View Full Version : Smashwords and ePub format question


yarndog
04-03-2011, 11:31 AM
I created an open office writer .doc of my business book and used the Smashwords style guide to format it, then submitted it. It was accepted and entered into the Smashwords Premium Catalog. However, I noticed my book never got shipped to Apple, even though I followed all the Apple guidelines. Upon further searching I discovered I should download the ePub version of my book from Smashwords and run it through an epub Validator, such as threepress.org. I have done so and fixed all the errors in the ePub version, however Smashwords only accepts .doc format. I do not see a straight forward way to fix listed ePub errors in my .doc document. What I currently do, is tell open office to save it in html format, then check the ePub errors, then go see what that translates to in html, then go find that spot in my .doc document, then I copy this document into Sigil, create a new ePub version, then run it through threepress.org. This seems very convoluted and inefficient. Is there a better way? What I am discovering is that the epub validator doesn't like my blockquote, I think, because by default a blockquote tag indents. My solution is to go into the quotations style on open office and modify the style so it doesn't indent. This just seems like pulling teeth. Is there a better way to do this? The Smashwords help and FAQ just say to download the epub from their website, run it through the validator, fix the errors, then upload the revised version. Very mysterious on how to translate an ePub error into a fix in my open office .doc document. Finally, when I upload a new version I end up going through the whole process of review and waiting and then approval, then wait to ship. My book has been shipped to everyone but Apple, but it seems now I have to start over. Is there a better way? Thanks for listening and thanks in advance for your answers.

yarndog
04-03-2011, 06:50 PM
I created an open office writer .doc of my business book and used the Smashwords style guide to format it, then submitted it. It was accepted and entered into the Smashwords Premium Catalog. However, I noticed my book never got shipped to Apple, even though I followed all the Apple guidelines. Upon further searching I discovered I should download the ePub version of my book from Smashwords and run it through an epub Validator, such as threepress.org. I have done so and fixed all the errors in the ePub version, however Smashwords only accepts .doc format. I do not see a straight forward way to fix listed ePub errors in my .doc document. What I currently do, is tell open office to save it in html format, then check the ePub errors, then go see what that translates to in html, then go find that spot in my .doc document, then I copy this document into Sigil, create a new ePub version, then run it through threepress.org. This seems very convoluted and inefficient. Is there a better way? What I am discovering is that the epub validator doesn't like my blockquote, I think, because by default a blockquote tag indents. My solution is to go into the quotations style on open office and modify the style so it doesn't indent. This just seems like pulling teeth. Is there a better way to do this? The Smashwords help and FAQ just say to download the epub from their website, run it through the validator, fix the errors, then upload the revised version. Very mysterious on how to translate an ePub error into a fix in my open office .doc document. Finally, when I upload a new version I end up going through the whole process of review and waiting and then approval, then wait to ship. My book has been shipped to everyone but Apple, but it seems now I have to start over. Is there a better way? Thanks for listening and thanks in advance for your answers.

I am thinking it just plain might be easier if I submit my epub version to Apple directly rather than going through Smashwords. It just seems so convoluted and I have spend a great deal of time trying to figure out how to make the corrections in my open office document to comply with the epub validator. I have already submitted an application to Apple for the iBookstore, I hope they accept me. Then I will just submit my epub file straight away and make any corrections straight away. No fuss, no muss. I would hope that Smashwords would give better guidance on how to fix ePub validator errors successfully in a .doc document.

Toxaris
04-04-2011, 05:48 AM
The validator errors usually have little to do with the doc, but with the conversion or creation. Blockquotes are totally allowed, but not within a paragraph.

What did the validation within Sigil have to say about the epub?

yarndog
04-04-2011, 08:04 AM
I make it through Sigil just fine, I make it through meatgrinder just fine. The problem is I get rejected by Apple because my final epub doesn't make it through the threepress.org epub checker. It doesn't like blockquotes so I Changed them to <div> tags. I am now guessing based on your feedback it is because I probably have the blockquotes inside <p> tags. Anyways that part is fixed. It also doesn't seem to like the attribute tags inside the <img> tag, meaning border=0 align=left, etc. Again, in the straight epub file I can fix those, in fact I downloaded the Smashwords created epub file, made the changes and ran it through threepress.org and it came out fine. My problem is how to change my doc document so it makes it though both smashwords meatgrinder(which it will), and the threepress.org epub check. I also seem to be getting a missing dc:title and dc:language metadata error, of which I don't know how to fix in the doc document either. Anyways, that is where I am at and thanks for the info.

