Do we need to merge similar services: weblab-resource-iterator and folder-source-reader?
These two services are providing the same functionnality, according to their POM description.
- folder-source-reader : Use this component to crawl a folder with RDF weblab resources.
- weblab-resource-iterator : This service is a simple SourceReader that allow to browse a folder containing WebLab resources.
Main differences are:
- folder-source-reader implements SourceReader + Configurable in a single class (not allowed now with CXF) while weblab-resource-iterator implements QueueManager and Configurable in two files.
- folder-source-reader provides ways of filtering the content of the folder to crawl (e.g. only crawl xml files, ignore .svn folders...) while weblab-resource-iterator does not (it takes every single file).
- folder-source-reader enable to delete crawled files, while weblab-resource-iterator does not. (But I don't think a SourceReader should, while a QueueManager can)
- folder-source-reader can enrich the resource with a dc:source with the path to the file, while weblab-resource-iterator does not. (But I don't think it's useful)
On top of that both are wheel, compared with the file iterator and the filter provided by commons-io. This should be rationalised one day to prevent the need of maintainance of duplicate services.