rtf/doc/odt to epub: Couldn't opened in Kobo

[WolfBack]

This problem drives me crazy!!! I try all of my day to find a solution for this...
I have a book that I made some years ago, 640 pages long in .doc formatted.
I want to formatted in .epub, but with no success...

Any of my tries give me an "Ooops! This document couldn't be opened. This document is protected by Adobe Digital Management (DRM) licence witch has expired"
Same occurs on my Aura Classic (3.13.1) and my H2O (3.13.1) in nickel.
(in Koreader the .epub formatted file works fine).

Things I try:

1. .doc to .rtf
2. .rtf to epub with calibre (no success...)

3. .doc to .odt with OpenOffice and then to .epub (with write2epub). No success either.

4. .rtf to .epub with Sigil 0.8.6. (just the same...)

That's it!!!! My hands are tighten!

All the above formatted files works fine with Koreader, but not in Nickel...

Any suggestion?

[oren64]

I also made a book with Office, i copy the file as .docx or .htm to Calibre and convert it to epub.

[WolfBack]

Originally Posted by oren64

I also made a book with Office, i copy the file as .docx or .htm to Calibre and convert it to epub.

I will try that!
I hope that make the things work!
Thnx, oren64!

[eschwartz]

calibre can natively convert DOCX to EPUB -- this will most likely provide the most accurate conversion.

If you are getting false DRM errors, try checking it with Sigil or calibre editor's errorcheckers.


I could swear I have heard of this problem before, but I can't remember where nor find it. Possibly something to do with obfuscated fonts???

[GeoffR]

Originally Posted by WolfBack

Any of my tries give me an "Ooops! This document couldn't be opened. This document is protected by Adobe Digital Management (DRM) licence witch has expired"
Same occurs on my Aura Classic (3.13.1) and my H2O (3.13.1) in nickel.
(in Koreader the .epub formatted file works fine).

I think the DRM error is just the message you get when the book wasn't recognised as a valid ePub for whatever reason, the ePub reader assumes that if it can't find what it expects to find, that it must be because the book is encryped.

Any suggestion?

One thing I would check if you are reloading multiple versions of the same book (same filename) onto the device is that you delete the old version from the device before reloading the new one. Some information about the book is stored in the database, and if the new version doesn't match what is in the database (because it has a different internal file structure, different TOC, etc.) then strange things can happen, just as if the book was corrupted.

The other thing to check, especially with books converted from non-ebook formats, is that the markup is sane. The ePub reader appears to make certain assumptions about the size of objects and can't cope if those assumptions are wrong. With normal books the assumptions always hold, but with very badly converted books they might not. One thing that can go wrong is that the conversion results in a whole chapter being contained in one giant paragraph for example, but the ePub reader assumes that paragraphs are not insanely large.

[PeterT]

Any possibility the date might be wrong on the Kobo?

[WolfBack]

I want to thank all of you guys for your help!

I try the suggestions you gave me and I think (accidentally) find a solution for that.


Converting my file to .htm and then to .epub (via Calibre) did the trick and the file opens without errors in Kobo IF the epub sideloaded on the device directly as it is.

When I drop the file into Sigil for a few adjustments (such as generate TOC and make some chapters) the saved file won't load in Kobo and send me a false DRM error. Something's strange with the structure of the file? The epub itself have a lot of sections (about 65) labeled as: Agkhone2_split_000.htm, Agkhone2_split_001.htm e.t.c.

My final change is to rename the sections as: Section001.htm, Section002.htm e.t.c and voila! (the sun shines again!) The sideloaded epub works fine on my device!

I don't know what wrong is happened... Something with the underscores or the long section names?

Anyway, I can now read my book at nickel... Finally!


Thank you all, again!

[oren64]

Glad it work for you .
i did all changes in the word file add table contents...
you don't need to covert the .doc file as .htm, you can do "save as" .htm in Word.

[WolfBack]

Originally Posted by oren64

Glad it work for you .
i did all changes in the word file add table contents...
you don't need to covert the .doc file as .htm, you can do "save as" .htm in Word.

Thanks oren64!
I'll keep that in mind the next time I face the same problem.

