I'm looking to develop an HCI stage for converting scanned documents into text-searchable documents using OCR software.
It's been suggested I use ABBYY's finereader which seems to work well.
So far, I've successfully used their SDK to produce a stand alone app that converts a single tif document into a text-based PDF.
I'd like to migrate that code into a Stage... and I've had some limited success. I was able to create a jar that includes the ABBYY wrapper and can execute via the plugin test harness.
However, I need LD_LIBRARY_PATH to point to the mountain of shared library objects (.so files) so that the underlying JNI function may dynamically link to these .so's.
I should note, I'm not loading the .so's myself. They are loaded implicitly by the ABBYY wrapper jar.
The other issue I have relates to process(), but I'll save that for a separate question.
If there is no good way to do this, one idea I've had is to take the closed source JNI and wrap it into a tomcat/docker container and essentially provide a RESTful micro service interface to the ABBYY SDK.
Thanks in advance!