Errors while validating EPUB

crankypants · 12-04-2015, 09:59 AM

I have a fun project where I'm making EPUBs for to see how it's made. I'm using the validator at http://validator.idpf.org/ and am getting some errors which I don't know how to handle.

ERROR 1. "Mimetype file entry is missing or is not the first file in the archive."

The mimetype file is actually in the EPUB. Here's my EPUB listing from 7zip.

Code:

   Date      Time    Attr         Size   Compressed  Name
------------------- ----- ------------ ------------  ------------------------
2015-12-04 09:47:54 ....A          260          177  META-INF\container.xml
2015-12-04 09:24:16 ....A           20           20  mimetype
2015-12-04 09:47:54 ....A         1130          512  OEBPS\content.opf
2015-10-30 18:33:34 D....            0            0  OEBPS\images
2015-10-30 18:33:34 D....            0            0  OEBPS\styles
2015-12-04 09:32:10 ....A          943          522  OEBPS\styles\style01.css
2015-11-04 09:19:14 D....            0            0  OEBPS\Text
2015-12-04 09:47:54 ....A         1381          777  OEBPS\Text\00cover.xhtml
2015-12-04 09:47:54 ....A         9334         3678  OEBPS\Text\31book.xhtml
2015-10-30 18:33:34 D....            0            0  OEBPS\themes
2015-12-04 09:47:54 ....A         1675          611  OEBPS\toc.ncx
------------------- ----- ------------ ------------  ------------------------
                                 21449         7839  9 files, 4 folders

I have several more errors in OEBPS/content.opf

ERROR 2 on line 12 of content.opf: "Error while parsing file 'value of attribute "id" is invalid; must be an XML name without colons'."

Here's my whole content.opf file:

Code:

<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<package xmlns="http://www.idpf.org/2007/opf" unique-identifier="BookId" version="2.0">
  <metadata xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:opf="http://www.idpf.org/2007/opf">
    <dc:title>Book of the Farm, 1844, Vol 2</dc:title>
    <dc:creator opf:role="aut">Henry Stephens, FRSE</dc:creator>
    <dc:language>en</dc:language>
    <dc:date opf:event="modification">2015-12-04</dc:date>
    <dc:identifier id="bookid" opf:scheme="ISBN">bookofthefarm-vol1-1847</dc:identifier>  </metadata>

<manifest>
  <item href="toc.ncx" id="ncx" media-type="application/x-dtbncx+xml" />
  <item href="Text/00cover.xhtml" id="00cover.xhtml" media-type="application/xhtml+xml" /> # THIS IS LINE 12
  <item href="Text/31book.xhtml" id="31book.xhtml" media-type="application/xhtml+xml" />
  <item href="styles/style01.css" id="style01.css" media-type="text/css" />
</manifest>

<spine toc="ncx">
  <itemref idref="00cover.xhtml" /> # LINE 18
  <itemref idref="31book.xhtml" /> # ERROR ON LINE 19
</spine>

<guide>
  <reference href="Text/00cover.xhtml" title="Cover" type="cover" />
</guide>
</package>

Line 12 needs an XML scheme attribute? What are the values of the scheme attribute?

PeterT · 12-04-2015, 10:08 AM

The mimetype HAS to be the first file and HAS to be stored uncompressed.

eschwartz · 12-04-2015, 11:01 AM

How are you zipping up the book?

You have to do mumb-jumbo with the info-zip command-line tools (the spell is documented in various places on the web), or alternatively use special tools designed specifically to pack up an EPUB (heavily advised).

Otherwise you will certainly fail to do it properly.

The other error is, I believe, because an id cannot start with a number.
As you can tell, EpubCheck doesn't have good error messages.

Try FlightCrew, which is usually a lot more clear about your actual problem.

crankypants · 12-04-2015, 12:02 PM

Quote:

The mimetype HAS to be the first file and HAS to be stored uncompressed.

An EPUB is just a zipped file with certain files and directories in it. How in the world do I force the mimetype file to not be compressed and be the first file using 7zip?

Why would the creators of EPUB even ask for something bizarre like that? There should be a mimetype file in there, and the EPUB validator should find it regardless of what physical order it has in the zip/epub file. Can someone help me understand?

Quote:

Try FlightCrew, which is usually a lot more clear about your actual problem.

Is Flightcrew free?

DaleDe · 12-04-2015, 12:12 PM

Quote:

Originally Posted by crankypants

An EPUB is just a zipped file with certain files and directories in it. How in the world do I force the mimetype file to not be compressed and be the first file using 7zip?

Why would the creators of EPUB even ask for something bizarre like that? There should be a mimetype file in there, and the EPUB validator should find it regardless of what physical order it has in the zip/epub file. Can someone help me understand?

Is Flightcrew free?

Zip the mimetype first by itself uncompressed and then add the other items. The reason they wanted it that way is so the file could be identified as to filetype without having to unzip it using, for example, the Unix file command.

Flightcrew is a free program, part of Sigil.

Also try "BookId" with caps in the place.

Dale

Doitsu · 12-04-2015, 12:18 PM

Quote:

Originally Posted by crankypants

ERROR 2 on line 12 of content.opf: "Error while parsing file 'value of attribute "id" is invalid; must be an XML name without colons'."

Code:

  <item href="Text/31book.xhtml" id="31book.xhtml" media-type="application/xhtml+xml" />

Ids in epub2 books must start with a letter.

pdurrant · 12-04-2015, 12:20 PM

Quote:

Originally Posted by crankypants

An EPUB is just a zipped file with certain files and directories in it. How in the world do I force the mimetype file to not be compressed and be the first file using 7zip?

If you were on a Mac, you could use my ePub zip/unzip AppleScript.

