We need to copy contents to azure blob storage (this is already done by the archiver).
The idea is once that copy is done, send those contents for computing multiple indexes on them (mimetype, language detection, sloc count, etc...).