Peregrine Releases

From BioAssist
Revision as of 17:13, 28 February 2010 by Dmitry Katsubo (Talk | contribs)

Jump to: navigation, search

Peregrine Releases

v0.3

What's new

  • Peregrine WS/RMI installator is implemented (RT#213).
  • Added support of modifiable ontologies (RT#214).
  • Added multi-language support (RT#114).
Changes in disambiguation API
Previous version Current version
  • Indexing procedure created disambiguation information enriched objects not depending or whether disambiguation step will be taken afterwards.
  • Disambiguation results where represented as accompanying objects, created next to indexing result objects.
  • Depending on further processing steps indexing step creates objects of two types: light-weight and disambiguation information enriched ones.
  • Disambiguation results are aggregated into indexing results.



v0.2

What's new

  • Peregrine WS implementation (RT#65) and RMI interface (RT#66) are available.
  • caseSensitive property has been renamed to caseInsensitive (impacts TermId interface and SQL Schema).
  • Added Ontology#getTerm(Term.Id) interface method.
  • Implemented keyword positions (RT#63).
  • Reorganised test resources: now test ontology can be shared among projects.

Complete list of changes

  • Renamed TermId to TermIdImpl.
  • Unified dependency library versioning across all POM files.
  • Improved WS implementation: fixed IndexingResult JAXB annotations, introduced independent WS interface (RT#65).
  • Implemented more fine file ontology hierarchy.
  • All classes that took part in RMI interaction should be Serializable.
  • Common tests for Peregrine interface.
  • Adding SVN build number to maven artifacts when making a release.
  • Made Concept#disambiguationType property optional with default value.



v0.1

What's new

  • Peregrine: re-uses common XSD types (RT#119)
  • Class diagram sequence diagram is available in wiki (RT#120)
  • DB Ontology was implemented (RT#121)
  • File Ontology was implemented (RT#122)
  • Compound ontology was implemented (RT#123)

Complete list of changes

  • Improved testing of flyweight processing interface in case when processing is aborted.
  • Marked database Ids as optional list.
  • Added line indexes to error message reports for grammar parser.
  • Added transaction support to DB ontology.
  • Simplified FlyweightProcessingOntology interface.
  • Implemented FileFlyweightProcessingOntologyImpl that can read ontologies from compressed sources.
  • Implemented modifiable SQL schema for ontology. Made all entities serializable. More correct and efficient buffered iterator implementation.
  • Added UTF-8 support for random-read files.
  • Corrected the lazy initialisation of concept ID map in file ontology. Improved unit tests.

Known issues and limitations

  • Tokenizer should have a default constructor to be recovered from serialization.