If you have access to a command-line zip, the commands (with the unzipped ePub directory as current directory)

zip -X0 <quoted form of ePubFilePath> mimetype
zip -rDX9 <quoted form of ePubFilePath> * -x mimetype

Doitsu · 12-04-2015, 12:22 PM

If you have a Windows machine, you can use ePubPack.

eschwartz · 12-04-2015, 12:53 PM

Quote:

Originally Posted by crankypants

An EPUB is just a zipped file with certain files and directories in it. How in the world do I force the mimetype file to not be compressed and be the first file using 7zip?

Like I said, either use the magical incantation with zip.exe, as blessed by the IPDF, or special tools written specifically to "zip up an EPUB properly".

It is not enough to simply create a zipped folder with 7-Zip.

...

Didn't you mention somewhere you were using perl? Why not use the perl module for creating EPUB automatically as part of the scripting?

http://search.cpan.org/~oty/EBook-EPUB-0.6/

Quote:

Why would the creators of EPUB even ask for something bizarre like that? There should be a mimetype file in there, and the EPUB validator should find it regardless of what physical order it has in the zip/epub file. Can someone help me understand?

As DaleDe said, the mimetype is needed for the zipfile itself, not the unpacked representation.

Quote:

Is Flightcrew free?

It is released under the same umbrella as Sigil (these days, it is no longer a builtin but decoupled and turned into a plugin).

https://github.com/Sigil-Ebook/flightcrew

Binary releases as a Sigil plugin are available, but you should also be able to build it as a standalone gui or cli tool.

crankypants · 12-04-2015, 12:58 PM

It appears my root problem might be that 7-zip 15.16 cannot control compression, so the mimetype file gets compressed. Let me fix that first, I'm download info-zip right now and will try that.

crankypants · 12-04-2015, 01:00 PM

Quote:

Originally Posted by Doitsu

Code:

  <item href="Text/31book.xhtml" id="31book.xhtml" media-type="application/xhtml+xml" />

Ids in epub2 books must start with a letter.

All id attributes? Or just the ones for the xhtml files?

That's seriously archaic.

eschwartz · 12-04-2015, 01:06 PM

Quote:

Originally Posted by crankypants

All id attributes? Or just the ones for the xhtml files?

That's seriously archaic.

Go bring it up with the World Wide Web Consortium. They made the rule.

(Your chances of getting the HTML4 spec retroactively changed like that are zero to none.)

Doitsu · 12-04-2015, 01:24 PM

Quote:

Originally Posted by crankypants

All id attributes? Or just the ones for the xhtml files?

If ids that start with numbers are important to you, you'll have to generate an ePub3 book.

(Since ePub3 (mostly) supports HTML5 files, you can use ids that start with a number in HTML5 files, but nowhere else.)

DaleDe · 12-04-2015, 04:28 PM

Quote:

Originally Posted by crankypants

It appears my root problem might be that 7-zip 15.16 cannot control compression, so the mimetype file gets compressed. Let me fix that first, I'm download info-zip right now and will try that.

Info zip will be fine. I use it. I have 7zip for adding and viewing content.

Dale

DaleDe · 12-04-2015, 04:46 PM

Another very simple tool to work with ePub files is Tweak EPUB described in our wiki. This is a free program that will allow taking a ePub apart to edit the pieces. It will allow you to view the resultant ePub from inside the tool to view any changes you make prior to rebuilding the file.

Dale

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Validating Epub	Thom*	Editor	6	07-08-2015 04:17 AM
validating epub fails - how do I correct?	BeccaPrice	ePub	14	06-12-2013 02:01 PM
Validating epub returns warning	writerkit	Conversion	2	03-30-2013 02:35 AM
Issues validating EPUB for iPad	godsmac	Calibre	13	06-26-2011 09:02 PM

12-04-2015, 10:08 AM	#2
PeterT Grand Sorcerer Posts: 13,293 Karma: 78876004 Join Date: Nov 2007 Location: Toronto Device: Libra H2O, Libra Colour	The mimetype HAS to be the first file and HAS to be stored uncompressed.

12-04-2015, 11:01 AM	#3
eschwartz Ex-Helpdesk Junkie Posts: 19,421 Karma: 85400180 Join Date: Nov 2012 Location: The Beaten Path, USA, Roundworld, This Side of Infinity Device: Kindle Touch fw5.3.7 (Wifi only)	How are you zipping up the book? You have to do mumb-jumbo with the info-zip command-line tools (the spell is documented in various places on the web), or alternatively use special tools designed specifically to pack up an EPUB (heavily advised). Otherwise you will certainly fail to do it properly. The other error is, I believe, because an id cannot start with a number. As you can tell, EpubCheck doesn't have good error messages. Try FlightCrew, which is usually a lot more clear about your actual problem.

12-04-2015, 12:22 PM	#8
Doitsu Grand Sorcerer Posts: 5,680 Karma: 23983815 Join Date: Dec 2010 Device: Kindle PW2	If you have a Windows machine, you can use ePubPack.

12-04-2015, 12:58 PM	#10
crankypants Hmm. Posts: 124 Karma: 2016606 Join Date: Oct 2015 Device: Android 4.2 Google Play Reader	It appears my root problem might be that 7-zip 15.16 cannot control compression, so the mimetype file gets compressed. Let me fix that first, I'm download info-zip right now and will try that.

12-04-2015, 04:46 PM	#15
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	Another very simple tool to work with ePub files is Tweak EPUB described in our wiki. This is a free program that will allow taking a ePub apart to edit the pieces. It will allow you to view the resultant ePub from inside the tool to view any changes you make prior to rebuilding the file. Dale

Advert

Advert