EPUB output - Page 29

llasram · 03-19-2009, 01:27 PM

Quote:

Originally Posted by tompe

I checked the 4.0 specification and a number is allowed (the value of name shoulc be CDATA) and the # reference is an URI and there numbers is allowed.

The relevant place to check is whether or not @id and @name attributes (type ID and NAME) are allowed to begin with a number, which they are not. See http://www.w3.org/TR/html401/types.html#h-6.2 for details. (XHTML, which is the relevant spec for EPUB, does not differ from HTML 4.0 on this point.)

ChaosInc · 03-20-2009, 12:02 AM

It does make some type of sense though, as it ultimately becomes an Xml, and those tags are a bit different from normal htm tags, guess I will have to learn xml now. lets see, no a href, now item href, hummmm.....

Lets see, that makes PML, Html, and learning how to use python... I guess I learn another one.

I have looked over the file(s) and I do now know what happened. It just seems stupid sometimes that each format of an ebook file needs different anchors. Can't they be more alike so as to be easier on the person who makes the html that they all come from?

Example:

Lit hated the page-break-before if the h1 were Id'ed, and would not break before the chapter heads/titles. It did want to have a cover htm. So separate htm for that ebook. (1 so far)
Mobi/prc had to add to the anchor points to get them to work in its own way. Had the cover separate, and with the ToC separate also, another htm to create that one. (2 so far)
The lrf seems to create extra anchors. Would seperate the Chapter 1 from the Chapter title by a page break (3 so far)
The pdf created its own anchors and then called them bookmarks. The file had to have the cover picture included in the htm also. (4 htm's so far)
The ePub hated the idea of having a size specified for the headings and created it's own links, that didn't work in portable devises but seemed fine on the desktop.

And on and on...

That's a lot of work to create one good ebook in various formats.

DaleDe · 03-27-2009, 11:40 AM

One problem that has surfaced is the need for a unique ID to mark the ePUB document. Some readers have a problem if the unique ID is not unique. A UUID or ISBN number would be a good choice but there needs to be the ability in Calibre to set or change this number. Hadien has identified this as a serious problem with Calibre processing of ePUB.

Dale

kovidgoyal · 03-27-2009, 01:49 PM

calibre uses a uuid on its generated EPUB files. And if you specify an ISBN in the metadata it will also be in the EPUB files, so I have no idea what Hadrien is talking about.

tompe · 03-27-2009, 02:06 PM

Quote:

Originally Posted by llasram

The relevant place to check is whether or not @id and @name attributes (type ID and NAME) are allowed to begin with a number, which they are not. See http://www.w3.org/TR/html401/types.html#h-6.2 for details. (XHTML, which is the relevant spec for EPUB, does not differ from HTML 4.0 on this point.)

Aha, there was additional restriction. So then this is a bug in Calibre that it generated ePub files that do not follow the specification.

starrigger · 03-30-2009, 06:26 PM

Another question about EPUB output:

I've noticed that in many of the files I've converted to epub for my own reading on the Sony, centering of text seems to be lost, and everything winds up left-justified. This seems to occur mostly in files I've converted from .prc and .rtf, I think. Is that just a matter of the source files not being good enough for more accurate output?

kovidgoyal · 03-30-2009, 07:32 PM

Most probably, hard to say without seeing one of these files

CommanderROR · 03-30-2009, 09:02 PM

I'm pretty new to the sony reader and the ePub file format. I was used to the ease of use the mobipocket Reader software offered.
Calibre seems to be the only real choice for ePub conversion out there right now, but I've had some problems with file conversion from HTML and PDF to ePub with Calibe throwing out errors along the way or producing files that don't work very well on the Sony Reader.

However, I made a very interesting discovery today...I was trying to get a lengthy .pdf file into ePub format for comfortable reading on the Sony 505, but it never really worked with calibre (did no want to handle the 3900 Page PDF properly).

What I did was the following...I threw the PDF file at mobipocket Reader and then took the .prc and converted it to ePub in Calibre...and voila...the file is only half the size (down from 16MB to 7MB), all the formatting looks fine and nothing seems to be missing.
I have not checked the entire file of course (it's over 6000 Pages on the Reader after all), but as far as I can tell this combination seems to do the trick.
I guess Mobipocket creates a clean base for Calible to work with and everybody is happy...although I would like to eventually have something like Mobipocket Reader for ePub files...maybe Calbre will be able to fill that gap eventually...

kovidgoyal · 03-30-2009, 10:35 PM

You can actually convert the file to MOBI in calibre and then the MOBI to eoub to acheive the same effect. The reason it happens is that conversion to MOBI throws away a lot of extraneous markup. In your case it doesn't matter since that markup was not important, but in general converting to MOBI forst and EPUB from the MOBI, you could lose some formatting.

