epub validation error

[drwinn]

ePub structure:

mimetype
metadata.opf
META-INF (folder: container.xml)
OEBPS (folder: html and toc.ncx )


In Flightcrew error is:

name of file/META-INF/container.xml
line 1
invalid document structure

This is the structure in the container.xml file:

<?xml version="1.0"?>
<container version="1.0" xmlns="urn:oasis:names:tc:opendocument:xmlns:conta iner">
<rootfiles>
<rootfile full-path="/metadata.opf" media-type="application/oebps-package+xml"/>
</rootfiles>
</container>

at http://threepress.org/document/epub-validate/
The error message reads:

ERROR: TheVRabbit_44.epub: I/O error: invalid CEN header (bad compression method)

What to do?

Winn

[Jellby]

The epubcheck message suggests that the file is not properly zipped, and this might be confusing flightcrew... How are you creating the epub? Are you sure you are compressing it as a ZIP and not some variant (like 7-zip or whatever), with the "deflate" method?

Otherwise, the full-path="/metadata.opf" looks wrong, as it points to an absolute path (it starts with a slash). It might work, but don't you mean full-path="metadata.opf"?

[drwinn]

Working in Windows 7 64 bit

I removed the / but with the same results.

I am using WinZip 14.5 First dragging the mimetype across then the other file and folders. mimetype is listed first and from what I can see it is uncompressed. The other files have various compression.

Change zipx to epub. Run validation.

[DMSmillie]

Originally Posted by drwinn

... I am using WinZip 14.5 ... Change zipx to epub ...

I think that might be your problem - you're zipping the files using WinZip 14's default "ZipX" compression, which is relatively new and may not yet be fully supported by other utilities. I'd suggest zipping the files using the "Legacy Zip 2.0" setting instead, and see if that solves the problem.

[drwinn]

Thanks,

What a difference an "x" makes.

What a joy to see:

"No errors or warnings detected"

Winn

[drwinn]

Okay,

The test above was to provide a template.

But, now with a "real" project. I am running into a different set of problems.

When I validate, I get the following errors and don't know exactly what to look for:

ERROR: miller_cross_purposes.epub/toc.ncx (line 78):

The value of attribute "src" associated with an element type "null" must not contain the '<' character.

This is what is on line 77/78 in the toc.ncx
<content src="Cross_Purposes_NOOK_final_32011.html/>
</navPoint>

ERROR: miller_cross_purposes.epub/toc.ncx: The value of attribute "src" associated with an element type "null" must not contain the '<' character.

ERROR: miller_cross_purposes.epub/toc.ncx (line 78):

The value of attribute "src" associated with an element type "null" must not contain the '<' character.

ERROR: miller_cross_purposes.epub/toc.ncx: The value of attribute "src" associated with an element type "null" must not contain the '<' character.

ERROR: miller_cross_purposes.epub/Cross_Purposes_NOOK_final_32011.html: Malformed byte sequence: Invalid byte 1 of 1-byte UTF-8 sequence. Check encoding

what does "check encoding" mean in the last one. I'm a newbie to ePub stuff.

Thanks for any help.

Winn

[DMSmillie]

Glad to see you got the first error(s) sorted out successfully.

Originally Posted by drwinn

ERROR: miller_cross_purposes.epub/toc.ncx (line 78):

The value of attribute "src" associated with an element type "null" must not contain the '<' character.

This is what is on line 77/78 in the toc.ncx
<content src="Cross_Purposes_NOOK_final_32011.html/>
</navPoint>

OK... this is a simple typo error - there's a missing double quote character after "32011.html". At the moment, the parser is reading the closing "/>" and then the opening "<" from "</navPoint>" as part of the "src" value. So it should read:

<content src="Cross_Purposes_NOOK_final_32011.html"/>
</navPoint>

Originally Posted by drwinn

ERROR: miller_cross_purposes.epub/Cross_Purposes_NOOK_final_32011.html: Malformed byte sequence: Invalid byte 1 of 1-byte UTF-8 sequence. Check encoding

what does "check encoding" mean in the last one. I'm a newbie to ePub stuff.

It's not specifically an EPUB thing. It's referring to the character encoding in the HTML file.

It looks like the file contains a declaration that it's UTF-8, either in an XML declaration at the start of the file:

<?xml version="1.0" encoding="UTF-8"?>

or in a META tag:

<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />

- but whatever text/HTML editor last saved the file was set to use a different character encoding (most likely "Windows-1252" or "ISO-8859-1"). You need to check the settings in your text/HTML editor to ensure it's set to save the file using "UTF-8" character encoding. The help files in whatever editor you're using should hopefully provide some further info on this if you search for "character encoding".

