
+            >>> filter.matches(artifact)
+            True
+        """
+        return (
+            self._matches_type(artifact)
+            and self._matches_producer(artifact)
+            and self._matches_correlation(artifact)
+            and self._matches_tags(artifact)
+            and self._matches_visibility(artifact)
+            and self._matches_time_range(artifact)
+        )
+
+    def _matches_type(self, artifact: Artifact) -> bool:
+        """Check if artifact type matches filter."""
+        if not self.canonical_types:
+            return True
+        return artifact.type in self.canonical_types
+
+    def _matches_producer(self, artifact: Artifact) -> bool:
+        """Check if artifact producer matches filter."""
+        if not self.produced_by:
+            return True
+        return artifact.produced_by in self.produced_by
+
+    def _matches_correlation(self, artifact: Artifact) -> bool:
+        """Check if artifact correlation ID matches filter."""
+        if not self.correlation_id:
+            return True
+        if artifact.correlation_id is None:
+            return False
+        return str(artifact.correlation_id) == self.correlation_id
+
+    def _matches_tags(self, artifact: Artifact) -> bool:
+        """Check if artifact has all required tags."""
+        if not self.tags:
+            return True
+        return self.tags.issubset(artifact.tags)
+
+    def _matches_visibility(self, artifact: Artifact) -> bool:
+        """Check if artifact visibility kind matches filter."""
+        if not self.visibility_kinds:
+            return True
+        return artifact.visibility.kind in self.visibility_kinds
+
+    def _matches_time_range(self, artifact: Artifact) -> bool:
+        """Check if artifact creation time is within range."""
+        if self.start and artifact.created_at < self.start:
+            return False
+        if self.end and artifact.created_at > self.end:
+            return False
+        return True

flock/storage/in_memory/history_aggregator.py

@@ -0,0 +1,115 @@
+"""Agent history aggregation for in-memory storage.
+
+Handles aggregation of produced/consumed artifacts for agent history summaries.
+Extracted from store.py to reduce complexity from B (7) to A (4).
+"""
+
+from __future__ import annotations
+
+from collections import defaultdict
+from typing import TYPE_CHECKING, Any
+
+
+if TYPE_CHECKING:
+    from flock.core.store import ArtifactEnvelope
+
+
+class HistoryAggregator:
+    """
+    Aggregate agent history from artifact envelopes.
+
+    Provides focused aggregation methods for produced and consumed artifacts,
+    keeping complexity low through functional patterns.
+    """
+
+    def aggregate(
+        self, envelopes: list[ArtifactEnvelope], agent_id: str
+    ) -> dict[str, Any]:
+        """
+        Aggregate produced and consumed statistics for an agent.
+
+        Args:
+            envelopes: List of artifact envelopes with consumptions
+            agent_id: Agent to aggregate history for
+
+        Returns:
+            Dictionary with produced and consumed statistics:
+            {
+                "produced": {"total": int, "by_type": dict[str, int]},
+                "consumed": {"total": int, "by_type": dict[str, int]}
+            }
+
+        Examples:
+            >>> aggregator = HistoryAggregator()
+            >>> summary = aggregator.aggregate(envelopes, "agent1")
+            >>> summary["produced"]["total"]
+            42
+        """
+        produced = self._aggregate_produced(envelopes, agent_id)
+        consumed = self._aggregate_consumed(envelopes, agent_id)
+
+        return {
+            "produced": {
+                "total": sum(produced.values()),
+                "by_type": dict(produced),
+            },
+            "consumed": {
+                "total": sum(consumed.values()),
+                "by_type": dict(consumed),
+            },
+        }
+
+    def _aggregate_produced(
+        self, envelopes: list[ArtifactEnvelope], agent_id: str
+    ) -> defaultdict[str, int]:
+        """
+        Count artifacts produced by agent, grouped by type.
+
+        Args:
+            envelopes: Artifact envelopes to analyze
+            agent_id: Producer to match
+
+        Returns:
+            Dict mapping artifact types to counts
+        """
+        from flock.core.store import ArtifactEnvelope
+
+        produced_by_type: defaultdict[str, int] = defaultdict(int)
+
+        for envelope in envelopes:
+            if not isinstance(envelope, ArtifactEnvelope):
+                raise TypeError("Expected ArtifactEnvelope instance")
+
+            artifact = envelope.artifact
+            if artifact.produced_by == agent_id:
+                produced_by_type[artifact.type] += 1
+
+        return produced_by_type
+
+    def _aggregate_consumed(
+        self, envelopes: list[ArtifactEnvelope], agent_id: str
+    ) -> defaultdict[str, int]:
+        """
+        Count artifacts consumed by agent, grouped by type.
+
+        Args:
+            envelopes: Artifact envelopes with consumption records
+            agent_id: Consumer to match
+
+        Returns:
+            Dict mapping artifact types to consumption counts
+        """
+        from flock.core.store import ArtifactEnvelope
+
+        consumed_by_type: defaultdict[str, int] = defaultdict(int)
+
+        for envelope in envelopes:
+            if not isinstance(envelope, ArtifactEnvelope):
+                raise TypeError("Expected ArtifactEnvelope instance")
+
+            artifact = envelope.artifact
+            for consumption in envelope.consumptions:
+                if consumption.consumer == agent_id:
+                    consumed_by_type[artifact.type] += 1
+
+        return consumed_by_type

flock/storage/sqlite/__init__.py

@@ -0,0 +1,10 @@
+"""SQLite storage backend components."""
+
+from flock.storage.sqlite.query_builder import SQLiteQueryBuilder
+from flock.storage.sqlite.schema_manager import SQLiteSchemaManager
+
+
+__all__ = [
+    "SQLiteQueryBuilder",
+    "SQLiteSchemaManager",
+]