why ePub -> PDF pages as images?

xristy · 12-25-2010, 10:12 PM

I have occasion to want to generate a searchable PDF from an ePub and thought that would be trivial in Calibre (0.7.35); however, it seems like what is generated from an ePub is a PDF of page images rather than searchable text. Am I doing something wrong or is this just how it is?

Manichean · 12-26-2010, 05:34 AM

That should only happen if the ePub consists only of images in the first place.

xristy · 12-26-2010, 07:55 PM

That is what I thought; however, I have run several tests on pure text ePub documents and the resulting PDFs appear to be sequences of images. When I try to select text for highlighting I get an alert:

Quote:

This page contains only an image of a scanned page. There are no text characters. Would you like to run character analysis to try to make the text on this page accessible?

I've attached a simple ePub and PDF generated by Calibre from the ePub to illustrate what I'm seeing. Calibre is version 0.7.35 running on Mac 10.6.5.

user_none · 12-26-2010, 09:43 PM

Non searchable text and PDF links (internal and external) are a limitation of Qt's PDF printing which calibre uses for PDF output.

AGB · 12-26-2010, 10:01 PM

Quote:

Originally Posted by user_none

Non searchable text and PDF links (internal and external) are a limitation of Qt's PDF printing which calibre uses for PDF output.

You've posthumously added to my list of why I left the platform a couple of years ago.

Not that that will help the OP one bit, but man, as if the list of workarounds-to-make-things-function-properly wasn't long enough in OS X.

user_none · 12-27-2010, 06:47 AM

Quote:

Originally Posted by AGB

You've posthumously added to my list of why I left the platform a couple of years ago.

Not that that will help the OP one bit, but man, as if the list of workarounds-to-make-things-function-properly wasn't long enough in OS X.

This doesn't have anything to do with OS X. This is an issue wtih Qt (the cross platform tool kit by Nokia that calibre uses for various things like the GUI). This same issue will be on all platforms.

AGB · 12-27-2010, 07:12 AM

Quote:

Originally Posted by user_none

This doesn't have anything to do with OS X. This is an issue wtih Qt (the cross platform tool kit by Nokia that calibre uses for various things like the GUI). This same issue will be on all platforms.

Oh, I thought Qt was short for Quicktime. In my world, "QT" is Quicktime. Guess I was way too many years on that sucky platform (more than two decades), lol.

Manichean · 12-27-2010, 07:42 AM

Well, QT is QuickTime, and Qt is the framework

AGB · 12-27-2010, 08:03 AM

Quote:

Originally Posted by Manichean

Well, QT is QuickTime, and Qt is the framework

Yeah, well, I learn every day

DoctorOhh · 12-27-2010, 09:52 AM

For the record: Windows XP, calibre 0.7.35 and the PDF file you attached was non-searchable and I was unable to cut and paste from it.

Quote:

Originally Posted by xristy

I've attached a simple ePub and PDF generated by Calibre from the ePub to illustrate what I'm seeing. Calibre is version 0.7.35 running on Mac 10.6.5.

When I convert your sample ePub to pdf (using calibre) and open it in Adobe, I am able to copy and paste the text from it and I am able to search it. (see attached)

PDF output settings

Paper Size: Letter
Orientation: Portrait

Spoiler:

Quote:

Originally Posted by user_none

Non searchable text and PDF links (internal and external) are a limitation of Qt's PDF printing which calibre uses for PDF output.

Are you saying calibre's viewer has this limit or the resultant pdf? I converted the example file to pdf and can search it and copy the text fine.

user_none · 12-27-2010, 10:21 AM

Quote:

Originally Posted by dwanthny

Are you saying calibre's viewer has this limit or the resultant pdf? I converted the example file to pdf and can search it and copy the text fine.

The PDF. I have not seen it produce a searchable file in Linuxor OS X. However, I have never tried in Windows. I will need to investigate this further...

kovidgoyal · 12-27-2010, 12:12 PM

That's weird, I converted you epub to pdf in linux and calibre generated a text pdf for me, see attached.

EDIT: Crappy hotel internet is not letting me upload the file, but it has selectable text as per okular.

user_none · 12-27-2010, 12:33 PM

Quote:

Originally Posted by kovidgoyal

That's weird, I converted you epub to pdf in linux and calibre generated a text pdf for me...

Hm... PDF output is producing non-searchable files. If I use print to PDF via the ebook viewer I get a searchable file. The ebook viewer produces the PDF using the same technique as conversion...

Edit: I'm using OS X.

kovidgoyal · 12-27-2010, 12:38 PM

dont have access to an OS X env right now, but it maybe that QtWebKit does something different on OS X

user_none · 12-28-2010, 07:10 PM

I've been playing with it and I've run into some major OS X issues with Qt's printing. I'm still stumped as to why ebook-convert is producing non-searchable PDFs and ebook-viewer can produce searchable ones...

12-25-2010, 10:12 PM	#1
xristy Connoisseur Posts: 54 Karma: 210 Join Date: Sep 2007 Device: iPad	why ePub -> PDF pages as images? I have occasion to want to generate a searchable PDF from an ePub and thought that would be trivial in Calibre (0.7.35); however, it seems like what is generated from an ePub is a PDF of page images rather than searchable text. Am I doing something wrong or is this just how it is?

12-27-2010, 12:12 PM	#12
kovidgoyal creator of calibre Posts: 43,851 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	That's weird, I converted you epub to pdf in linux and calibre generated a text pdf for me, see attached. EDIT: Crappy hotel internet is not letting me upload the file, but it has selectable text as per okular. Last edited by kovidgoyal; 12-27-2010 at 12:14 PM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Classic Split PDF pages into smaller pages (images into tiles)	Astro	Barnes & Noble NOOK	4	06-12-2020 10:56 AM
Calibre PDF>>EPUB N° de pages	micheljo	Software	2	10-13-2011 02:03 PM
PDF to Epub - Images with Text	ebahm	Calibre	2	09-19-2010 03:23 PM
PDF to Epub (problem with pages)	violentlyserene	Calibre	1	08-22-2010 10:38 AM
Sony reader for PDF files: pages as images	claudioita	Sony Reader	3	07-30-2007 02:46 PM

12-26-2010, 05:34 AM	#2
Manichean Wizard Posts: 3,130 Karma: 91256 Join Date: Feb 2008 Location: Germany Device: Cybook Gen3	That should only happen if the ePub consists only of images in the first place.

12-26-2010, 09:43 PM	#4
user_none Sigil & calibre developer Posts: 2,488 Karma: 1063785 Join Date: Jan 2009 Location: Florida, USA Device: Nook STR	Non searchable text and PDF links (internal and external) are a limitation of Qt's PDF printing which calibre uses for PDF output.

12-27-2010, 07:42 AM	#8
Manichean Wizard Posts: 3,130 Karma: 91256 Join Date: Feb 2008 Location: Germany Device: Cybook Gen3	Well, QT is QuickTime, and Qt is the framework

12-27-2010, 12:38 PM	#14
kovidgoyal creator of calibre Posts: 43,851 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	dont have access to an OS X env right now, but it maybe that QtWebKit does something different on OS X

12-28-2010, 07:10 PM	#15
user_none Sigil & calibre developer Posts: 2,488 Karma: 1063785 Join Date: Jan 2009 Location: Florida, USA Device: Nook STR	I've been playing with it and I've run into some major OS X issues with Qt's printing. I'm still stumped as to why ebook-convert is producing non-searchable PDFs and ebook-viewer can produce searchable ones...

Advert

Advert