[drwinn]

Here I am again. Running Flightcrew. Getting the following messages:

File: miller_cross_purposes.epub/Cross_Purposes_NOOK_final_32011.html/
Line: N/A
Message: This resource is reachable but not present in the OPF <manifest>.

From Manafest file:

<manifest>
<item id="ncx" href="toc.ncx" media-type="application/x-dtbncx+xml"/>
<item id="millerhtml" href="Cross_Purposes_NOOK_final_32011.html" media-type="application/xhtml+xml"/>
</manifest>
<spine toc="ncx">
<itemref idref="millerhtml"/>
</spine>

File: miller_cross_purposes.epub/toc.ncx
Line 78
Message: This <content> element's "src" attribute value is "Cross_Purposes_NOOK_final_32011.html/", but that file does not exist.

Line 78 in toc.ncx: <content src="Cross_Purposes_NOOK_final_32011.html/">

File: miller_cross_purposes.epub/toc.ncx
Line 79
no character data is allowed by content model

Line 79 in toc.ncx: </navPoint>

Complete cluster from toc.ncx:
<navPoint id="navpoint-11" playOrder="11">
<navLabel>
<text>Bibliography</text>
</navLabel>
<content src="Cross_Purposes_NOOK_final_32011.html/">
</navPoint>

File: miller_cross_purposes.epub/toc.ncx
Line 79
Message: expected end of tag 'content'

Line 79 is: </navPoint>

File: miller_cross_purposes.epub/Cross_Purposes_NOOK_final_32011.html
Line: 170
Message: entity 'nbsp' not found

Line 170: &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n bsp;&nbsp;&nbsp;&nbsp;And to the earth be peace.<br />

Any help would be appreciated.

[DMSmillie]

Hi Winn

I think all of these errors except the NBSP one are down to a single typo in the toc.ncx file.

Originally Posted by drwinn

Line 78 in toc.ncx: <content src="Cross_Purposes_NOOK_final_32011.html/">

You've got the " and / characters the wrong way round at the end of that line. It should be:

Line 78 in toc.ncx: <content src="Cross_Purposes_NOOK_final_32011.html"/>

Originally Posted by drwinn

File: miller_cross_purposes.epub/Cross_Purposes_NOOK_final_32011.html/
Line: N/A
Message: This resource is reachable but not present in the OPF <manifest>.

Note the / at the end of the file name - result of typo above, I think.

Originally Posted by drwinn

File: miller_cross_purposes.epub/toc.ncx
Line 78
Message: This <content> element's "src" attribute value is "Cross_Purposes_NOOK_final_32011.html/", but that file does not exist.

Again, the / at the end of the file name.

Originally Posted by drwinn

File: miller_cross_purposes.epub/toc.ncx
Line 79
no character data is allowed by content model

This is caused by the / being included in the <content> element's "src" value, and so the parser not recognising that the <content> tag has ended, and reading "</navPoint>" as being included in the <content> element.

Originally Posted by drwinn

File: miller_cross_purposes.epub/toc.ncx
Line 79
Message: expected end of tag 'content'

Same issue here.

Originally Posted by drwinn

File: miller_cross_purposes.epub/Cross_Purposes_NOOK_final_32011.html
Line: 170
Message: entity 'nbsp' not found

Line 170: &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&n bsp;&nbsp;&nbsp;&nbsp;And to the earth be peace.<br />

This one... I'm not sure, but do you see the space between "N" and "BSP" in the 9th "&nbsp;"? I think perhaps there's a non-standard character lurking there. That space shows in your original post, and is showing when I preview this post, but isn't showing in the raw quoted text that I'm seeing here as I'm typing this. So not sure if it's some sort of artifact when the forum software displays the post, or an indication of something actually in the code. Perhaps deleting and retyping that string of "&nbsp;"s will get rid of this error?

[drwinn]

It seems link I fix one thing and a bunch more pops up.

The problem is that I don't exactly know what I'm looking for. Is there a list of epub errors laying around somewhere?

Here is a sample of the latest:

no character data is allowed by content model notes line with the following

<blockquote>
  The word <span class="italics">atonement</span>, which is almost the only theological term of English origin, has a curious history. The verb "atone," from the adverbial phrase "at one" (M.E. <span class="italics">at oon</span>), at first meant to reconcile, or make "at one;" from this it came to denote the action by which such reconciliation was effected.<a id="ftn5" href="#bkftn5">[5]</a>
  </blockquote>

element 'span' is not allowed for content model '(p|h1|h2|h3|h4|h5|h6|div|ul|ol|dl|pre|hr|blockquo te|address|fieldset|table|switch|form|noscript|ins |del|script)'

Same as above, just one line below

element 'a' is not allowed for content model '(p|h1|h2|h3|h4|h5|h6|div|ul|ol|dl|pre|hr|blockquo te|address|fieldset|table|switch|form|noscript|ins |del|script)'

</blockquote> is on the line denoted.

