In my quick experiments, Nutch 0.7.1 (with bundled PDFBox which I thought wouldn't handle multibyte text) can index Chinese PDFs but not Japanese PDFs. Japanese PDFs can be hit by entering an English word that happens to appear in them but the digest lines are shown corrupted. Apparantly, PDFBox is having problem extracting text only from Japanese PDF but not other language PDFs. It's not a multibyte vs single byte issue. Has anybody has any success indexing Japanese PDFs?
-kuro
