django-bulk-hooks 0.1.280__py3-none-any.whl → 0.2.1__py3-none-any.whl

django_bulk_hooks/operations/analyzer.py

@@ -0,0 +1,208 @@
+"""
+Model analyzer service - Combines validation and field tracking.
+
+This service handles all model analysis needs:
+- Input validation
+- Field change detection
+- Field comparison
+"""
+
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+class ModelAnalyzer:
+    """
+    Analyzes models and validates operations.
+
+    This service combines the responsibilities of validation and field tracking
+    since they're closely related and often used together.
+    """
+
+    def __init__(self, model_cls):
+        """
+        Initialize analyzer for a specific model.
+
+        Args:
+            model_cls: The Django model class
+        """
+        self.model_cls = model_cls
+
+    # ========== Validation Methods ==========
+
+    def validate_for_create(self, objs):
+        """
+        Validate objects for bulk_create operation.
+
+        Args:
+            objs: List of model instances
+
+        Raises:
+            TypeError: If objects are not instances of model_cls
+        """
+        self._check_types(objs, operation="bulk_create")
+        return True
+
+    def validate_for_update(self, objs):
+        """
+        Validate objects for bulk_update operation.
+
+        Args:
+            objs: List of model instances
+
+        Raises:
+            TypeError: If objects are not instances of model_cls
+            ValueError: If objects don't have primary keys
+        """
+        self._check_types(objs, operation="bulk_update")
+        self._check_has_pks(objs, operation="bulk_update")
+        return True
+
+    def validate_for_delete(self, objs):
+        """
+        Validate objects for delete operation.
+
+        Args:
+            objs: List of model instances
+
+        Raises:
+            TypeError: If objects are not instances of model_cls
+        """
+        self._check_types(objs, operation="delete")
+        return True
+
+    def _check_types(self, objs, operation="operation"):
+        """Check that all objects are instances of the model class"""
+        if not objs:
+            return
+
+        invalid_types = {
+            type(obj).__name__ for obj in objs if not isinstance(obj, self.model_cls)
+        }
+
+        if invalid_types:
+            raise TypeError(
+                f"{operation} expected instances of {self.model_cls.__name__}, "
+                f"but got {invalid_types}"
+            )
+
+    def _check_has_pks(self, objs, operation="operation"):
+        """Check that all objects have primary keys"""
+        missing_pks = [obj for obj in objs if obj.pk is None]
+
+        if missing_pks:
+            raise ValueError(
+                f"{operation} cannot operate on unsaved {self.model_cls.__name__} instances. "
+                f"{len(missing_pks)} object(s) have no primary key."
+            )
+
+    # ========== Data Fetching Methods ==========
+
+    def fetch_old_records_map(self, instances):
+        """
+        Fetch old records for instances in a single bulk query.
+
+        This is the SINGLE point of truth for fetching old records.
+        All other methods should delegate to this.
+
+        Args:
+            instances: List of model instances
+
+        Returns:
+            Dict[pk, instance] for O(1) lookups
+        """
+        pks = [obj.pk for obj in instances if obj.pk is not None]
+        if not pks:
+            return {}
+
+        return {obj.pk: obj for obj in self.model_cls._base_manager.filter(pk__in=pks)}
+
+    # ========== Field Introspection Methods ==========
+
+    def get_auto_now_fields(self):
+        """
+        Get fields that have auto_now or auto_now_add set.
+
+        Returns:
+            list: Field names with auto_now behavior
+        """
+        auto_now_fields = []
+        for field in self.model_cls._meta.fields:
+            if getattr(field, "auto_now", False) or getattr(
+                field, "auto_now_add", False
+            ):
+                auto_now_fields.append(field.name)
+        return auto_now_fields
+
+    def get_fk_fields(self):
+        """
+        Get all foreign key fields for the model.
+
+        Returns:
+            list: FK field names
+        """
+        return [
+            field.name
+            for field in self.model_cls._meta.concrete_fields
+            if field.is_relation and not field.many_to_many
+        ]
+
+    def detect_changed_fields(self, objs):
+        """
+        Detect which fields have changed across a set of objects.
+
+        This method fetches old records from the database in a SINGLE bulk query
+        and compares them with the new objects to determine changed fields.
+
+        PERFORMANCE: Uses bulk query (O(1) queries) not N queries.
+
+        Args:
+            objs: List of model instances to check
+
+        Returns:
+            List of field names that changed across any object
+        """
+        if not objs:
+            return []
+
+        # Fetch old records using the single source of truth
+        old_records_map = self.fetch_old_records_map(objs)
+        if not old_records_map:
+            return []
+
+        # Track which fields changed across ALL objects
+        changed_fields_set = set()
+
+        # Compare each object with its database state
+        for obj in objs:
+            if obj.pk is None:
+                continue
+
+            old_obj = old_records_map.get(obj.pk)
+            if old_obj is None:
+                # Object doesn't exist in DB, skip
+                continue
+
+            # Check each field for changes
+            for field in self.model_cls._meta.fields:
+                # Skip primary key and auto fields
+                if field.primary_key or field.auto_created:
+                    continue
+
+                old_val = getattr(old_obj, field.name, None)
+                new_val = getattr(obj, field.name, None)
+
+                # Use field's get_prep_value for proper comparison
+                try:
+                    old_prep = field.get_prep_value(old_val)
+                    new_prep = field.get_prep_value(new_val)
+                    if old_prep != new_prep:
+                        changed_fields_set.add(field.name)
+                except (TypeError, ValueError):
+                    # Fallback to direct comparison
+                    if old_val != new_val:
+                        changed_fields_set.add(field.name)
+
+        # Return as sorted list for deterministic behavior
+        return sorted(changed_fields_set)

