What you are kind of looking for, is an indexer for a search engine. You might look around for small server search engines, and configure them in a way that it indexes the directories you need.
If you happen to be on OSX, you might look at the way you could control Spotlight (via AppleScript), but that would not really work from within a PDF.
If it has to be from a PDF, you would first have to run Catalog which then creates an according index, and then you can search through it.
If you have a specific set of search terms, you could also set the Metadata of the PDF accordingly, and then parse just the metadata, which means that you would not have to open the PDF in a PDF viewer.
Hope this can help.