truthound-dashboard 1.3.1__py3-none-any.whl → 1.4.1__py3-none-any.whl

truthound_dashboard/core/triggers/evaluators.py

@@ -0,0 +1,609 @@
+"""Trigger evaluator implementations.
+
+Provides concrete implementations for all trigger types:
+- CronTrigger: Cron expression based scheduling
+- IntervalTrigger: Fixed time interval scheduling
+- DataChangeTrigger: Profile-based change detection
+- CompositeTrigger: Combine multiple triggers
+- EventTrigger: Respond to system events
+- ManualTrigger: API-only execution
+"""
+
+from __future__ import annotations
+
+import logging
+from datetime import datetime, timedelta
+from typing import Any
+
+from apscheduler.triggers.cron import CronTrigger as APCronTrigger
+from apscheduler.triggers.interval import IntervalTrigger as APIntervalTrigger
+
+from .base import (
+    BaseTrigger,
+    TriggerContext,
+    TriggerEvaluation,
+    TriggerRegistry,
+)
+
+logger = logging.getLogger(__name__)
+
+
+@TriggerRegistry.register("cron")
+class CronTrigger(BaseTrigger):
+    """Cron expression based trigger.
+
+    Uses standard cron format: minute hour day month weekday
+
+    Config:
+        expression: Cron expression string
+        timezone: Optional timezone (default: UTC)
+    """
+
+    def _validate_config(self) -> None:
+        """Validate cron configuration."""
+        expression = self.config.get("expression")
+        if not expression:
+            raise ValueError("Cron trigger requires 'expression' field")
+
+        # Validate by attempting to parse
+        try:
+            APCronTrigger.from_crontab(expression)
+        except Exception as e:
+            raise ValueError(f"Invalid cron expression: {e}")
+
+    async def evaluate(self, context: TriggerContext) -> TriggerEvaluation:
+        """Evaluate if cron trigger should fire.
+
+        For cron triggers, we check if we're past the scheduled time
+        since the last run.
+        """
+        expression = self.config.get("expression")
+        trigger = APCronTrigger.from_crontab(expression)
+
+        # Get next fire time from last run (or from epoch if never run)
+        base_time = context.last_run_at or datetime(2000, 1, 1)
+        next_fire = trigger.get_next_fire_time(None, base_time)
+
+        if next_fire is None:
+            return TriggerEvaluation(
+                should_trigger=False,
+                reason="No upcoming scheduled time",
+            )
+
+        should_trigger = context.current_time >= next_fire
+
+        return TriggerEvaluation(
+            should_trigger=should_trigger,
+            reason=(
+                f"Scheduled time reached ({next_fire.isoformat()})"
+                if should_trigger
+                else f"Waiting for scheduled time ({next_fire.isoformat()})"
+            ),
+            next_evaluation_at=next_fire if not should_trigger else None,
+            details={
+                "expression": expression,
+                "next_fire_time": next_fire.isoformat(),
+            },
+        )
+
+    def get_next_evaluation_time(
+        self, context: TriggerContext
+    ) -> datetime | None:
+        """Get next cron fire time."""
+        expression = self.config.get("expression")
+        try:
+            trigger = APCronTrigger.from_crontab(expression)
+            return trigger.get_next_fire_time(None, context.current_time)
+        except Exception:
+            return None
+
+    def get_description(self) -> str:
+        """Get human-readable description."""
+        return f"Cron: {self.config.get('expression', 'not configured')}"
+
+
+@TriggerRegistry.register("interval")
+class IntervalTrigger(BaseTrigger):
+    """Fixed time interval trigger.
+
+    Config:
+        seconds: Interval in seconds
+        minutes: Interval in minutes
+        hours: Interval in hours
+        days: Interval in days
+    """
+
+    def _validate_config(self) -> None:
+        """Validate interval configuration."""
