Python/SQLAlchemy : Event Listeners after_flush fires for decimal values, 即使值没有改变

Question

这是我的方法:

def member_field_change_after_flush(session, flush_context):
    """
    Keep track of MemberField objects changing per-session.
    """
    member_fields = field_changes[session]
    session_member_fields = filter(
        lambda model: isinstance(model, MemberField),
        session.new | session.dirty | session.deleted
    )

    for member_field in session_member_fields:
        for mapper_property in object_mapper(member_field).iterate_properties:
            if isinstance(mapper_property, ColumnProperty):
                attribute_state = inspect(member_field).attrs.get(mapper_property.key)
                history = attribute_state.history
                if history.has_changes():
                    member_fields[member_field.field_id] = {
                        'member_id': member_field.member_id,
                        'old_value': history.deleted[0] if history.deleted else None,
                        'new_value': attribute_state.value,
                        'field_type': member_field.field.ref_field_type.column_key,
                        'account_id': member_field.account_id,
                    }

如果我更改数据库中的值 100.2至100.2 , SQLAlchemy认为值实际上已更改并登录 history.deleted[0] dict 中的属性如上所述。

我对Python相当陌生，所以我不确定如何解决这个问题。

Answer 1

每当值发生更改时，您的事件就会被触发。但是，要检查值是否实际更改，您需要调用 session.is_modified(...)。

示例:

for obj in session.dirty:
    # Here, the field might have been set to the same value
    if session.is_modified(obj):
        # Here, the fields value has actually changed

我猜这样实现的原因是优化。 docs say something similar :

is_modified(instance, include_collections=True, passive=True)

It is in effect a more expensive and accurate version of checking for the given instance in the Session.dirty collection; a full test for each attribute’s net “dirty” status is performed.