[GUI Plugin] EpubSplit

[JimmXinu]

Originally Posted by shamanNS

@JimmXinu: Is it possible to make future version of plugin copy embedded and referenced font files (CSS declaration of font-family remains in CSS after splitting, but .ttf files are left behind) or at least notify user when splitting that omnibus has some font files embedded? So that user knows that he should add those font files manually post-split.

That's a good suggestion.

Can you (or anyone else) share or point me to some epubs containing embedded fonts? I don't think I have any to experiment and test with.

[theducks]

JimmXinu

2 ways I have added back fonts (for personal use) (that were commented out because of license restrictions)

@fonts is in in the:
Pagefile.css
stylesheet (for that segment) CSS


I might attack the issue by determining what is needed per Split:

1) Is there a Font-Family: value, other than generic (serif, sans-serif, monospace) used for any of the stylesheets used in the split off section. << Some anthologies use different sheets per story

2)is there a matching (set) of @fonts in any of the CSS in use that points to an embedded file << I mention this, because it might depend on a 'master' set of fonts on the reading device or the code is just left dirty

[JimmXinu]

Attached is a test version that copies all font files in the source epub into each split epub.

You can use Polish to remove unused fonts afterwards.

[MarjaE]

What if the book doesn't have section or chapter breaks?

[theducks]

Originally Posted by MarjaE

What if the book doesn't have section or chapter breaks?

Then you will probably have to use 'The Editor' if the normal Calibre Convert options offer no solution (there are books that just do not have any clearly defined divide point)

Splitting assumes you have a good starting point.

But... this PI is about splitting OUT pieces (whole files), not breaking up a large file.

If your goal is BOTH

you will need to break the book up Internally first.
A common break is files-per-chapter or file-per-story (eg anthology)

[JimmXinu]

Originally Posted by MarjaE

What if the book doesn't have section or chapter breaks?

EpubSplit can split either on file boundaries or on ToC entries within files.

So you have to use an editing problem to either split files or insert ToC entries before you can use EpubSplit on it. Calibre's builtin editor can do that.

[JimmXinu]

Originally Posted by theducks

But... this PI is about splitting OUT pieces (whole files), not breaking up a large file.

It can split files, but only where there are ToC entries.

[theducks]

Originally Posted by JimmXinu

It can split files, but only where there are ToC entries.

Thank,
I have only used it a couple of times and obviously have not discovered all the features

[JimmXinu]

Originally Posted by JimmXinu

Attached is a test version that copies all font files in the source epub into each split epub.

You can use Polish to remove unused fonts afterwards.

Soooo...

Has anyone had either success or failure using this test version?

[shamanNS]

I've tested it on a dummy epub (2 html files, one paragraph in each, one referencing embedded font one not) created using Calibre Editor and for that epub file plugin works correctly, it copied .ttf file(s). But it's not working for my copy of The Malazan Book of the Dead omnibus 1-10, probably because of non standard internal directories inside of epub...

When you said "you can use Polish to remove unused fonts" were you referring to Calibre's "Polish Books" toolbar button? Because I don't see options for removing unused fonts (files, not CSS), just for embedding referenced fonts and subset-ing embedded fonts.

I'm just wandering if I'm missing out on some Calibre functionality...
I'm used to checking every book in Calibre Editor and while checking/editing CSS I always check if the book contains unwanted fonts (most often deleting Charis SIL font files )

[JimmXinu]

Originally Posted by shamanNS

I've tested it on a dummy epub (2 html files, one paragraph in each, one referencing embedded font one not) created using Calibre Editor and for that epub file plugin works correctly, it copied .ttf file(s). But it's not working for my copy of The Malazan Book of the Dead omnibus 1-10, probably because of non standard internal directories inside of epub...

I don't have any epubs myself with embedded fonts except for a couple I made for testing.

Unless I can get copies of epubs to test with containing other examples, I'm not going to be able to support them...

Originally Posted by shamanNS

When you said "you can use Polish to remove unused fonts" were you referring to Calibre's "Polish Books" toolbar button? Because I don't see options for removing unused fonts (files, not CSS), just for embedding referenced fonts and subset-ing embedded fonts.

Yes, Calibre's "Polish Books" feature. Subseting embedded fonts will also remove font files that aren't used.

[bob_bookworm]

Hi, I keep getting a 'memory error' message while trying to split a massive (120MB) Epub file.
Am I doing something wrong?

[JimmXinu]

Originally Posted by bob_bookworm

Hi, I keep getting a 'memory error' message while trying to split a massive (120MB) Epub file.
Am I doing something wrong?

I assume that you're running into a memory limit.

The biggest epub I have is ~17M and it split fine for me.

I've never tested with anything approaching 120M as even my million+ word books are <10M. I don't even know where I'd get such a thing.

Using a 64bit version of Calibre might help if you're not.

[bob_bookworm]

Originally Posted by JimmXinu

I assume that you're running into a memory limit.

The biggest epub I have is ~17M and it split fine for me.

I've never tested with anything approaching 120M as even my million+ word books are <10M. I don't even know where I'd get such a thing.

Using a 64bit version of Calibre might help if you're not.

It's a 11-book file [Story of Civilization by Will Durant], and I would be surprised if it's depending on either RAM or storing capacity because it is a Win10 64bit new and potent machine.
The same problem in both Calibre 2.50 and 2.51.

I am trying to split the file into smaller epub files because trying to convert the big file from epub to mobi is getting a similar error message.

[JimmXinu]

Originally Posted by bob_bookworm

It's a 11-book file [Story of Civilization by Will Durant], and I would be surprised if it's depending on either RAM or storing capacity because it is a Win10 64bit new and potent machine.
The same problem in both Calibre 2.50 and 2.51.

Which is why I said 'memory limit' instead of 'running out of memory'. I suspect there's a limit somewhere in Calibre or Python (one that seemed a reasonable assumption to the writer at the time) that is being exceeded.

Originally Posted by bob_bookworm

I am trying to split the file into smaller epub files because trying to convert the big file from epub to mobi is getting a similar error message.

Which further suggests that Calibre (or Python) has some internal limitation, unfortunately.

If you run Calibre in debug mode (under Calibre's Preferences drop down), capture a log of the error message and post it here, that may help.