I have the Extract Text module installed on an Omeka-S installation and when uploading pdfs the OCR content is correctly taken and put in the ‘extracted text’ property of an Item. The problem im having is the search function does not seem to look at this property. Both…
‘Search full-text’ and ‘Search by value’ (=extracted text)…
return zero results when searching for words that are definitely in the extracted text property.
This is strange. There is functionally no difference between extracttext:extracted_text and any other property. Can you make a successful search otherwise?
I can’t reproduce this for either “full text” or “by value”. Everything works as expected.
Are your extracted text values marked as private? Private values will not show up during a “full text” search. Even so, private values should show up for a “by value” search. Are you sure your “by value” search is correct (e.g. using “contains” instead of “is exactly”)?
Id set the ‘extracted text’ to private because when the text appeared on an Item page, it made the page so long.*
This meant when not logged in (ie using site as public user) both ‘search full text or ‘search by value’ did not work.
When I was logged-in the search by value did work, but search full text did not (as you pointed out)
I have since made the ‘extracted text’ as public. Now ‘search full text’ and ‘search by value’ both get results.
*extract text making the page so long. Basically I wanted to make the extracted text hidden instead of private. Or instead of hidden at least display it below all other properties including the Universal Viewer, on the Item page. I imagine this would be the preference of most users so if anyone has tips on how theyve done this i would be keen to hear! I don’t know if it can be done using Blocks Disposition or if it requires CSS editing