Bibliotek-o to BIBFRAME is a lossy RDF to RDF conversion where a significant portion of the data will remain the same (things in the bf: namespace). It thus seems that SPARQL UPDATE to delete bibliotek-o triples and replace them with BIBFRAME equivalent would be a good approach?
The conversion is lossy because bibliotek-o is more expressive than BIBFRAME.
The Python make_bteko2bf_sparql.py program is designed to write a SPARQL UPDATE script bteko2bf.ru based on data from bibliotek-o to BIBFRAME mappings. Generate with:
> python make_bteko2bf_sparql.py -v
INFO:root:Done, written 90 mappings to bteko2bf.ru
Assuming the SPARQL UPDATE code is bteko2bf.ru, one can run it on a bibliotek-o file testdata/ex_listingCredits_1_bteko.ttl with:
> update --dump --data=testdata/ex_listingCredits_1_bteko.ttl --update=bteko2bf.ru
<http://example.org/obj1> <http://id.loc.gov/ontologies/bibframe/credits> "Credit" .
the output BIBFRAME is written to STDOUT (just one triple in this example).
See for notes on running SPARQL UPDATE as a command-line tool. There are also notes about running against a Fuseki triplestore server.
The program test_bteko2bf_sparql.py uses the Python unittest framework to run tests against all examples *_bteko.ttl -> *_bf.ttl in the testdata directory. Run with:
> python test_bteko2bf_sparql.py
.
----------------------------------------------------------------------
Ran 1 test in 3.018s
OK
The code is designed to run with Python3.5 or higher.
The tests rely on Apache Jena tools, on OSX these can be installed with homebrew:
> brew install jena
I'm not sure how sensitive any of this is to version on Jena. Developement has been done against 3.1.0:
> update --version
Jena: VERSION: 3.1.0
Jena: BUILD_DATE: 2016-05-10T11:59:39+0000
ARQ: VERSION: 3.1.0
ARQ: BUILD_DATE: 2016-05-10T11:59:39+0000
RIOT: VERSION: 3.1.0
RIOT: BUILD_DATE: 2016-05-10T11:59:39+0000
and the travis tests run aginst 3.4.0 (as of 2017-08-07).
WARNING - Mappings currently work in progress, the versions here are an incomplete dump*
The directory vendor/bibliottek-o_bibframe_usage has TSV dumps of the current mapping spreadsheets from the LD4P bibliottek-o work. This will be updated when real/final mappings are published.