PropertyAccessorTest history

Click on the commit message for inspecting the diff and how the code has been rewritten.

Command line:

svn log -r31996:1 http://svn.osgeo.org/geotools/trunk/modules/library/metadata/src/test/java/org/geotools/metadata/PropertyAccessorTest.java
Rev. Date Author Message
308522008-07-03desruisseauxPropertyMap now converts automatically from String to a variety of types (InternationalString, File, Number...) and add singleton to collection. UML identifier name are now accepted as well as property name inferred from method name.
306402008-06-12acusterCopyright headers: lib/metadata, this time with feeling (and the el in Toolkit)
306372008-06-12acusterCopyright headers: lib/metadata, now normalized to LGPL v2.1 only
305192008-06-05acusterMetadata header cleanup and copyright review.
302582008-05-08acusterReshuffle the top level repo: drop uDig, move up trunk, tags, and branches.
302572008-05-08acusterMove trunk/gt/ directory contents up to trunk/ and drop gt
297682008-04-02desruisseauxUpgrated the metadata and referencing tests to JUnit 4.
289222008-01-24acusterBump the (at)since version to 2.5 since WKTParser was cut from 2.4
285402007-12-29acusterHide buttons which are not yet used
283742007-12-14desruisseauxUse some java 5 features. Added javadoc. Removed AbstractCoverage.dispose() deprecated method (replaced by dispose(boolean)). Cleaning.
282372007-12-03desruisseauxFixed the search for implemented interfaces (looks deeper in the hierarchy).
280512007-11-26desruisseauxMore Java 5 parameterized types. Include a fix for GEOT-1599 (CitationImpl.getIdentifiers() should returns a set of Identifiers, not a set of Strings).
251932007-04-18desruisseauxFixed some typos in metadata revealed by the new ISOTest. First draft of hints review (still in process).
251752007-04-16desruisseauxRemoved 'equals', 'hashCode', 'toString' and 'freeze' methods from every metadata implementation. They are now implemented by generic methods inherited from AbstractMetadata, which perform their work using Java reflection. This way we avoid lot of redundancy, some errors introduced by maintenance ('equals' method not updated in the way they should), more systematic implementations (espcially for the 'toString' method, which was not implemented at all for a majority of metadata), more support for future introduction of other metadata standards than ISO 19115, etc. The cost is more runtime overhead, but it still possible to provide custom implementations for 'equals' and 'hashCode' only in the class that appear to be bottleneck, if we want.
251572007-04-12desruisseauxMetadata review. Also provided a first draft of support classes for new 'equals(...)' copy constructor implementations backed by Java reflection.