Automatically download Tesseract training files during an import

Currently, OCR-7 allows to download training files from the platform administration, which is already quite good.

The idea now would be to propose by default every available data file when importing a document, and download them if they are not present in the store.