+        has_interval = any(
+            self.config.get(key)
+            for key in ["seconds", "minutes", "hours", "days"]
+        )
+        if not has_interval:
+            raise ValueError(
+                "Interval trigger requires at least one of: seconds, minutes, hours, days"
+            )
+
+    def _get_total_seconds(self) -> int:
+        """Calculate total interval in seconds."""
+        total = 0
+        total += self.config.get("seconds", 0)
+        total += self.config.get("minutes", 0) * 60
+        total += self.config.get("hours", 0) * 3600
+        total += self.config.get("days", 0) * 86400
+        return total or 3600  # Default to 1 hour
+
+    async def evaluate(self, context: TriggerContext) -> TriggerEvaluation:
+        """Evaluate if interval trigger should fire."""
+        interval_seconds = self._get_total_seconds()
+
+        # If never run, trigger immediately
+        if context.last_run_at is None:
+            return TriggerEvaluation(
+                should_trigger=True,
+                reason="First run (never executed before)",
+                details={"interval_seconds": interval_seconds},
+            )
+
+        # Calculate next run time
+        next_run = context.last_run_at + timedelta(seconds=interval_seconds)
+        should_trigger = context.current_time >= next_run
+
+        return TriggerEvaluation(
+            should_trigger=should_trigger,
+            reason=(
+                f"Interval elapsed ({interval_seconds}s since last run)"
+                if should_trigger
+                else f"Waiting for interval ({(next_run - context.current_time).seconds}s remaining)"
+            ),
+            next_evaluation_at=next_run if not should_trigger else None,
+            details={
+                "interval_seconds": interval_seconds,
+                "next_run": next_run.isoformat(),
+            },
+        )
+
+    def get_next_evaluation_time(
+        self, context: TriggerContext
+    ) -> datetime | None:
+        """Get next interval fire time."""
+        if context.last_run_at is None:
+            return context.current_time
+        return context.last_run_at + timedelta(seconds=self._get_total_seconds())
+
+    def get_description(self) -> str:
+        """Get human-readable description."""
+        parts = []
+        if self.config.get("days"):
+            parts.append(f"{self.config['days']}d")
+        if self.config.get("hours"):
+            parts.append(f"{self.config['hours']}h")
+        if self.config.get("minutes"):
+            parts.append(f"{self.config['minutes']}m")
+        if self.config.get("seconds"):
+            parts.append(f"{self.config['seconds']}s")
+        return f"Every {' '.join(parts)}" if parts else "Interval: not configured"
+
+
+@TriggerRegistry.register("data_change")
+class DataChangeTrigger(BaseTrigger):
+    """Data change detection trigger.
+
+    Triggers when profile metrics change by more than a threshold.
+
+    Config:
+        change_threshold: Minimum change percentage (0.0-1.0)
+        metrics: List of metrics to monitor
+        check_interval_minutes: How often to check
+    """
+
+    DEFAULT_METRICS = ["row_count", "null_percentage", "distinct_count"]
+
+    def _validate_config(self) -> None:
+        """Validate data change configuration."""
+        threshold = self.config.get("change_threshold", 0.05)
+        if not 0.0 <= threshold <= 1.0:
+            raise ValueError("change_threshold must be between 0.0 and 1.0")
+
+    async def evaluate(self, context: TriggerContext) -> TriggerEvaluation:
+        """Evaluate if data has changed enough to trigger.
+
+        Compares current profile against baseline and checks if
+        any monitored metrics have changed beyond the threshold.