[GeoffR]

Originally Posted by WolfBack

I don't know what wrong is happened... Something with the underscores or the long section names?

Underscores shouldn't cause any problem, any book that as been converted with Calibre will usually have underscores in the filenames, and lots of books from the big publishers have them too.

Filename length shouldn't normally be a problem either, there is probably some limit on length, but it must be a very high limit because some of the books that come from Smashwords have crazy long filenames, and they work okay (well, they usually have a lot of problems that need fixing, but the ePub reader doesn't completely refuse to display them.) Here is a typical example:

tmp_ca8960047058be7ad0839209d3b86dd7_zbbG9l.ch.fixed.fc.tidied.stylehacked.xfixed_split_003.html

Edit: But I thought of one possible problem with the filename length: in some places the firmware might reference the internal files in the epub by their full pathname, i.e. <name of epub>/<name of internal file> so there might be a limit on the total length of that pathname, in which case if the epub filename is very long then a long internal filename could be enough to make the total too long.

How long is the full pathname of the epub once calibre has transferred it to the device? With the default template, Calibre can generate some very long pathnames if the book has multiple authors. (I have changed my template to produce short pathnames that only use the surname of the first author.)

[davidfor]

Originally Posted by WolfBack

I want to thank all of you guys for your help!

I try the suggestions you gave me and I think (accidentally) find a solution for that.


Converting my file to .htm and then to .epub (via Calibre) did the trick and the file opens without errors in Kobo IF the epub sideloaded on the device directly as it is.

When I drop the file into Sigil for a few adjustments (such as generate TOC and make some chapters) the saved file won't load in Kobo and send me a false DRM error. Something's strange with the structure of the file? The epub itself have a lot of sections (about 65) labeled as: Agkhone2_split_000.htm, Agkhone2_split_001.htm e.t.c.

My final change is to rename the sections as: Section001.htm, Section002.htm e.t.c and voila! (the sun shines again!) The sideloaded epub works fine on my device!

I don't know what wrong is happened... Something with the underscores or the long section names?

Sigil shouldn't be causing a problem. It's been a while since I have used it, but it was always a very safe way to get a book to work on a Kobo device. What version of Sigil are you using? Maybe there is something wrong in the recent versions. And did you try using the epub check within Sigil? That might have shown something.

As GeoffR said, undescores should not be a problem. I use them all the time. But, there are other characters that could be at fault. I also don't see the file name length being it. I think we would have seen this before as some of the file names in commercial epubs can be very long.

I'm curious enough to want to try this. Any chance of getting the original RTF or the epub after the conversion? If there are any copyright concerns, don't post it, but send me a PM to arrange something.

[WolfBack]

Originally Posted by GeoffR

How long is the full pathname of the epub once calibre has transferred it to the device? With the default template, Calibre can generate some very long pathnames if the book has multiple authors. (I have changed my template to produce short pathnames that only use the surname of the first author.)

GeoffR, the file name was short enough: "Anghone - Antonis WolfBack.epub".
You're right about the underscores. Lot's of my books have lot's of those and my Aura read the files fine and seamlessly.

Originally Posted by davidfor

I'm curious enough to want to try this. Any chance of getting the original RTF or the epub after the conversion? If there are any copyright concerns, don't post it, but send me a PM to arrange something.

Hahahahaha! I like your courage, but I thing that make you a lot of troubles and headaches. There's no copyright concerns, the book was a 640 pages novel made by myself in my mother language (Greek) and contains lots of special characters and symbols. Trouble to display and editing in a foreign language, thought.
Btw, all seems to work normal on my Aura 6'' after renaming those sections.

My version of Sigil is 0.8.6 (the new version causes me a lot of crashes...), the original .doc file was converted to .htm and then .epub with Calibre. I was made all the necessary step before adding the file on my device (check on both software -Sigil and Calibre- with no errors at all).

One last thing that I will try soon:
Start all over again with the original file. I will make the changes with Sigil (except the renaming thing. Generate chapters and TOC's only) and I'll try the converted .epub on my other device (H2O). Both devices runs on 3.13.1 firmware, but maybe the problem occurs only on Aura 6'' (due some firmware fault...).

I will keep in touch!

Thank you all, guys!