Sugesstion needed: pdf to swf page flipping in drupal

http://www2.islandpacket.com/special-sections/destination_hilton_head_is...

We used a script that converted PDF > SWF, extacted text (for searchability), made a thumbnail, then imported a CSV file of everything into a 'print page' CCK node. The page flipping is done with a contrib module whose name I need to look up, and the UI is done with jQuery.

If this is similar to what you're looking for, I can write something up in greater detail.

Edit: I started to write something up a year ago... but it's very general:
http://www.christianyates.com/blog/drupal/how-save-crapload-money-conver...

Chris Yates
http://christianyates.com

We wrote a generic python daemon for managing workflows - moving files, running and monitoring processes. We used it for a lot of things, including this task.

That tool handled
• Converting PDF to SWF with swftools
• Extracting text from the PDFs with pdftotext (part of xpdf)
• Creating thumbnail images with imagemagick

Then it ran a perl script that gathered all of this stuff and generated and uploaded an RSS file to a server. Then came the only human interaction - someone would manually import the feed into Drupal (with simplefeed IIRC).

No reason this couldn't all be built into a module that monitored an import directory and imported PDFs via functions called from hook_cron. At the time, it was faster to do it offline.

I chose to make each page a node for the purpose of searchability, so that a user could run a keyword search and go straight to the page. If that's not important, you could use swftools to create a SWF from a multipage PDF, and use javascript and/or actionscript to implement the flipping.

Of course, searchability is at the mercy of the quality of the PDF. In our case, a lot of the ads were put into our advertising front end (which generated the PDFs) as TIFFs. This is why we considered including OCR in the process. Nowadays, you could also look at Robert Douglass' excellent apachesolr module, which allows attachment searching, and include the original (or a downsampled) PDF file in the feed.

I'd honestly build it all in Drupal as a module that called the external binaries that did the processing steps.

I don't really have any code I could share, because the workflow manager was a multipurpose tool we had lying around, and the Perl script I used to assemble the products into an RSS feed was pretty one-off.

I'd think you could use some functions called from hook_cron in a drupal module to do the same thing, and skip the packaging into XML step - fire off a function to convert ODF files in a directory into images and text, and another to watch a folder where those files get dumped, and build nodes with file path references in CCK, images in image.module, or file attachments, then write a custom .tpl file to theme that content type.

Was just searching for a similar solution and found Page Flip.