BookReader OCR search sometimes fails

Project:RUcore/NJDH/Partner Portal Search
Version:8.1
Component:User Interface - Service Agent
Category:bug report
Priority:normal
Assigned:chadmills
Status:closed
Description

When using the BookReader OCR search in conjunction with ABBYY formatted OCR depending on the ABBYY results the BookReader search option may hang/never complete.

The issue is with blank OCR entries in the ABBYY OCR that are concatenated together to form blank "word" entries for analysis.

I have a fix but I don't think it is critical for 8.1; although a very suitable 8.1.1 candidate. This was uncovered using a separate test case on staging that was not used on test. Konin will ingest that staging test case on the test machine and test using that.

On staging the troublesome resource is: https://rucore-staging.libraries.rutgers.edu/rutgers-lib/201542/JPEG/rea...

Comments

#1

Same content from staging was replicated here:

https://rucore-test.libraries.rutgers.edu/rutgers-lib/204080/JPEG/read/

Please test.

#2

Assigned to:pkonin» chadmills
Status:test» active

Works as it should on test server.

#3

Status:active» fixed

#4

The variety of the TIFF sources in production will most likely expose this as a more common issue if not addressed in 8.1. I didn't fully appreciate this when suggesting it might be a worthwhile 8.1.1 candidate.

#5

Status:fixed» closed

Back to top