The id/nam issue

paulpeer · 01-15-2010, 06:29 PM

Sigil (or Tidy) will change <a href="#note01" id="to_note01"> to <a href="#note01" id="to_note01" name="to_note01">. However, when parsing such a file in Validator (http://threepress.org/document/epub-validate/), you get the message:
ERROR: attribute "name" not allowed at this point.

The developers of Tidy are aware of this bug (see issue 226), but I wonder what we can do in the mean time. I can delete all the wrong attributes in a few minutes using wildcards, but when I save the file and open it again, all the wrong name-attributes are there again.

Does anyone have a good trick? Or do you just leave the name attributes?

Valloric · 01-15-2010, 07:12 PM

Quote:

Originally Posted by paulpeer

The developers of Tidy are aware of this bug (see issue 226)

Ha! Look at the VCS logs, no one really works on Tidy anymore. Official development is pretty much dead.

I'll fix it myself when I find the time. Probably for 0.2.1. Until then, don't worry about it. AFAIK "name" is merely deprecated in XHTML, not disallowed. Baaah, I checked the wrong version of the XHTML spec. In 1.1, it is officially removed for "a" elements, not just deprecated.

Still, don't worry about it. No renderer is ever going to flake out on that. But like I said, I'll fix it when I find the time.

netseeker · 01-15-2010, 07:44 PM

The name attribute for <a>-elements is forbidden in XHTML 1.1 if doctype "strict" is used. When using XHTML 1.0 (all doctypes) or 1.1 with another doctype than "strict" (eg. "transitional") then it's just deprecated. Anyway, the OPS-2.0 spec relies on XHTML 1.1 and if doctype "strict" is used then the name attribute is disallowed.

sjkramer · 02-24-2010, 04:23 PM

I've just run into this issue as well. I'd like not to worry about it, but my distributor (one of the largest ones in the US) will not allow us to upload files with the name attribute present. So if it's possible to make it a slightly higher priority fix, that would be great. Until it is, I won't be able to use Sigil (or at least will have eventually to open my file in another editor before uploading and make the fixes manually). Thanks!

Valloric · 02-24-2010, 07:04 PM

Quote:

Originally Posted by sjkramer

I've just run into this issue as well. I'd like not to worry about it, but my distributor (one of the largest ones in the US) will not allow us to upload files with the name attribute present. So if it's possible to make it a slightly higher priority fix, that would be great. Until it is, I won't be able to use Sigil (or at least will have eventually to open my file in another editor before uploading and make the fixes manually). Thanks!

It's already a high priority issue on the tracker, one of three such issues, out of 80+ currently opened.

I planned to take a look at it for 0.2.1, but if it's impacting you this much (i.e., not able to work), I'll try to take a look at it for beta2 of 0.2.0 (beta1 will be out in a few days).

Valloric · 02-24-2010, 08:03 PM

I felt bad about you not being able to work until this issue was fixed so I took a look at it now. It turned out to be remarkably simple to fix: Tidy (wrongly) specifies that the "name" attribute is present in the XHTML 1.1. spec, and has a special function for "fixing" (actually breaking) anchors.

So while I was analyzing this function for a not-too-intrusive way to turn it off, I noticed Tidy has an option for turning on/off this "name fixing" behavior. The option is switched on by default, but Tidy usually changes options according to the required doctype. So this should have been turned off internally, but it wasn't.

Either way, I now explicitly turned it off and it all works. No more name attribute.

You can build Sigil from source and get the fix right now if you need it ASAP, or you can wait for the first beta of 0.2.0.

sjkramer · 02-24-2010, 08:05 PM

You're really the best. And I know that you have tons and tons of issues to fix. I just thought I'd put another vote in for this one—and also that it's an issue that effects not just me, but it seems to be the way things are going. I really really appreciate it and all of your hard work—especially in the face of whiners like me!