CommanderROR · 03-31-2009, 05:28 AM

hmm...that is strange...I tried to convert the PDF to mobi in Calibre but it would not complete the job...maybe I'll give it another try sometime.

JSWolf · 03-31-2009, 07:57 AM

Quote:

Originally Posted by starrigger

Another question about EPUB output:

I've noticed that in many of the files I've converted to epub for my own reading on the Sony, centering of text seems to be lost, and everything winds up left-justified. This seems to occur mostly in files I've converted from .prc and .rtf, I think. Is that just a matter of the source files not being good enough for more accurate output?

In a case like that, I would use mobi2oeb to then have a look at the HTML and see what's going on. Mobipocket HTML can be a real mess sometimes.

DaleDe · 03-31-2009, 09:59 AM

Quote:

Originally Posted by CommanderROR

hmm...that is strange...I tried to convert the PDF to mobi in Calibre but it would not complete the job...maybe I'll give it another try sometime.

PDF is the most difficult file format to use as a source for conversion. Many times a particular file will not convert properly and it seems to vary depending on the conversion tool. PDF is just a finicky format, intended only for full size page layup. Never make a judgment based on one PDF sample.

Dale

starrigger · 04-01-2009, 01:16 AM

Quote:

Originally Posted by JSWolf

In a case like that, I would use mobi2oeb to then have a look at the HTML and see what's going on. Mobipocket HTML can be a real mess sometimes.

I'll try that the next time I see it happen. Mostly it's in books I just want to throw onto my Reader, and it's not like I'm going to go to extra trouble to fix them up. But it would be good to understand what's going on.

Just as I'd like to understand why, in mobipocket format particularly (but others, too), blank lines spaces--such as between a chapter head and the following text--so often get closed up. (For example, in my own Eternity's End, which I recently was preparing for release, I found that the line space went away in the .prc file made by Mobi Creator, but not in the .mobi file made by Calibre. In that case, there were nonbreaking spaces in the blank lines.)

I guess I just strayed from the Epub topic. Sorry.

darkmonk · 04-03-2009, 11:28 AM

Quote:

Originally Posted by kovidgoyal

You can actually convert the file to MOBI in calibre and then the MOBI to eoub to acheive the same effect. The reason it happens is that conversion to MOBI throws away a lot of extraneous markup. In your case it doesn't matter since that markup was not important, but in general converting to MOBI forst and EPUB from the MOBI, you could lose some formatting.

This sounds really interesting - size is important, but some formatting is too. so, I ask if these things will remain:

Pictures
tables
chapters
font sizes

note that my input format will be pdf...

kovidgoyal · 04-03-2009, 12:15 PM

Quote:

Originally Posted by darkmonk

This sounds really interesting - size is important, but some formatting is too. so, I ask if these things will remain:

Pictures
tables
chapters
font sizes

note that my input format will be pdf...

pictures may or may not depending on the PDF. Tables probably not. Fonts sizes and so on should remain.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[Old Thread] Epub Output: Line Height	greenapple	Conversion	20	01-27-2013 09:27 AM
EPUB output justification	toki08	Calibre	10	01-08-2011 04:14 PM
Calibre epub output details and Nook	squidward	Calibre	6	11-24-2010 03:21 PM
epub output metadata	troymc	Calibre	5	05-22-2010 12:23 AM
Problem with epub output in Cybook Gen3	fjf	Calibre	3	02-03-2010 02:23 AM

03-20-2009, 12:02 AM	#422
ChaosInc Enthusiast Posts: 43 Karma: 31740 Join Date: Feb 2009 Location: Washington State Device: iPaq 3800 series	It does make some type of sense though, as it ultimately becomes an Xml, and those tags are a bit different from normal htm tags, guess I will have to learn xml now. lets see, no a href, now item href, hummmm..... Lets see, that makes PML, Html, and learning how to use python... I guess I learn another one. I have looked over the file(s) and I do now know what happened. It just seems stupid sometimes that each format of an ebook file needs different anchors. Can't they be more alike so as to be easier on the person who makes the html that they all come from? Example: Lit hated the page-break-before if the h1 were Id'ed, and would not break before the chapter heads/titles. It did want to have a cover htm. So separate htm for that ebook. (1 so far) Mobi/prc had to add to the anchor points to get them to work in its own way. Had the cover separate, and with the ToC separate also, another htm to create that one. (2 so far) The lrf seems to create extra anchors. Would seperate the Chapter 1 from the Chapter title by a page break (3 so far) The pdf created its own anchors and then called them bookmarks. The file had to have the cover picture included in the htm also. (4 htm's so far) The ePub hated the idea of having a size specified for the headings and created it's own links, that didn't work in portable devises but seemed fine on the desktop. And on and on... That's a lot of work to create one good ebook in various formats.

