Warning Attribute 'lang' is not declared, docx to epub, with FlightCrew

jkeinonen · 09-27-2016, 07:01 AM

Hi!

This is my first post here. I read the stickies, don't think I need to attach files, searched but didn't find a similar case.

My problem is that when converting with Calibre (2.68 64-bit) from Word 2016 docx (Windows 7 64-bit) to epub, I get warnings on FlightCrew 0.72 of the kind "attribute 'lang' is not declared for element html (or i, or p)".

The warning does not affect the usability of the epub. However, the two biggest e-bookstores in the Finnish market, Elisa Kirja and iTunes, seem to have a similar validation system, so the warnings have to be fixed. I have been able to do that manually.

Here is a what the second line of index_split_000.html reads like:

<html xmlns="http://www.w3.org/1999/xhtml" lang="fi" xml:lang="fi">

Not understanding code very much, I have by trial and error learned that, in Calibre Edit mode, removing the part lang="fi" will make the warning go away.

Using the Find/Replace tool, in All text files mode, it has been quite easy to remove all the unwished lang="fi". But the recent conversion job had quotes in many other languages, leading to lang="de" etc. causing extra work and human error risk.

So, my questions and notes are:

1) Is there a switch or button in Calibre that I have missed, which could fix the mini-problem?

2) The "blame" probably lies at Microsoft

as always, but if someone knows if there is a setting in Word 2016 that "kills" the wrong type of language marking, I'm happy to learn that instead.

3) Regarding the new multi-language issue vs. just one language (Finnish): I probably should, in Word 2016, select Ctrl-A and force everything to one language. So that's another workaround.

Finally, a huge thanks to Kovid Goyal and everyone helping with developing Calibre. I made a donation when I started using it, and will make another one if this issue proceeds.

kovidgoyal · 09-27-2016, 07:20 AM

Just use the search and replace settings in the conversion dialog, no need to involve the editor at all.

Code:

Search:<([^>+]) lang=\S+([^>]*)>
Replace: <\1 \2>

Should do the trick.

jkeinonen · 09-27-2016, 08:17 AM

Thank you Kovid. I'm really afraid of using anything that looks like code, because I am totally amateur in that area, but I will have a deep look at this when I can concentrate!

09-27-2016, 07:01 AM	#1
jkeinonen Junior Member Posts: 7 Karma: 10 Join Date: Aug 2016 Device: Kindle, Kobo, tablet	Warning Attribute 'lang' is not declared, docx to epub, with FlightCrew Hi! This is my first post here. I read the stickies, don't think I need to attach files, searched but didn't find a similar case. My problem is that when converting with Calibre (2.68 64-bit) from Word 2016 docx (Windows 7 64-bit) to epub, I get warnings on FlightCrew 0.72 of the kind "attribute 'lang' is not declared for element html (or i, or p)". The warning does not affect the usability of the epub. However, the two biggest e-bookstores in the Finnish market, Elisa Kirja and iTunes, seem to have a similar validation system, so the warnings have to be fixed. I have been able to do that manually. Here is a what the second line of index_split_000.html reads like: <html xmlns="http://www.w3.org/1999/xhtml" lang="fi" xml:lang="fi"> Not understanding code very much, I have by trial and error learned that, in Calibre Edit mode, removing the part lang="fi" will make the warning go away. Using the Find/Replace tool, in All text files mode, it has been quite easy to remove all the unwished lang="fi". But the recent conversion job had quotes in many other languages, leading to lang="de" etc. causing extra work and human error risk. So, my questions and notes are: 1) Is there a switch or button in Calibre that I have missed, which could fix the mini-problem? 2) The "blame" probably lies at Microsoft as always, but if someone knows if there is a setting in Word 2016 that "kills" the wrong type of language marking, I'm happy to learn that instead. 3) Regarding the new multi-language issue vs. just one language (Finnish): I probably should, in Word 2016, select Ctrl-A and force everything to one language. So that's another workaround. Finally, a huge thanks to Kovid Goyal and everyone helping with developing Calibre. I made a donation when I started using it, and will make another one if this issue proceeds.

09-27-2016, 07:20 AM	#2
kovidgoyal creator of calibre Posts: 43,858 Karma: 22666666 Join Date: Oct 2006 Location: Mumbai, India Device: Various	Just use the search and replace settings in the conversion dialog, no need to involve the editor at all. Code: Search:<([^>+]) lang=\S+([^>]*)> Replace: <\1 \2> Should do the trick.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
ePub validation error help "attribute border is not declared for element img	bmacklin	ePub	12	11-06-2015 11:00 AM
Attribute 'target' is not declared for element 'a'	Alda	Sigil	5	01-30-2015 08:49 AM
attribute 'name' is not declared for element 'a'	lballas	Sigil	2	01-09-2014 06:03 PM
Help! ePub errors -- attribute link/text not declared	katyya	ePub	1	11-17-2011 02:13 PM
Validation: attribute 'start' is not declared for element 'ol'	WS64	Sigil	4	10-06-2011 03:15 AM

09-27-2016, 08:17 AM	#3
jkeinonen Junior Member Posts: 7 Karma: 10 Join Date: Aug 2016 Device: Kindle, Kobo, tablet	Thank you Kovid. I'm really afraid of using anything that looks like code, because I am totally amateur in that area, but I will have a deep look at this when I can concentrate!

Advert