+        """
+        threshold = self.config.get("change_threshold", 0.05)
+        metrics = self.config.get("metrics", self.DEFAULT_METRICS)
+
+        # Need both profiles to compare
+        if context.profile_data is None:
+            return TriggerEvaluation(
+                should_trigger=False,
+                reason="No current profile data available",
+                details={"error": "missing_current_profile"},
+            )
+
+        if context.baseline_profile is None:
+            # First profile - trigger to establish baseline
+            return TriggerEvaluation(
+                should_trigger=True,
+                reason="First profile (no baseline to compare)",
+                details={"reason": "no_baseline"},
+            )
+
+        # Calculate changes for each metric
+        changes = {}
+        max_change = 0.0
+        triggered_metrics = []
+
+        for metric in metrics:
+            current = self._get_metric_value(context.profile_data, metric)
+            baseline = self._get_metric_value(context.baseline_profile, metric)
+
+            if current is None or baseline is None:
+                continue
+
+            # Calculate percentage change
+            if baseline == 0:
+                change = 1.0 if current != 0 else 0.0
+            else:
+                change = abs(current - baseline) / abs(baseline)
+
+            changes[metric] = {
+                "current": current,
+                "baseline": baseline,
+                "change": change,
+            }
+
+            if change >= threshold:
+                triggered_metrics.append(metric)
+                max_change = max(max_change, change)
+
+        should_trigger = len(triggered_metrics) > 0
+
+        return TriggerEvaluation(
+            should_trigger=should_trigger,
+            reason=(
+                f"Data change detected: {', '.join(triggered_metrics)} changed by >= {threshold*100:.0f}%"
+                if should_trigger
+                else f"No significant changes (max: {max_change*100:.1f}%, threshold: {threshold*100:.0f}%)"
+            ),
+            details={
+                "threshold": threshold,
+                "max_change": max_change,
+                "changes": changes,
+                "triggered_metrics": triggered_metrics,
+            },
+            confidence=max_change if should_trigger else 1.0 - max_change,
+        )
+
+    def _get_metric_value(
+        self, profile: dict[str, Any], metric: str
+    ) -> float | None:
+        """Extract metric value from profile data."""
+        # Handle top-level metrics
+        if metric in profile:
+            return float(profile[metric])
+
+        # Handle nested column metrics
+        if "columns" in profile:
+            for col in profile["columns"]:
+                if metric in col:
+                    return float(col[metric])
+
+        # Handle summary metrics
+        if "summary" in profile and metric in profile["summary"]:
+            return float(profile["summary"][metric])
+
+        return None
+
+    def get_description(self) -> str:
+        """Get human-readable description."""
+        threshold = self.config.get("change_threshold", 0.05)
+        return f"Data change >= {threshold * 100:.0f}%"
+
+
+@TriggerRegistry.register("composite")
+class CompositeTrigger(BaseTrigger):
+    """Composite trigger that combines multiple triggers.
+
+    Config:
+        operator: "and" or "or"
+        triggers: List of trigger configurations
+    """
+
+    def _validate_config(self) -> None:
+        """Validate composite configuration."""
+        operator = self.config.get("operator", "and").lower()
+        if operator not in ("and", "or"):
+            raise ValueError("Composite operator must be 'and' or 'or'")
+
+        triggers = self.config.get("triggers", [])
+        if len(triggers) < 2:
+            raise ValueError("Composite trigger requires at least 2 sub-triggers")
+
+        # Validate sub-triggers don't contain composites (prevent nesting)
+        for trigger in triggers:
+            if trigger.get("type") == "composite":
+                raise ValueError("Nested composite triggers are not supported")
+
+    async def evaluate(self, context: TriggerContext) -> TriggerEvaluation:
+        """Evaluate composite trigger based on operator."""
+        operator = self.config.get("operator", "and").lower()
+        trigger_configs = self.config.get("triggers", [])
+
+        sub_results = []
+
+        for trigger_config in trigger_configs:
+            trigger_type = trigger_config.get("type")
+            trigger = TriggerRegistry.create(trigger_type, trigger_config)
+
+            if trigger is None:
+                sub_results.append({
+                    "type": trigger_type,
+                    "should_trigger": False,
+                    "reason": f"Unknown trigger type: {trigger_type}",
+                })
+                continue
+
+            result = await trigger.evaluate(context)
+            sub_results.append({
+                "type": trigger_type,
+                "should_trigger": result.should_trigger,
+                "reason": result.reason,
+            })
+
+        # Apply operator logic
+        if operator == "and":
+            should_trigger = all(r["should_trigger"] for r in sub_results)
+            reason = "All conditions met" if should_trigger else "Not all conditions met"
+        else:  # or
+            should_trigger = any(r["should_trigger"] for r in sub_results)
+            reason = "At least one condition met" if should_trigger else "No conditions met"
+
+        triggered_count = sum(1 for r in sub_results if r["should_trigger"])
+
+        return TriggerEvaluation(
+            should_trigger=should_trigger,
+            reason=f"{reason} ({triggered_count}/{len(sub_results)} triggers)",
+            details={
+                "operator": operator,
+                "sub_results": sub_results,
+                "triggered_count": triggered_count,
+                "total_count": len(sub_results),
+            },
+        )
+
+    def get_description(self) -> str:
+        """Get human-readable description."""
