Glo + Shelves + different formats

[FencepostErr1r]

Hello everyone,
Is there any way to generate separate shelves for different file formats?
Say I've got Book A as an original_epub and I converted it using Calibre to epub.
Book 'A' is placed in a shelve called 'B'.
When sending the book to the device, I'd like to have both versions in two different shelves called ePub and original_ePub.

Is this possible and if it is, can I automate this?

Why all this madness?
I was stuck one too many times with the broken original ePub I pulled off my favourite fan-fiction site and unable to fix it right away by running it through the conversion in calibre and retransferring it to my device.

Yes, I know that there is a fan-fiction download plugin - no this does not help my problem.

Gods of Calibre, please enlighten me for I am hopeless.

-FencepostErr1r

[eschwartz]

No device will read files of type ".original_epub" ?

[FencepostErr1r]

Originally Posted by eschwartz

No device will read files of type ".original_epub" ?

This is only how Calibre displays the originally added version of the ePub, it still carries the extension ".ePub".

To clarify I'll append a screenshot of what I was referring to.



-FencepostErr1r

/edit: And now it hit me to look up the containing folder. Damn, but this shouldn't be a problem? We'd just have Calibre change the extension back for the transfer...

[eschwartz]

calibre has no facility for changing the extension on the books it sends. You can, however, copy the files manually, or create a new book record for them.

Or write a custom plugin to transfer ORIGINAL_EPUB, similar to the KoboTouchExtended plugin.

[BetterRed]

@FencepostErr1r - Maybe you could store the 'original' epubs in a separate physical library (as epubs) and export them from there to a separate shelf on your Glo.

But if its 'broken' then I assume it would it would not work on your Glo, so why would you...

BR

[davidfor]

I can't think of an automated way to do this, but I can think of some things to help.

Firstly, as the others have said, calibre has no way to send the ORIGINAL_EPUB version of a book to a device. And the Kobo has no way of processing it without the extension being renamed.

What it sounds like you are doing is getting a new epub from somewhere and sending it to the Kobo. Then, when you read it, you realise there are problems. You fix those in calibre and resend the book. That replaces the original version with the new version. But, you want to have both on the device, just in case.

If that is right, what I can think of is some tricks with some Reading Lists and the send to template. So:

1. Create a custom column called "fix_state". It needs to be a text column.
2. Install the Reading List plugin and create two lists. Something like "Unfixed" and "Fixed". In the configuration for these, have them set the column "fix_state" to values like "Unfixed" and "Fixed" respectively.
3. In the driver configuration, set the save template to something like:
   

   {author_sort}/{title} - {authors}-{#fix_state}

   and add "#fix_state" to the list of columns used for the shelves. If you already have columns there, you will need to separate them with commas.
4. When you get a new epub, add it to the "Unfixed" reading list.
5. Send new epub the to the device. The file name will end with "-Unfixed.epub" and it will be put on a shelf called "Unfixed".
6. When you fix the epub, move it from the "Unfixed" reading list to the "Fixed" reading list.
7. Send the fixed epub to the device. The file name will end with "-Fixed.epub" and it will be put on a shelf called "Fixed".

That is the nearest thing to what I think you want is. But, it won't work. When I got to the last line, I realised that while the two different versions of the book will both be on the device, the driver will put both of them on the same shelf. When you send the fixed version, both will be put onto the "Fixed" shelf and there will be no way to differentiate them.

The only thing I can think of to do is to either use a plugboard to put the contents of #fix_state into the title of the book as well as the save template. When it is sent, the new version will have the state in the title and the file name. Then you will be able to find it.

After writing all this, I hope I have misunderstood what you want to do. I can't think of a reason to keep the old version of the book on the device. If I find a problem in a book I am reading, I annotate the problem and fix it when I get back to the PC. Then I simply send them to the device in place of the version that is there and continue reading from where I was up to. And mark any other errors I find. I do keep the original version around, but it is on the PC where I can recover it if I need to.

[FencepostErr1r]

Hi,
Davidfor, you might just have solved my problem! I could make do without a special shelve for original/fixed books, if I can differentiate them in my usual shelves.
I'm off to work now, but this looks promising.

-FencepostErr1r

[FencepostErr1r]

Originally Posted by davidfor

I can't think of a reason to keep the old version of the book on the device. If I find a problem in a book I am reading, I annotate the problem and fix it when I get back to the PC. Then I simply send them to the device in place of the version that is there and continue reading from where I was up to. And mark any other errors I find. I do keep the original version around, but it is on the PC where I can recover it if I need to.

When I notice that the book is broken, it's too late. My computer is either out of reach or I don't have a suitable cable or I am too lazy to get up from my couch.
I'd really like just to be able to switch to a pre-fixed version that is already on the device when necessary.

-FencepostErr1r

[davidfor]

Originally Posted by FencepostErr1r

When I notice that the book is broken, it's too late. My computer is either out of reach or I don't have a suitable cable or I am too lazy to get up from my couch.
I'd really like just to be able to switch to a pre-fixed version that is already on the device when necessary.