django_bulk_hooks/operations/bulk_executor.py

@@ -0,0 +1,151 @@
+"""
+Bulk executor service for database operations.
+
+This service coordinates bulk database operations with validation and MTI handling.
+"""
+
+import logging
+from django.db import transaction
+
+logger = logging.getLogger(__name__)
+
+
+class BulkExecutor:
+    """
+    Executes bulk database operations.
+
+    This service coordinates validation, MTI handling, and actual database
+    operations. It's the only service that directly calls Django ORM methods.
+
+    Dependencies are explicitly injected via constructor.
+    """
+
+    def __init__(self, queryset, analyzer, mti_handler):
+        """
+        Initialize bulk executor with explicit dependencies.
+
+        Args:
+            queryset: Django QuerySet instance
+            analyzer: ModelAnalyzer instance (replaces validator + field_tracker)
+            mti_handler: MTIHandler instance
+        """
+        self.queryset = queryset
+        self.analyzer = analyzer
+        self.mti_handler = mti_handler
+        self.model_cls = queryset.model
+
+    def bulk_create(
+        self,
+        objs,
+        batch_size=None,
+        ignore_conflicts=False,
+        update_conflicts=False,
+        update_fields=None,
+        unique_fields=None,
+        **kwargs,
+    ):
+        """
+        Execute bulk create operation.
+
+        NOTE: Coordinator is responsible for validation before calling this method.
+        This executor trusts that inputs have already been validated.
+
+        Args:
+            objs: List of model instances to create (pre-validated)
+            batch_size: Number of objects to create per batch
+            ignore_conflicts: Whether to ignore conflicts
+            update_conflicts: Whether to update on conflict
+            update_fields: Fields to update on conflict
+            unique_fields: Fields to use for conflict detection
+            **kwargs: Additional arguments
+
+        Returns:
+            List of created objects
+        """
+        if not objs:
+            return objs
+
+        # Execute bulk create - validation already done by coordinator
+        return self._execute_bulk_create(
+            objs,
+            batch_size,
+            ignore_conflicts,
+            update_conflicts,
+            update_fields,
+            unique_fields,
+            **kwargs,
+        )
+
+    def _execute_bulk_create(
+        self,
+        objs,
+        batch_size=None,
+        ignore_conflicts=False,
+        update_conflicts=False,
+        update_fields=None,
+        unique_fields=None,
+        **kwargs,
+    ):
+        """
+        Execute the actual Django bulk_create.
+
+        This is the only method that directly calls Django ORM.
+        We must call the base Django QuerySet to avoid recursion.
+        """
+        from django.db.models import QuerySet
+
+        # Create a base Django queryset (not our HookQuerySet)
+        base_qs = QuerySet(model=self.model_cls, using=self.queryset.db)
+
+        return base_qs.bulk_create(
+            objs,
+            batch_size=batch_size,
+            ignore_conflicts=ignore_conflicts,
+            update_conflicts=update_conflicts,
+            update_fields=update_fields,
+            unique_fields=unique_fields,
+        )
+
+    def bulk_update(self, objs, fields, batch_size=None):
+        """
+        Execute bulk update operation.
+
+        NOTE: Coordinator is responsible for validation before calling this method.
+        This executor trusts that inputs have already been validated.
+
+        Args:
+            objs: List of model instances to update (pre-validated)
+            fields: List of field names to update
+            batch_size: Number of objects to update per batch
+
+        Returns:
+            Number of objects updated
+        """
+        if not objs:
+            return 0
+
+        # Execute bulk update - use base Django QuerySet to avoid recursion
+        # Validation already done by coordinator
+        from django.db.models import QuerySet
+
+        base_qs = QuerySet(model=self.model_cls, using=self.queryset.db)
+        return base_qs.bulk_update(objs, fields, batch_size=batch_size)
+
+    def delete_queryset(self):
+        """
+        Execute delete on the queryset.
+
+        NOTE: Coordinator is responsible for validation before calling this method.
+        This executor trusts that inputs have already been validated.
+
+        Returns:
+            Tuple of (count, details dict)
+        """
+        if not self.queryset:
+            return 0, {}
+
+        # Execute delete via QuerySet
+        # Validation already done by coordinator
+        from django.db.models import QuerySet
+
+        return QuerySet.delete(self.queryset)