Creating PDFs with links for web searches

Report · Jan 30, 2012

Hi everyone. That title is not exactly the problem I'm having, but it's a good start.

We have found that Firefox/IE/Chrome search our PDFs quite nicely, but when the link is clicked in the search results, all that happens is that the PDF opens to page 1. Users have to search again to find what they're looking for. Perhaps that is just how such a search is supposed to happen, but it seems that I've seen somewhere that a PDF opened to the page with the search term on it.

What can we add to the FrameMaker source files or the PDF (before it's posted to the web) to ensure that a search link takes the user precisely to the text being searched for? We're TCS3 with unstructured FM but might soon be, um, strongly encouraged to switch to the structured version.

I've looked through the FM and Acrobat forums for an answer, to no avail.

Thanks for the help!

Report · Jan 30, 2012

This might get you started, from this forum.

Permanent html bookmarks and/or pdf nameddests

Named Destinations is the key.

Naturally, they don't work quite like you might expect.

Report · Jan 30, 2012

Thanks for the reply, Error. That's not quite what I'm looking for, unfortunately.

Say I'm trying to figure out how to place a test call on a Polycom HDX system. I enter "place test calls HDX" in my search text box and get a few results, one of which is in the Administrator's Guide for Polycom HDX Systems. You can see the text I searched for highlighted in the results, but when you click the link, the PDF just opens to page 1, rather than to the page where that phrase is. It doesn't seem that creating a named destination for everything a user might search for is practical, let alone possible.

I wonder whether it's a limitation of the search engine, rather than FM or Acrobat.

Thoughts, anyone?

Dimi Everette

Report · Jan 30, 2012

> I wonder whether it's a limitation of the search engine, rather than FM or Acrobat.

That would be my guess. Search engines probably cannot search many PDFs in the raw at all (you don't have to turn on much security to turn a PDF binary into an encrypted mess). But if the PDF allows reading and/or printing, they can render the whole thing, and index what's displayed. The named destinations may not be recovered in that process, plus, you're not always assured that there will be an ND nearby everything you're indexing.

Adobe might even provide some APIs that would allow deeper linking into PDFs as a search result, but it may also be that few enough have useful NDs that it just isn't worth the bother to look. NDs inflate the size of the PDF, so many publishers don't turn them on (we don't, and probably won't, until we need deliberate deep linking).

If you want the info to be accessible like a web page, you multi-flow to HTML or a similar non-atomic format as well.