01-15-2010, 06:29 PM	#1
paulpeer Zealot Posts: 147 Karma: 56 Join Date: Dec 2009 Location: Antwerpen Device: iPhone, Sony PRS-505, EPUBreader	The id/nam issue Sigil (or Tidy) will change <a href="#note01" id="to_note01"> to <a href="#note01" id="to_note01" name="to_note01">. However, when parsing such a file in Validator (http://threepress.org/document/epub-validate/), you get the message: ERROR: attribute "name" not allowed at this point. The developers of Tidy are aware of this bug (see issue 226), but I wonder what we can do in the mean time. I can delete all the wrong attributes in a few minutes using wildcards, but when I save the file and open it again, all the wrong name-attributes are there again. Does anyone have a good trick? Or do you just leave the name attributes?

02-24-2010, 04:23 PM	#4
sjkramer Connoisseur Posts: 68 Karma: 526028 Join Date: Nov 2009 Location: New York, NY Device: iphone	I've just run into this issue as well. I'd like not to worry about it, but my distributor (one of the largest ones in the US) will not allow us to upload files with the name attribute present. So if it's possible to make it a slightly higher priority fix, that would be great. Until it is, I won't be able to use Sigil (or at least will have eventually to open my file in another editor before uploading and make the fixes manually). Thanks! Last edited by sjkramer; 02-24-2010 at 04:24 PM. Reason: forgot to mention...

02-24-2010, 08:03 PM	#6
Valloric Created Sigil, FlightCrew Posts: 1,982 Karma: 350515 Join Date: Feb 2008 Device: Kobo Clara HD	I felt bad about you not being able to work until this issue was fixed so I took a look at it now. It turned out to be remarkably simple to fix: Tidy (wrongly) specifies that the "name" attribute is present in the XHTML 1.1. spec, and has a special function for "fixing" (actually breaking) anchors. So while I was analyzing this function for a not-too-intrusive way to turn it off, I noticed Tidy has an option for turning on/off this "name fixing" behavior. The option is switched on by default, but Tidy usually changes options according to the required doctype. So this should have been turned off internally, but it wasn't. Either way, I now explicitly turned it off and it all works. No more name attribute. You can build Sigil from source and get the fix right now if you need it ASAP, or you can wait for the first beta of 0.2.0. Last edited by Valloric; 02-24-2010 at 08:05 PM.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Battery issue - please help!	Akuli	Astak EZReader	7	05-21-2010 02:21 PM
Good afternoon from Saigon, Viet Nam	bindzus	Introduce Yourself	9	03-14-2010 10:28 PM
FB2 Issue	Jaie	Other formats	4	04-26-2009 05:16 AM
0.5.2 Issue	drmathprog	Calibre	3	03-25-2009 10:33 PM
Next issue.....	DeathtoToasters	iRex	5	12-05-2007 02:10 PM

01-15-2010, 07:44 PM	#3
netseeker sleepless reader Posts: 4,763 Karma: 615547 Join Date: Jan 2008 Location: Germany, near Stuttgart Device: Sony PRS-505, PB 360° & 302, nook wi-fi, Kindle 3	The name attribute for <a>-elements is forbidden in XHTML 1.1 if doctype "strict" is used. When using XHTML 1.0 (all doctypes) or 1.1 with another doctype than "strict" (eg. "transitional") then it's just deprecated. Anyway, the OPS-2.0 spec relies on XHTML 1.1 and if doctype "strict" is used then the name attribute is disallowed.

02-24-2010, 08:05 PM	#7
sjkramer Connoisseur Posts: 68 Karma: 526028 Join Date: Nov 2009 Location: New York, NY Device: iphone	You're really the best. And I know that you have tons and tons of issues to fix. I just thought I'd put another vote in for this one—and also that it's an issue that effects not just me, but it seems to be the way things are going. I really really appreciate it and all of your hard work—especially in the face of whiners like me!

Advert

Advert