Since 1.4.2, I noticed that an “Anywhere” search seems to be unable to find pdf files based on text in Skim notes. I tried this both with .pdf+.skim and .pdfd files:
I tried out the link you sent me, but after re-indexing the problem still occurs.
To rule out problems with my library, I’ve created a new one, imported an English pdf with Skim notes containing Dutch words, closed the library and rebuilt the index. Still no luck.
I’ve also re-downloaded EagleFiler to replace my old (but updated) copy, but that doesn’t change anything either. This is really a weird problem.
Apparently, Skim is not able to store the notes in the extended attributes of these Postscript-pdfs (verified using commandline Skim.app/Contents/SharedSupport/skimnotes tool). They can only be found in the backup .skim file (if this is enabled).
Hence, it looks like for some pdf files extended attributes can be used, whereas for others only the .skim file contains the right information.
It turns out to be a PDFKit problem with encodings. For certain pdfs, notes are attached, but contains weird symbols because of the way PDFKit handles text encodings. So, it’s neither a Skim nor an EagleFiler problem.