SPARQL Engine

File Release Notes and Changelog

Release Name: 0.8

Notes:
Version 0.8 Release Notes

This release is huge and indicates a small change in my SPARQL philosophy of the past.  When I first
wrote the interpreter I saw it from a implementor point of view, but after spending a couple months
using it, I began to notice certain features that it was lacking, in comparison to other query languages.

To that end, the language is now a superset of SPARQL and while I will always release a version that is
pure SPARQL and will never hopefully contradict the specification, I think it's a step forward for the
language.

See the CHANGELOG for a full breakdown of the changes.  The new features will be documented in the coming
days so they can be useful, but for now the engine itself should be greatly improved.  Besides bug fixes,
there is now reordering logic that dramatically increases the speed of poorly ordered queries.  See
SPARQLQueryLogic which is now the facade for the engine after it has already been adapted to work
with a RDF source via the SelectQuery/ConstructQuery/DescribeQuery/AskQuery execute methods.

Changes:
-----------------------
Version 0.8, 2006-07-16
-----------------------

- Updated the grammar to the latest W3C version, including unicode and escapes
- Fixed multiple UNION bug, certain types of GRAPH query bugs, and several other big ones
- Fixed several logic errors stemming from my misunderstanding of unbound versus null values
- Introduced an intersection ordering logic to reorder the intersected sets to prevent cross-products
- Extended the language (parser and interpreter) to support functional predicates and WITH extensions
- Extended the language to allow expressions in SELECT bindings and CONSTRUCT templates
- Implemented a small set of core value functions, this will be expanded greatly in the next release
- Added a logic facade (SPARQLQueryLogic) so you can easily change the default logic for every query
- Value factory and logic binding is done at query parse time, which should make it easier to deal with Query objects
- Added a debugging system through SPARQLQueryLogic, can output intermediate set logic steps
- Added a GUI console that can be used to execute queries and then debug them
- Added more documentation, but it seems to be neverending :)

-----------------------
Version 0.71, 2006-02-22
-----------------------

- Fixed bugs 1377140 and 1436935 and added test cases to assure their success
- Added a bunch of final modifiers for completeness

-----------------------
Version 0.7, 2005-11-28
-----------------------

- Fixed CONSTRUCT queries so they seem to work now
- Fixed SELECT * queries
- Redesigned the visitor pattern in the AST parser so it's a lot cleaner

-----------------------
Version 0.6, 2005-10-02
-----------------------

- Moved set logic from RdfBindingSetUtils to it's own logic in the factory
- Created much more efficient streamed/indexed logic steps to use by default
- Changed DESCRIBE so that it should work correctly

-----------------------
Version 0.5, 2005-08-31
-----------------------

- All javadoc finished
- The engine now passes all SELECT tests in the specification
- Extensive test parsing and running module added
- All Values are now created with ValueFactory factories
- Fixed all previous bugs from the bug list
- Moved anything Sesame specific to it's own project

-----------------------
Version 0.2, 2005-08-18
-----------------------

- Finished almost all the javadoc
- Fixed several major bugs with optional constraints, ask queries, and syntax shortcuts
- Reworked and simplified the logic factory design
- Added pre-execution variable access
- Removed Comparable from the main Value interfaces and put it in it's own logic
- Removed Javadoc from the build, users can javadoc it themselves

-----------------------
Version 0.1, 2005-08-12
-----------------------

- First public release of the source code
- Included Sesame/adapter distribution for testing purposes