Hello all,
a very basic version of the Corpus Server is now available,
it should be good enough be used for the development
of the Corpus Refiner and the Wikinews Importer. It
will of course improve over time, especially after we are done
with the Lucene and Derby integration.
The Corpus Sever project consists of two parts, the server itself
and command line tool.
The server itself can host a corpus, and this corpus contains
an UIMA type systme and UIMA CASes.
The server is packaged as a war file and can be deployed to
a tomcat (e.g.). After that is done the command line tool
can create with this command a corpus:
CreateCorpus http://localhost:8080/corpus-server/rest/corpora wikinews
../wikinews-importer/samples/TypeSystem.xml
and add a xmi file via this command:
CASImporter http://localhost:8080/corpus-server/rest/corpora/wikinews
../wikinews-importersamples/FTC_begins_antitrust_inquiry_of_Google.xmi
The rest interface of the server is defined in:
CorpusResource and should be easy to use, as the sample code in the
CreateCorpus and CASImporter classes show.
All the mentioned code here can be found in our sandbox.
Jörn