03-27-2009, 11:40 AM	#423
DaleDe Grand Sorcerer Posts: 11,470 Karma: 13095790 Join Date: Aug 2007 Location: Grass Valley, CA Device: EB 1150, EZ Reader, Literati, iPad 2 & Air 2, iPhone 7	One problem that has surfaced is the need for a unique ID to mark the ePUB document. Some readers have a problem if the unique ID is not unique. A UUID or ISBN number would be a good choice but there needs to be the ability in Calibre to set or change this number. Hadien has identified this as a serious problem with Calibre processing of ePUB. Dale

03-27-2009, 01:49 PM	#424
kovidgoyal creator of calibre Posts: 45,382 Karma: 27756918 Join Date: Oct 2006 Location: Mumbai, India Device: Various	calibre uses a uuid on its generated EPUB files. And if you specify an ISBN in the metadata it will also be in the EPUB files, so I have no idea what Hadrien is talking about.

03-30-2009, 06:26 PM	#426
starrigger Jeffrey A. Carver Posts: 1,355 Karma: 1107383 Join Date: Aug 2008 Location: Massachusetts, USA Device: Lenovo Yoga Tab Plus, Droid phone, Nook HD+	Another question about EPUB output: I've noticed that in many of the files I've converted to epub for my own reading on the Sony, centering of text seems to be lost, and everything winds up left-justified. This seems to occur mostly in files I've converted from .prc and .rtf, I think. Is that just a matter of the source files not being good enough for more accurate output?

03-30-2009, 07:32 PM	#427
kovidgoyal creator of calibre Posts: 45,382 Karma: 27756918 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Most probably, hard to say without seeing one of these files

03-30-2009, 09:02 PM	#428
CommanderROR eink fanatic Posts: 2,022 Karma: 4924 Join Date: Mar 2006 Location: Germany Device: STAReBOOK, iRex Iliad, Sony 505, Kindle 2	I'm pretty new to the sony reader and the ePub file format. I was used to the ease of use the mobipocket Reader software offered. Calibre seems to be the only real choice for ePub conversion out there right now, but I've had some problems with file conversion from HTML and PDF to ePub with Calibe throwing out errors along the way or producing files that don't work very well on the Sony Reader. However, I made a very interesting discovery today...I was trying to get a lengthy .pdf file into ePub format for comfortable reading on the Sony 505, but it never really worked with calibre (did no want to handle the 3900 Page PDF properly). What I did was the following...I threw the PDF file at mobipocket Reader and then took the .prc and converted it to ePub in Calibre...and voila...the file is only half the size (down from 16MB to 7MB), all the formatting looks fine and nothing seems to be missing. I have not checked the entire file of course (it's over 6000 Pages on the Reader after all), but as far as I can tell this combination seems to do the trick. I guess Mobipocket creates a clean base for Calible to work with and everybody is happy...although I would like to eventually have something like Mobipocket Reader for ePub files...maybe Calbre will be able to fill that gap eventually...

03-30-2009, 10:35 PM	#429
kovidgoyal creator of calibre Posts: 45,382 Karma: 27756918 Join Date: Oct 2006 Location: Mumbai, India Device: Various	You can actually convert the file to MOBI in calibre and then the MOBI to eoub to acheive the same effect. The reason it happens is that conversion to MOBI throws away a lot of extraneous markup. In your case it doesn't matter since that markup was not important, but in general converting to MOBI forst and EPUB from the MOBI, you could lose some formatting.

03-31-2009, 05:28 AM	#430
CommanderROR eink fanatic Posts: 2,022 Karma: 4924 Join Date: Mar 2006 Location: Germany Device: STAReBOOK, iRex Iliad, Sony 505, Kindle 2	hmm...that is strange...I tried to convert the PDF to mobi in Calibre but it would not complete the job...maybe I'll give it another try sometime.