Problems reading epub on prs-505

[gonzule]

Hello, i just updated my sony reader and really wanted to try some .epub books to see how "wonderful" it was, but after putting the books on the reader and opening them, they display with several question marks ( ? ) scattered all around the pages, making it almost impossible to read.

I thought that maybe the books came corrupted, so after that, i oppened the .epub file and extracted the contents (with winrar) and opened the .html pages with firefox/explorer but the text is fine.

Does anybody else have this problem or have any idea on how to solve it?

Heres a screen capture of the epub book oppened with "eBook Library" on windows.

[akiburis]

A guess, based on the fact that the question marks seem to come between syllables: perhaps your books use the HTML soft-hyphen entity to indicate possible end-of-line word breaks; and perhaps Digital Editions can't interpret it, and renders it as a question mark. You could check by opening the extracted HTML in a text editor (or open it in a browser and use the view-source setting). The fix, if this is the case, would presumably be a global search-and-delete of the entity in the HTML files, since you probably can't change DE's behavior.

[gonzule]

mmm, you are right about the syllabes part, but i opened the file to view the source code but doesnt come with any "soft-hypens" or something indicating the word breaks, probably this is done by the reader or by the "epub software" installed inside it.

i also checked some webpages detailing the process of creation of an epub book but couldnt find anything related to this problem (anyway, i could only understand half of what the explained).

Guess i'll have to get these books from somewhere else. i'll keep searching for a solution thoug.

[kovidgoyal]

try converting the epub to LRF

[gonzule]

Wait, i just found the problem. It was as akiburis said. i opened the .xhtml files again, but this time not with explorer/firefox source code viewer but with MS Wordpad and now it shows the letters generating the problem with the question marks.

Example:
The barÂ*riÂ*er is arÂ*tiÂ*fiÂ*cial, genÂ*erÂ*atÂ*ed by huge alien arÂ*tiÂ*facts. Time is passÂ*ing faster outÂ*side the barÂ*riÂ*er than inÂ*side–more than a hunÂ*dred milÂ*lion years per day on Earth. At this rate, the death throes of the sun are onÂ*ly about forty years in our fuÂ*ture.

So all i have to do now is open the dcuemnts one by one and remove those characters.

thanks everybody

[HarryT]

They're "soft hyphens".

[gonzule]

yeah, that's what i meant, haha.

Anyway, i just removed them from the text, then replaced a lot of "€" signs from the text that where supposed to represent "-" , " ' " , "-" etc (too much work, haha) and then save the document and latter open it with firefox and NOW its firefox who displays question marks:

One night in Oc�to�ber when he was ten years old, Tyler Dupree stood in his back yard and watched the stars go out. They all flared in�to bril�liance at once, then dis�ap�peared, re�placed by a flat,

I saved it and stored it in epub format again and imported it to the reader and now it shows up all blank (no text at all after where the first question mark was)

I checked the encoding on the file and its UTF-8. Tried changing it to all possible combinations with no result. It look's i'll have to transform it to lrf after all.

Will still post a solution if i find one. Maybe its just a bad series of book i downloaded



edit:

i also get an error when opening these books with calibre (last version)

'NoneType' object has no attribute 'metadata'
Detailed traceback:
Traceback (most recent call last):
File "main.py", line 448, in add_books
File "main.py", line 460, in _add_books
File "calibre\ebooks\metadata\meta.pyo", line 82, in get_metadata
File "calibre\ebooks\metadata\__init__.pyo", line 214, in smart_update
File "C:\Program Files\calibre\library.zip\calibre\ebooks\metadata\ opf.py", line 210, in __get__
File "C:\Program Files\calibre\library.zip\calibre\ebooks\metadata\ opf.py", line 239, in get_title
AttributeError: 'NoneType' object has no attribute 'metadata'

[kovidgoyal]

that error indicates a malformed OPF file. Where did you get these epub files from?

[llasram]

Originally Posted by gonzule

Anyway, i just removed them from the text, then replaced a lot of "€" signs from the text that where supposed to represent "-" , " ' " , "-" etc (too much work, haha) and then save the document and latter open it with firefox and NOW its firefox who displays question marks:

Your text editor is not interpreting the file in the proper encoding.

[pilotbob]

Originally Posted by gonzule

but with MS Wordpad and now it shows the letters generating the problem with the question marks.

You should probably get an editor like Notepad2 or Slickedit or Notepadd++ which support proper encoding and unicode. Wordpad isn't the sharpest tool in the shed when it comes to text editing.