yarndog
04-04-2011, 11:41 AM
I think I found another, somewhat convoluted, solution. I tried an experiment on a Microsoft Word 2007 document and it worked, so I am thinking the same thing with Open Office. I will save my Open Office document as an xml file, then using an XML editor go in and make the changes. Then I will open the modified xml file back in Open Office, then save that as an HTML file, then open in Sigil, where I will save it as an epub document. I do realize there is an epub extenstion for open office I could use, but I like sigil for a variety of reasons. Well, off to experiment. See if I write long enough and babble long enough and search long enough and then babble some more, the answers start flowing like a river.

yarndog
04-05-2011, 11:35 AM
This is proving to be a little more challenging than I originally thought. I may have to just stick with my original solution of just dealing with Apple directly so I can edit the epub and leaving Smashwords out. Well, one more try.

yarndog
04-05-2011, 09:09 PM
Well, I feel whipped. I switched over to LibreOffice. My document path goes something like LibreOffice>Doc>HTML>Sigil>ebup>threepress>errors, doesn't like border or align attributes listed with my image. Also missing dc:title and dc:language I tried saving as XML and finding those attributes to delete, but not successful as I was unable to find those attributes, and don't know how to control them from Libre. I was really hoping to get to Apple via Smashwords, but without a solution to my dilemma I will have to go directly to Apple as I can edit the epub directly. Bummer. Any help is appreciated! I find myself talking to myself :(

Toxaris
04-06-2011, 05:10 AM
When you create the epub in Sigil, you have to fill in the author, title and language. It is in the menu (meta editor). Those are required according to the specs and you have to fill that in manually of course.

yarndog
04-06-2011, 08:06 AM
The problem I have is that Smashwords only accepts .doc documents, meaning no epub. That is my struggle. They only thing I can come up with, based on my experimenting, is to open my document in Microsoft Word 2007, save it as a .docx document, then change the .docx extension to .zip, then unpack it, make the changes to the document.xml file in terms of picture names, border, and align attributes, just strip them out, then I have to find where the content.opf information is and add the dc:title, and dc:language metatags, then compress it back into a .zip, then change extension to .docx, then open in Microsoft Word 2007 and resave into .doc, and hope the changes took. Then upload to Smashwords. This is so convoluted. I did apply to Apple to be able to load books there directly. If they say yes then I am going that route because then i can just use epub straight away and that is much, much easier. Thanks for you info.

st_albert
04-06-2011, 04:25 PM
It seems to me that if Smashwords only accepts .doc format, and your .doc follows Smashwords' extensive style guide, and Smashwords conversion of you doc produces an invalid epub, then perhaps you should take this up with Smashwords' technical support.

Just a suggestion.

yarndog
04-06-2011, 10:48 PM
Unfortunately the issue isn't Smashwords and meatgrinder. I pass through Smashwords and my book is in the Premium Catalog and has been shipped to everyone but Apple. It gets run through another epub checker(threepress) and unless the Smashwords meatgrinder generated epub file passes threepress, I get rejected by Apple for their iBookstore. The only thing Smashwords says in their FAQ is to make sure my file passed threepress epub checker and if there are any to fix the errors and resubmit the .doc file. So I end up having to create my own epub file from my .doc file and find the errors, which are quite hard to fix in the .doc file. Thank you for suggesting I talk to Tech Support at Smashwords, for perhaps they have a better way. I can't be the only one with these struggles. In fact, I know I am not.

Toxaris
04-07-2011, 03:40 AM
Again, you cannot 'correct' the DOC for those kind of errors in the ePub. If the ePub is not valid, you either have to correct it yourself or let the one creating the ePub correct it.

yarndog
04-07-2011, 11:54 AM
That is what I am up against. I have seen posts of people who unzip a .docx file, make the fixes, then compress, then convert to epub to do a check. I have attempted to do that and only get so far.

Toxaris
04-07-2011, 03:36 PM
As soon as you convert, the docx (xml) will be converted to xhtml. This WILL result in loss of some items that are specific for the Word format. Again, you either need to change the epub (not docx) yourself or ask Smashword to convert better.

You did not respond to my earlier question. What does Flightcrew (validator of Sigil) say when you put the epub from Smashwords through it?

yarndog
04-07-2011, 04:21 PM
I will have to run the epub from smashwords through sigil. I ran my own epub that I made through both sigil and threepress and the errors are similar, but worded differently. For example on error in threepress is "align not allowed here" for an <img> statement, where flightcrew would say "align not inside element", etc.

Toxaris
04-08-2011, 05:38 AM
Flightcrew checks more than threepress and the error descriptions are usually much more usable and understandable.

