I somehow got it working. In Dolphin, going to a pdf's properties -> information -> configure and checking "content" revealed that the contents of various pdfs were indexed. One showed nothing so I manually ran nepomukindexer on it, went back to properties -> information in Dolphin and indexed content showed up. Then magically ALL indexed content showed up! Apparently my pdfs were indexed but Dolphin wasn't tapping into the database. I'll see if it continues to work on reboot.

molecule-eye wrote:I somehow got it working. In Dolphin, going to a pdf's properties -> information -> configure and checking "content" revealed that the contents of various pdfs were indexed. One showed nothing so I manually ran nepomukindexer on it, went back to properties -> information in Dolphin and indexed content showed up. Then magically ALL indexed content showed up! Apparently my pdfs were indexed but Dolphin wasn't tapping into the database. I'll see if it continues to work on reboot.

When you select a file with Dolphin is indexed if is not indexed.

molecule-eye wrote:I somehow got it working. In Dolphin, going to a pdf's properties -> information -> configure and checking "content" revealed that the contents of various pdfs were indexed. One showed nothing so I manually ran nepomukindexer on it, went back to properties -> information in Dolphin and indexed content showed up. Then magically ALL indexed content showed up! Apparently my pdfs were indexed but Dolphin wasn't tapping into the database. I'll see if it continues to work on reboot.

I'm seeing the exact same behaviour. I'm searching for 'receptor' in my research papers I downloaded and only those that have the word in their title come up. However it's a word that is in almost all research papers I have, believe me. So I just open one that did not come up, search for the word, find it, then search again in dolphin. However I have to search for the word in every pdf separately in order for dolphin to take it into account. Apparently, it should not work like that.

Can you try running the "nepomukindexer" program against each PDF file, and see if it is completing successfully? Also, is the directory enabled for indexing in System Settings > Desktop Search?

The "issue" with indexing PDFs is that you need to put their text inside the Virtuoso index. However Virtuoso does not like too large elements of text pushed into it, and can go crazy on the CPU, therefore only a part of it is pushed inside (a predefined size starting from the beginning).