You have me baffled. It sounds like you are getting a new epub and fixing problems in it. Then you want to send the original unfixed version and the fixed version to the device. But, you are going to read the unfixed version and only change to the fixed version if you find problems.

Or do I have it backwards. You want to read the fixed version and revert to the unfixed if you find problems you didn't fix?

Part of my confusion is that I'm not sure what you mean by "pre-fixed version". Is that "pre" as in "before the fix" or "pre" as in "one I prepared earlier".

In either case, I'm not sure of the point. I edit just about any epub I get to clean up the code and make them look good on the device. I even have a calibre plugin to automate this for books from a couple of sources. The only reason I put the unedited version on my devices is for the curiosity of seeing how they look with the original code. Unless I have completely screwed up the changes, the version that gets to the device is always better than the original. Reverting back to the original version wouldn't help me.

But, if that's what you want to do, I think the easiest way to do it is to treat the original and fixed versions as two books. Add the original book to calibre, change the title or author to include something to say it is the unedited version. Then add the book again and edit it. Then you can send both versions to the device and they will be treated as different books. Then you can swap between them with no problems.

[cybmole]

Originally Posted by davidfor

But, if that's what you want to do, I think the easiest way to do it is to treat the original and fixed versions as two books. Add the original book to calibre, change the title or author to include something to say it is the unedited version. Then add the book again and edit it. Then you can send both versions to the device and they will be treated as different books. Then you can swap between them with no problems.

in the rare case that a conversion seems problematic, or is causing one or more devices to glitch, then I re-add the original book to calibre & append "original formatting" to the title. that way I don't have to wrestle with original_epub formats and I can see what happens when unedited book meets device.
instances:
Stephen King:Wizard & Glass - which had some very convoluted chapter headers code
other books ( I forget which ones) where chapter was used as a style class name, thus confusing calibre on its default conversion settings. One side effect of that glitch can be that chapters get split into 2 files, one with just the header, one with everything else.

[FencepostErr1r]

Originally Posted by davidfor

You have me baffled. It sounds like you are getting a new epub and fixing problems in it. Then you want to send the original unfixed version and the fixed version to the device. But, you are going to read the unfixed version and only change to the fixed version if you find problems.

You are right on the money with this!

Where should I put up a bounty for a plugin that provides this functionality?
I'll flesh out the requirements there.

Thanks for your hints though!
-FencepostErr1r

[davidfor]

Originally Posted by FencepostErr1r

You are right on the money with this!

Really? Why fix them if you don't actually read the fixed version?

Where should I put up a bounty for a plugin that provides this functionality?
I'll flesh out the requirements there.

Apart from the plugin ideas thread, I don't know of a place.

[FencepostErr1r]

Originally Posted by davidfor

Really? Why fix them if you don't actually read the fixed version?

Why have a backup parachute if you have packed your main one well enough?

I should've described the kind of problem that happened to me before asking for a belt and suspenders solution.

The affected ePubs can not be recognised immediately, because they render just fine. The flaw they have usually shows up at the end of the first chapter.
Something about how the chapter-breaks are set up makes whole sections and sometimes most of the remaining chapters disappear.

This is frustrating as hell - you get sucked into a story and with the next page turn you find that you suddenly skipped forty pages.

Those ePubs are displayed just fine in Calibre. I like my Glo, I can't go use a Kindle again because I need backlighting.

I found that performing a ePub conversion in Calibre fixes this.
I'm a bit apprehensive of just carpet-bombing my library with ePub conversions, because that could lead to other problems.

I'll link this thread in the plugin forum, thanks!

-FencepostErr1r

[davidfor]

Originally Posted by FencepostErr1r

Why have a backup parachute if you have packed your main one well enough?

I should've described the kind of problem that happened to me before asking for a belt and suspenders solution.

The affected ePubs can not be recognised immediately, because they render just fine. The flaw they have usually shows up at the end of the first chapter.
Something about how the chapter-breaks are set up makes whole sections and sometimes most of the remaining chapters disappear.

This is frustrating as hell - you get sucked into a story and with the next page turn you find that you suddenly skipped forty pages.

Those ePubs are displayed just fine in Calibre. I like my Glo, I can't go use a Kindle again because I need backlighting.

I found that performing a ePub conversion in Calibre fixes this.
I'm a bit apprehensive of just carpet-bombing my library with ePub conversions, because that could lead to other problems.

I've seen problems like that. Generally, I have found it to be incorrect nesting of tags. A conversion usually fixes it, but, it can mean some loss of formatting. The calibre viewer is more forgiving of these sort of errors, but again, some formatting might be lost. Instead of a conversion, try opening it in the editor and use check book to find errors. Hopefully it will be obvious and easy to fix.

As to loading both versions on the Glo, I don't have a better solution than what I suggested.

[FencepostErr1r]

Hello davidfor,
Thank you for pointing me in the right direction!
It has something to do with the usage of the italic tag in conjunction with the paragraph tag.
It seems that once the Glo's parser hits this error it skips the rest of the chapter.
I have crafted a demo of this bug. Please see the attached ePub.
Is there any way to automate the fix needed for this?

Thanks,

-FencepostErr1r