It looks that the epub generator from smashword does not follow specs though.

cantovento
04-09-2011, 06:06 PM
My problem is similar to yarndog's. I get 52 times the same error message from epubcheck:

ERROR: van-oerknal-tot-oorbevolking.epub/tmp_6b25f4347228bfd9dd9d03c78e4d7602_Vez7s1.chapha ck.fixed.fc.tidied.stylehacked.xfixed_split_001.ht ml(1879): attribute "width" not allowed here; expected attribute "abbr", "align", "axis", "char", "charoff", "class", "colspan", "dir", "headers", "id", "rowspan", "scope", "style", "title", "valign" or "xml:lang"

My question: It should be easy to fix since it's 52x the same thing, but where do I find the 52 errors in the epub file?

Toxaris
04-10-2011, 06:42 AM
It looks like an error in images. Run the file through FlightCrew, it is much better than epubcheck, and then check the errors. It should be more clear what the problem is.

If the problem is really the same problem over and over, a well-phrased search and replace should fix it.

cantovento
04-10-2011, 02:50 PM
Sorry, I forgot to say that I wrote my book in WORD1997-2003. In accordance with smashwords instructions , I removed all formatting, then manually re-formatted the headings in 4 levels, then bookmarked about 200 headings and linked them to a TOC. After sending it through Meatgrinder, all formats were shipped except Amazon and Apple. When a website says "will be shipped, I believe it. Shipping to Apple is said to be on Thursdays, so I waited 4 Thursdays. Only after querying Smashwords did I learn that you have to validate the converted epub file through epubchecker.

The doc contains 27 images. I have reformatted them all and republished. When I'm through the Smashwords rigmarole I'll run the epub file through FlightCrew and report the result here.

Dank je wel!

cantovento
04-11-2011, 09:09 AM
After re-aligning all images & making a few other improvements, I uploaded my doc to Smashwords and passed the resulting epub thru Epubchecker & FlightCrew. The Epubchecker error msgs are still exactly the same, FlightCrew's are just as unintelligible to me, e.g:

This resource is reachable but not present in the OPF <manifest>

This resource is reachable but not present in the OPF <spine>

This resource is present in the OPF <manifest> but is not reachable (it's unused)

then it gets worse...
The msgs in FlightCrew will not copy & it's too much garble to transcribe manually.
Some have N/A under line, others have numbers.
How can I find which line is which?
Isn't there no software that will give understandable error msgs for plain people like me?

DaleDe
04-11-2011, 12:08 PM
It is saying you have an extra file or perhaps more than one in your ePUB document that should not be there. Change the extension of you ebook to .zip and then open it and take a look at the structure and see it there is an unusually named file. Sometimes Calibre can stick a bookmark file in the archive which will cause this error. Once you fix it then you can change the extension back to .epub. Note if you have 7zip then you can just open the archive without having to rename the extension.

Dale

yarndog
04-15-2011, 12:13 PM
I did hear from Mark Coker from Smashwords. It appears meatgrinder did not pick up some table tabs in my doc document. I fixed that problem. In addition when I quoted a person I used the Quotations style in LibreOffice. When this got converted to epub the Quotations style got converted to <blockquote> tags and meatgrinder didn't like them. My solution was to just change my quotes that used Quotation style to Content 5 style, and just tell that style to italicize. Once I did this, in addition to fixing the table problem, my book made it through meatgrinder and epub checker, and has been shipped to Apple. Yippee! Mr. Coker mentioned they are changing the way a user is notified of epub errors. They do understand the current system is too mysterious in terms of a user knowing if they have epub errors by having to check whether their book shipped/not shipped.

yarndog
04-15-2011, 12:16 PM
I did hear from Mark Coker from Smashwords. It appears meatgrinder did not pick up some table tabs in my doc document. I fixed that problem. In addition when I quoted a person I used the Quotations style in LibreOffice. When this got converted to epub the Quotations style got converted to <blockquote> tags and meatgrinder didn't like them. My solution was to just change my quotes that used Quotation style to Content 5 style, and just tell that style to italicize. Once I did this, in addition to fixing the table problem, my book made it through meatgrinder and epub checker, and has been shipped to Apple. Yippee! Mr. Coker mentioned they are changing the way a user is notified of epub errors. They do understand the current system is too mysterious in terms of a user knowing if they have epub errors by having to check whether their book shipped/not shipped. Just by fixing the blockquote problem got rid of most of the errors alone. Threepress cranks out so many errors for every occurrence of a problem.

yarndog
04-15-2011, 12:17 PM
I got an error for each occurrence of <blockquote>. I just used find and replace to replace each one of those <blockquotes> and the large morass of errors went away.

