I have worked through the tutorial at
https://onehippo-forge.github.io/content-export-import/tutorials-import-bins.html
It was easier to understand by exporting images, examining the output JSON, then move the files & edit lightly then import them back in as new images.
My question is: is there any way to read in binary files directly & build up the other properties along the way instead of pre-building JSON & base-64 encoding the file contents?
Are there other recommended approaches for importing thousands of JPEGs? It would be preferable to let the system configuration create all the imageSet variants during import rather than pre-generating them outside the system - even if that meant import was much slower, at least it would be internally consistent.
If not, are there any scripts that would help generate the JSON files from a folder of JPEG images that already exist?