OCR-D processor: Set @index and/or add TextEquiv on export

[kba]

    How about setting `@index`? How about merely **adding** the imported text in the front? 

This could be useful to compare automatic and manual versions visually later-on (e.g. with LAREX).

Originally posted by @bertsky in #9 (comment)