BOb

[vivaldirules]

I've upgraded my Sony 500 to the 505 and have installed the new firmware for the Reader and software upgrades to eBook Library and Adobe DE. I have downloaded and can successfully read DRM'd Adobe DE files that I've purchased both on my PC and on my new Reader. I've done the same with Adobe DE ebooks that I've borrowed from digital public libraries. And I've done the same with this ePub ebook that was uploaded to MR: https://www.mobileread.com/forums/showthread.php?t=26872

Great. But:

1. I've purchased a book from Books on Board (http://www.booksonboard.com/index.ph...ok&BOOK=201849) as an "Adobe ePub eBook" which I can read on my PC using Adobe DE and I can transfer from there to my Sony Reader using Adobe DE but when I go to read it on my Reader I get the dreaded "Page Error!" indicator. Is this likely a Sony firmware issue or an indication that the publisher didn't do a good job at making the "ePub" or what?

2. All of the ePub books I've seen so far are not fully justified when read either in Adobe DE or on my Reader. Is this a shortcoming of the ePub format, the reader software, or the publisher who created the files?

3. I've casually searched for ePub books on the internet and have found (from http://www.epubbooks.com/) some which are prominently labelled as ePub (http://www.snee.com/epubkidsbooks/) but are actually RAR files. When I open them, I find NCX, OPF, HTM files and a couple of folders (one with images). So where's "the ePub"? What would I load to my Reader? Or are these the components of an ePub that would need to be put together?

Sorry for asking what may be stupid questions. For the moment, I don't think I'll be bothering with any more of these. At least, I won't be buying any more.

[wallcraft]

Originally Posted by gonzule

they display with several question marks ( ? ) scattered all around the pages

Prompted by this thread, I have posted Soft Hyphens which illustrates that other ebook reading software has problems with soft hyphens too.

[gonzule]

Originally Posted by kovidgoyal

that error indicates a malformed OPF file. Where did you get these epub files from?

i got them from a bittorrents site, the guy said he made them himself, but the files are like a year old, so probably he made a mistake while creating the books or they are not meant to be read on a sony reader (becasue the firmware update is so new)

[wallcraft]

Originally Posted by vivaldirules

3. I've casually searched for ePub books on the internet and have found (from http://www.epubbooks.com/) some which are prominently labelled as ePub (http://www.snee.com/epubkidsbooks/) but are actually RAR files. When I open them, I find NCX, OPF, HTM files and a couple of folders (one with images). So where's "the ePub"? What would I load to my Reader? Or are these the components of an ePub that would need to be put together?

An ePub is all the NCX, OPF, HTM and image files assembled into a "container" which is essentially a ZIP of the files renamed from .zip to .epub. So perhaps RAR has been associated with ePub by mistake (or to allow the file to be exploded into its component parts).

I tested Dame Wonder's Picture Alphabet by Anonymous and it passed epubcheck at threepress with only minor issues.

[llasram]

Originally Posted by vivaldirules

1. I've purchased a book from Books on Board (http://www.booksonboard.com/index.ph...ok&BOOK=201849) as an "Adobe ePub eBook" which I can read on my PC using Adobe DE and I can transfer from there to my Sony Reader using Adobe DE but when I go to read it on my Reader I get the dreaded "Page Error!" indicator. Is this likely a Sony firmware issue or an indication that the publisher didn't do a good job at making the "ePub" or what?

The on-Reader version of Adobe DE punts the technical problem of reading large files using limited memory by giving up (and giving that error) if any HTML file inside the EPUB book is larger than 300k. Publisher’s apparently didn’t get the memo, but Books on Board posted here that they’ve informed the publishers of the issue and they and are in the process of fixing the problem.

Originally Posted by vivaldirules

2. All of the ePub books I've seen so far are not fully justified when read either in Adobe DE or on my Reader. Is this a shortcoming of the ePub format, the reader software, or the publisher who created the files?

Yeah, the current version of Adobe DE doesn’t do justified text at all. But on narrow screens and using naive paragraph-formation algorithms you get better grayness with left-flush text anyway, so no big loss IMHO.

Originally Posted by vivaldirules

3. I've casually searched for ePub books on the internet and have found (from http://www.epubbooks.com/) some which are prominently labelled as ePub (http://www.snee.com/epubkidsbooks/) but are actually RAR files. When I open them, I find NCX, OPF, HTM files and a couple of folders (one with images). So where's "the ePub"? What would I load to my Reader? Or are these the components of an ePub that would need to be put together?

As wallcraft describes, EPUB books are actually ZIP archives. Er, I imagine the person who made the RAR file “EPUB” books used a generic archiver and just accidentally used RAR instead of ZIP. If you convert it to a ZIP archive it should be “good enough” for Adobe DE.