Content of PDF documents is not indexed for suggestions, while on some ZIM it is the "core" of the ZIM. Having a utility in scraperlib to extract PDF info and get the document title would probably help. See https://github.com/openzim/warc2zim/issues/290 for one use-case.