+        operator = self.config.get("operator", "and").upper()
+        triggers = self.config.get("triggers", [])
+        return f"Composite: {len(triggers)} triggers ({operator})"
+
+
+@TriggerRegistry.register("event")
+class EventTrigger(BaseTrigger):
+    """Event-based trigger.
+
+    Triggers in response to specific system events.
+
+    Config:
+        event_types: List of event types to respond to
+        source_filter: Optional list of source IDs to filter
+    """
+
+    VALID_EVENTS = {
+        "validation_completed",
+        "validation_failed",
+        "schema_changed",
+        "drift_detected",
+        "profile_updated",
+        "source_created",
+        "source_updated",
+    }
+
+    def _validate_config(self) -> None:
+        """Validate event configuration."""
+        event_types = self.config.get("event_types", [])
+        if not event_types:
+            raise ValueError("Event trigger requires 'event_types' list")
+
+        invalid_events = set(event_types) - self.VALID_EVENTS
+        if invalid_events:
+            raise ValueError(f"Invalid event types: {invalid_events}")
+
+    async def evaluate(self, context: TriggerContext) -> TriggerEvaluation:
+        """Evaluate if event matches configured types."""
+        event_types = set(self.config.get("event_types", []))
+        source_filter = self.config.get("source_filter")
+
+        # Check if there's event data in context
+        if context.event_data is None:
+            return TriggerEvaluation(
+                should_trigger=False,
+                reason="No event data in context",
+            )
+
+        event_type = context.event_data.get("type")
+        event_source = context.event_data.get("source_id")
+
+        # Check event type match
+        if event_type not in event_types:
+            return TriggerEvaluation(
+                should_trigger=False,
+                reason=f"Event type '{event_type}' not in configured types",
+                details={"event_type": event_type, "configured_types": list(event_types)},
+            )
+
+        # Check source filter if configured
+        if source_filter and event_source not in source_filter:
+            return TriggerEvaluation(
+                should_trigger=False,
+                reason=f"Event source '{event_source}' not in filter",
+                details={"event_source": event_source, "filter": source_filter},
+            )
+
+        return TriggerEvaluation(
+            should_trigger=True,
+            reason=f"Event '{event_type}' matched",
+            details={
+                "event_type": event_type,
+                "event_source": event_source,
+                "event_data": context.event_data,
+            },
+        )
+
+    def get_description(self) -> str:
+        """Get human-readable description."""
+        events = self.config.get("event_types", [])
+        return f"Events: {', '.join(events[:2])}{'...' if len(events) > 2 else ''}"
+
+
+@TriggerRegistry.register("manual")
+class ManualTrigger(BaseTrigger):
+    """Manual-only trigger.
+
+    Only triggers when explicitly invoked via API.
+    """
+
+    def _validate_config(self) -> None:
+        """No validation needed for manual trigger."""
+        pass
+
+    async def evaluate(self, context: TriggerContext) -> TriggerEvaluation:
+        """Manual triggers never auto-fire."""
+        # Check if manual trigger was requested via context
+        force_trigger = context.custom_data.get("force_trigger", False)
+
+        return TriggerEvaluation(
+            should_trigger=force_trigger,
+            reason=(
+                "Manually triggered via API"
+                if force_trigger
+                else "Manual trigger only (use API to execute)"
+            ),
+            details={"manual_only": True},
+        )
+
+    def get_description(self) -> str:
+        """Get human-readable description."""