Toxaris
04-15-2011, 04:08 PM
<blockquote> is an allowed tag for epubs... It just should be inside <p> tags.

DrDln
09-19-2011, 11:11 PM
My problem is similar to yarndog's. I get 52 times the same error message from epubcheck:
ERROR: van-oerknal-tot-oorbevolking.epub/tmp_6b25f4347228bfd9dd9d03c78e4d7602_Vez7s1.chapha ck.fixed.fc.tidied.stylehacked.xfixed_split_001.ht ml(1879): attribute "width" not allowed here; expected attribute "abbr", "align", "axis", "char", "charoff", "class", "colspan", "dir", "headers", "id", "rowspan", "scope", "style", "title", "valign" or "xml:lang"
My question: It should be easy to fix since it's 52x the same thing, but where do I find the 52 errors in the epub file?

I seem to have similar problem. Please share if you find the solution. I will do the same. Here is my dilemma.

Nine of my 12 guides pass the epub test. One of them I fixed yesterday only. Below are the error messages that I get with other 3 guides. Any idea what do these mean!

I will continue to work on these. Most of these errors perhaps are link and image related. The major work is to locate them.

I will appreciate any clue on how to find these errors or to know what these mean.

1. <<ERROR: the-power-of-breathing.epub: The file tmp_56e5d5dbe37ad7496a77581a2e5789e3_VW6W9h_html_m 5b32798a.png does not appear to be of type image/png>>

2. <<ERROR: art-of-stress-free-living-eastern-and-western-approach.epub/tmp_90dd12c32a5b750b5645fd483b9830eb_aTzujH.ch.fix ed.fc.tidied.stylehacked.xfixed_split_011.html(38) : attribute "width" not allowed here; expected attribute "abbr", "align", "axis", "char", "charoff", "class", "colspan", "dir", "headers", "id", "rowspan", "scope", "style", "title", "valign" or "xml:lang" >>

3. << ERROR: HHL-epub.epub/tmp_6c6543aa0c591fd11631fa8ca3bcaedb_G67JJi.ch.fix ed.fc.tidied.stylehacked.xfixed_split_014.html(207 ): attribute "height" not allowed here; expected attribute "abbr", "align", "axis", "char", "charoff", "class", "colspan", "dir", "headers", "id", "rowspan", "scope", "style", "title", "valign" or "xml:lang"

Thanks in advance for your help.

LaoTseu
09-20-2011, 01:00 AM
<blockquote> is an allowed tag for epubs... It just should be inside <p> tags.

This is a very surprising statement to me. So far as I know, <blockquote> can be use as first child of <body> and do not need to be enclosed by <p>.

The following code validate with http://validator.w3.org/
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title>title</title>
</head>

<body>
<blockquote>
<p>test</p>
</blockquote>
</body>
</html>


Maybe you meant that when using strict HTML/XHTML, <blockquote> can only contain block-level elements like <p> so

<blockquote>
text 1
text 2
text 3
</blockquote>

is invalid for strict XHTML whereas

<blockquote>
<p>
text 1
text 2
text 3
</p>
</blockquote>

is valid.

Toxaris
09-20-2011, 02:41 AM
I have to check, but it looks that you are right. I wrote it down in the wrong order apparently...

DrDln
09-28-2011, 12:58 AM
Finally, all of my 12 guides pass ePub validation. But two of 12 are still pending approval for some reason.

DrDln
10-23-2011, 12:49 PM
I have all 12 books passing the ePub test. But I got a new problem with one of the books. I have done the suggestion in message below. But still have a problem. Any help will be highly appreciated. Here's the message that I get:

" #2 Your book contains some possibly corrupt formatting. If you take a look at the EPUB, you'll notice there are slight font size differentiations. This is caused by your mix of inconsistent on-screen formatting. One option is to do a CTRL-A and change everything to Normal style, then modify the style to enforce the characteristics you want. ||| When you're finished correcting your book, go to Dashboard: 'upload new version' to upload the new version. Thanks. "

Thanks in advance for your help.

fidelipub
01-04-2012, 11:38 AM
What I usually do is view the epub and see where the issues are, then go to those paragraphs and clear the formatting. This usually fixes the problems and that way you don't have to employ the nuclear method.

theducks
01-04-2012, 11:57 AM
What I usually do is view the epub and see where the issues are, then go to those paragraphs and clear the formatting. This usually fixes the problems and that way you don't have to employ the nuclear method.

I second this (for minor differences).
Use Tweezers not a Clam-shell crane to pick out bits