[Biojava-dev] bjv2

[Matthew Pocock]

Hi,

I've had a fairly serious hack arround with bjv2 over the last 2 weeks. 
I will put in a tag before BOSC. The code triggers a bug in javac or 
javadoc with beta2 build from the sun website, but this is fixed in the 
beta3 pre-release.

I've concentrated on core genomics so far, but we can start adding APIs 
for other things as soon as the core is out of alpha. So, wish-lists or 
use-cases or stories would be great.

Matthew


New since the last release:

Documentation of often-used features
  general IO
  sequence IO
  symbols and strings
  controling data integration & memory

Complete re-working of the APIs for biological symbols
  type-safe
  supports windowed views
  supports convertions to/from strings using trivial syntax
  supports systematic mappings of symbols e.g. dna->rna translations

Data projection and integration
  we can now map from underlying data to a 'view' bean based upon
    javabean properties
    annotations and map values
    rdf properties
  performance and scaleability fixes
  support for agregate functions
    min, max, average, all identical values

RDF
  there is a closer integration between RDFS data-models and bjv2 
introspection
  RDF is fully supported in the query language

New file formats
  now cover Embl, Fasta, Gff, Blast (not tested on many blast flavors)

Taverna plugin bridge
  uses the @Utility hooks to expose all facades as taverna processors
  same APIs could be used to expose all facades as web services, or whatever

Todo before ISMB
  swissprot IO
  alignments + alignment IO
  similarity search data model (populated from blast?)
  protein alphabet
  assemblies

Todo soon after ISMB
  wire RDF into the parser event framework propperly
  bridge RDFS and Annotation
  RDFS -> data model compiler
  RDF -> file format parser, and write the RDF for /lots/ of SRS-able 
formats
  use the RDF support to give us stable serialization, for RMI and for 
web services

Todo slightly later ISMB
  support data modification, transactions, set() methods.
  add APIs for common areas
    metabolomics
    trees
    snp analysis
    linkage
    ...