Ingesting HOCR derivatives as a media attachment

[dmer]

This work is in support of the plans to provide search term highlighting in Mirador started by @alxp Islandora/islandora#897

And continued by @patdunlavey here:
Islandora/islandora_mirador#17 (comment)

Functions like the OCR where the contents of the extracted text file are copied into a text field on the media (original_file) for indexing purposes.

USE CASE: I have a Islandora 7 repository with a very large amount of textual content in TIF file format - each page (TIF) has an associated HOCR file. I want to migrate the pages WITH their HOCR into Islandora.

I'd like to be able to batch in the HOCR files (either as part of the node-creating csv or as an add_media job) and have them attached to the appropriate file field on the media object.

Ideally I could pull these HOCR files directly from the Islandora7 datastream with a URL like I do for the OBJ (TIF) files.

Hopefully this is a clear definition of the ask - I'm happy to answer questions or add more details if requested.

[mjordan]

Related - #572.

[dmer]

Hi Mark - just checking in on this. I'm expecting to need this within a month or so. I can definitely volunteer some testing.