Handle JP2 files natively when performing OCR #31
Conversation
Tesseract does not natively handle JP2 files, so if the OBJ datastream is a JP2 we must create a TIFF from that JP2 to pass into tesseract.
|
One item to note, I'm note seeing anything show up in Solr, but that might be because things are still going through GSearch, but has been almost a day now. So, not quite sure what's going on. |
|
I wouldn't think it would take that long. Are the derivatives created in the object as one would expect? |
|
Yep. It looks like it. |
|
Hmmm -- well the OCR/HOCR derivative creation code isn't anything special. It commits those datastreams back to the Fedora repository, where GSearch should pick up the message that the object has changed and go get the content. Certain that GSearch is configured to pull content from the OCR/HOCR datastreams? Any error messages from GSearch? I find it odd that there are 5 versions of the OCR and HOCR datastreams. Can you account for that? |
|
I'm not seeing anything out of the ordinary in The 5 revisions really threw me too. I have no idea what that is all about. |
|
After some discussion in #islandora, I've come to the conclusion that the Solr/GSearch issue is a nuance of my particular installation, and not from the two related pull requests. 👍 for merge. |
| $jp2_file = islandora_ocr_get_uploaded_file($datastream); | ||
|
|
||
| // Create JP2 with kakadu. | ||
| module_load_include('inc', 'islandora_large_image', 'includes/utilities'); |
There was a problem hiding this comment.
Dependency should be noted in the .info file
|
Sometime, somewhere this already made it into the code base. So, I'll close this. |
Depends on Islandora/islandora_solution_pack_large_image#86