Although javadoc of Annotation.getAnnotationType() defines file name of .xml file describing the annotation's data in
reality the result of that method is used to match against content of <name> attribute. This way _all_ the registered
annotations' xml files first get loaded and then the one with the <name> attribute corresponding to the given A.gAT() is
used (see AnnotationTypes and AnnotationTypesFolder). With the present >80 annotaions this affects performance
negatively (it's currently done as a warmup task but still it consumes extra memory etc.)
As the javadoc already says it right it's just necessary to obey the contract and check if all the annotations have
their file name equal to their <name> attribute (they mostly should).