Process all Wikipedias by default
- add
wikipedias.json
with all Wikipedia language codes as a package resource file - the default behavior is to load it for both Spark scripts
- make sure the file gets packaged when the artifact is built, see latest build.