What should I be looking for?

The file opens in Adobe Digital Editions and the links seem to be working including the TOC internal page and the left margin of ADE.

I am preparing this file for B&N Nook. Is the only acceptable outcome: no errors. Does the list mean that it will work on one reader and not others?

I am using Math symbols for Greek Characters in the book like α for the Greek alpha. I can see them in ADE and in the PC Nook reader. Will that cause any kind of a problem?

Thanks again?

[Jellby]

Originally Posted by drwinn

<blockquote>
  The word <span class="italics">atonement</span>, which is almost the only theological term of English origin, has a curious history. The verb "atone," from the adverbial phrase "at one" (M.E. <span class="italics">at oon</span>), at first meant to reconcile, or make "at one;" from this it came to denote the action by which such reconciliation was effected.<a id="ftn5" href="#bkftn5">[5]</a>
  </blockquote>

element 'span' is not allowed for content model '(p|h1|h2|h3|h4|h5|h6|div|ul|ol|dl|pre|hr|blockquo te|address|fieldset|table|switch|form|noscript|ins |del|script)'

The <blockquote> element must contain block-level content, so try with:

<blockquote>
  <p>The word <span class="italics">atonement</span>, [...]
effected.<a id="ftn5" href="#bkftn5">[5]</a></p>
  </blockquote>

[drwinn]

Thanks Jellby and Donna for your help.

I am down to one set of errors:

I am using internal links for an internal TOC with

<p class="left"><a href="#foreword">Foreword</a></p>
to
<a id="foreword"></a>

The ERROR reference below is to the line (line 119) with the <a id="foreword"></a>. It references this each occurrence through the document with different IDs i.e., <a id="c1"></a> or <a id="c2"></a>

ERROR: miller_cross_purposes.epub/Cross_Purposes_NOOK_final_32011.html (line 119): element "a" from namespace "http://www.w3.org/1999/xhtml" not allowed in this context

These ERROR messages are generated by epubchecker.

FlightCrew is only unhappy with one line in the HTML file:

element 'a' is not allowed for content model '(p|h1|h2|h3|h4|h5|h6|div|ul|ol|dl|pre|hr|blockquo te|address|fieldset|table|switch|form|noscript|ins |del|script)'

This is assigned to the next to last line of the HTML file which reads:

</body>

I read somewhere that FlightCrew was better at this stuff than epubchecker. The first being in English and the latter one in Codelish. While I can read the English, I am still stumped by what it means.

What to do?

Thanks again.

[pdurrant]

Originally Posted by drwinn

element 'a' is not allowed for content model '(p|h1|h2|h3|h4|h5|h6|div|ul|ol|dl|pre|hr|blockquo te|address|fieldset|table|switch|form|noscript|ins |del|script)'

move the <a....></a> outside the <p>....</p> or other block level tags.

[drwinn]

Thanks pdurrant,

move the <a....></a> outside the <p>....</p> or other block level tags.

but that makes no sense to me.

The following is one example from the 86 footnotes in this piece with reciprocal links.

<p class="left"><a id="bkftn86" href="#ftn86">[86]</a>. Lee, Johnny, <span class="italics">Lookin' for Love</span>, Music, 1989, Q Records, Atlanta.</p>

should be:

<a id="bkftn86" href="#ftn86">[86]</a>. <p class="left">Lee, Johnny, <span class="italics">Lookin' for Love</span>, Music, 1989, Q Records, Atlanta.</p>

Moving the "a" link outside of the "p" makes the display of the footnote number on one line and the footnote reference on another line.

What am I missing?

Could you give me an example of what is meant?

[Jellby]

Originally Posted by pdurrant

move the <a....></a> outside the <p>....</p> or other block level tags.

You mean inside?

Just like <blockquote>, the main <body> cannot contain in-line, such as <a>, <span> or <i>. You have to put your <a> inside a <p> or a <div>.

But you don't need an empty anchor at all. If all you want is to put an anchor somewhere, you can add the id attribute to any existing element:

<p id="foreword">This is where the Foreword starts.</p>

<p>This <a href="#foreword">link</a> points to the previous paragraph.</p>