+        return "Manual trigger only"
+
+
+@TriggerRegistry.register("webhook")
+class WebhookTrigger(BaseTrigger):
+    """Webhook-based trigger.
+
+    Triggers when receiving webhook requests from external systems.
+
+    Config:
+        webhook_secret: Optional secret for HMAC validation
+        allowed_sources: List of allowed source identifiers
+        payload_filters: JSON path filters for payload matching
+        require_signature: Whether to require signature validation
+    """
+
+    def _validate_config(self) -> None:
+        """Validate webhook configuration."""
+        # Webhook config is optional, but if require_signature is True,
+        # webhook_secret must be provided
+        if self.config.get("require_signature") and not self.config.get("webhook_secret"):
+            raise ValueError(
+                "webhook_secret is required when require_signature is True"
+            )
+
+    async def evaluate(self, context: TriggerContext) -> TriggerEvaluation:
+        """Evaluate if webhook trigger should fire.
+
+        Checks:
+        1. Webhook data exists in context
+        2. Source is in allowed_sources (if configured)
+        3. Payload matches filters (if configured)
+        4. Signature is valid (if required)
+        """
+        # Check for webhook data in context
+        webhook_data = context.custom_data.get("webhook_data")
+        if webhook_data is None:
+            return TriggerEvaluation(
+                should_trigger=False,
+                reason="No webhook data in context",
+                details={"waiting_for_webhook": True},
+            )
+
+        source = webhook_data.get("source", "")
+        payload = webhook_data.get("payload", {})
+        signature_valid = webhook_data.get("signature_valid", True)
+
+        # Check signature if required
+        require_signature = self.config.get("require_signature", False)
+        if require_signature and not signature_valid:
+            return TriggerEvaluation(
+                should_trigger=False,
+                reason="Invalid webhook signature",
+                details={"error": "invalid_signature", "source": source},
+            )
+
+        # Check allowed sources
+        allowed_sources = self.config.get("allowed_sources")
+        if allowed_sources and source not in allowed_sources:
+            return TriggerEvaluation(
+                should_trigger=False,
+                reason=f"Source '{source}' not in allowed sources",
+                details={
+                    "source": source,
+                    "allowed_sources": allowed_sources,
+                },
+            )
+
+        # Check payload filters
+        payload_filters = self.config.get("payload_filters", {})
+        if payload_filters:
+            filters_match = self._check_payload_filters(payload, payload_filters)
+            if not filters_match:
+                return TriggerEvaluation(
+                    should_trigger=False,
+                    reason="Payload does not match filters",
+                    details={
+                        "source": source,
+                        "filters": payload_filters,
+                    },
+                )
+
+        return TriggerEvaluation(
+            should_trigger=True,
+            reason=f"Webhook received from '{source}'",
+            details={
+                "source": source,
+                "event_type": webhook_data.get("event_type", "unknown"),
+                "payload_keys": list(payload.keys()) if payload else [],
+            },
+        )
+
+    def _check_payload_filters(
+        self, payload: dict[str, Any], filters: dict[str, Any]
+    ) -> bool:
+        """Check if payload matches the configured filters.
+
+        Simple key-value matching. For nested paths, use dot notation.
+        """
+        for key, expected_value in filters.items():
+            # Support dot notation for nested keys
+            parts = key.split(".")
+            value = payload
+            for part in parts:
+                if isinstance(value, dict):
+                    value = value.get(part)
+                else:
+                    value = None
+                    break
+
+            if value != expected_value:
+                return False
+
+        return True
+
+    def get_description(self) -> str:
+        """Get human-readable description."""
+        allowed = self.config.get("allowed_sources", [])
+        if allowed:
+            return f"Webhook: {', '.join(allowed[:2])}{'...' if len(allowed) > 2 else ''}"
+        return "Webhook trigger"