CustomConversions now evaluates @ReadingConverter / @WritingConverter when adding Converter implementations. See DATACMNS-113 and the appropriate commit for details. Added unit test to verify StringToBigIntegerConverter does not get added as writing converter.
In case a query referencing an entity's id needs massaging before being executed (e.g. Strings that can be ObjectID) the massaging failed in case the query was built using _id already as we did not detect that one being an id reference as we compared it to the entity's id property name. We now also compare against it's field name.
Changed the setter parameters for ServerAddresses to use arrays instead of List. We now register the ServerAddressPropertyEditor to convert a ServerAddress[] and thus don't register a PropertyEditor for List which caused unwanted side effects before.
If an entity is handed into the template to be saved or inserted we now check that the auto-generated ObjectId can actually be applied to the id property after saving the object.
Throwing an exception if an update does not affect any documents doesn't make sense in all cases. Removed throwing an exception by default but made the relevant method (handleAnyWriteResultErrors(…)) protected so that subclasses might override this behavior.
JavaDoc, Formatting. Made dependencies in DefaultIndexOperations final. Reduced dependency to MongoOperations instead of depending on MongoTemplate directly. Added not-null assertion to constructor of DIO.
Cleaned up the id handling on query mapping and mapping in general. We now only try to convert id values into an ObjectId and store it as is using potentially registered custom converters. Register BigInteger<->String converters by default now.
The MANIFEST.MF in current MongoDB driver version is broken in terms of not stating package versions. Thus we unfortunately cannot refer to a particular version range but have to use the generic 0 as required version.
In case an extension of AbstractMongoEventListener does not define a parameter type we now default to Object as handled domain type as we'd cause a NullPointerException if not.
We now check whether a URL was passed in as map and/or reduce function and throw an exception in case the file either does not exist or cannot be read.
Query now exposes a withHint(…) method which will be applied to the DBCursor on query execution. Reduced CursorPreparer's visibility to the package and removed methods exposing it from MongoOperations.
QueryMapper now delegates to a MongoConverter instead of a plain ConversionService and invokes optional conversion on it. This optional conversion now removes type information from the created DBObject.
The decision whether a property value was handled as Collection or Map was based on inspecting the property's type which failed for classes using very open property declarations such as:
class MyClass {
Object something;
}
We now rather inspect the value type instead of the property.
<mongo:custom-conversions /> now has a base-package attribute that scans for Converter and GenericConverter beans. Added <tool:exports /> metadata for MappingMongoConverter.
CustomConversions now delegates to MongoSimpleTypes.HOLDER.isSimpleType(…) instead of maintaining an additional list of Mongo-primitive types. Added DBObject to the list of Mongo-primitive types.
MappingMongoConverter.convertToMongoType(…) not only converts elements of collections and maps but also converts the wrapper into the appropriate MongoDB type (BasicDBList, BasicDBObject).
If an Enum implements abstract methods, the Class object derived from ${ENUM}.getClass() does not return true for ….isEnum(). Thus we have to rather check Enum.class.isAssignableFrom(…) as this catches this scenario as well. Also see DATACMNS-99 for a related fix in simple type handling in the core infrastructure.
Oliver Gierke
Mark Pollack