Sigil doesn't like named entities

Notjohn · 07-13-2013, 09:27 AM

I am revising a book published a few years ago when I still trusted Word's export to html, but when I open it in Sigil, I get an error at the first named entity, which was an – and when I fixed that, it choked at the very next one, with many more waiting down the line.

Since Sigil has never before complained about my use of named entities (I prefer them to numerical entities because I can read them as I go), I'm guessing that this is because Microsoft has its own peculiar way of rendering html entities, which show on my text editor as ^R, ^V, and so forth.

Is there an easy fix for this, such as by telling Sigil there are two entity modes in use?

Mixing Word export and named entities has never caused me a problem in uploading html to Amazon or Barnes & Noble for conversion. Indeed, the very html file I have opened is already on sale as a Kindle and an epub book.

theducks · 07-13-2013, 10:46 AM

I use them all the time in Sigil
(recently used &bul; ° 10 ° &bul; as a chapter # decoration)

Something else is wrong. codepage?

Notjohn · 07-13-2013, 11:42 AM

Perhaps Sigil wants to see two codepages, one for Word-generated html, the other for named entities?

I ran the Word doc through word2cleanhtml.com online, and now all is fine, with the minor cavil that all the entities (Word's and mine) have become numerical entities. In any event, Sigil has no problem opening the book.

theducks · 07-13-2013, 11:48 AM

NJ did you think to Validate one of the failing files before feeding it to Sigil?
That might give a clue.

DiapDealer · 07-13-2013, 12:50 PM

Gotta be the declaration.
Named entities work fine, for the most part, but not unless the doctype and all the namespace declarations allow for them.

Hitch · 07-13-2013, 03:41 PM

Quote:

Originally Posted by DiapDealer

Gotta be the declaration.
Named entities work fine, for the most part, but not unless the doctype and all the namespace declarations allow for them.

Agreed. We used named entities all the time, it's our inhouse rule. Sigil may/mayn't change them, but going in, they're all named. Never had a problem except when I mistype them (which is, blush, often). ;-)

Hitch

theducks · 07-13-2013, 03:47 PM

Quote:

Originally Posted by Hitch

Never had a problem except when I mistype them (which is, blush, often). ;-)

Hitch

Welcome to that club

Which makes me wonder ? WHY ? EPUB3 supposedly does not allow named entities.
Try and remember a frickn' number

They call them (named entities) mnemonics for a reason

Notjohn · 07-13-2013, 05:08 PM

I wish I had a mnemonic so I would remember where that word came from!

Yes, "declaration" must be the term I was looking for. I just use a generic <!DOCTYPE HTML>. Odd that Microsoft's ^R would pass muster, and the named entity not, when both are in the same file. I'd be less surprised if it were the other way around.

Speaking of declarations, does MathML work in any e-book format?

Hitch · 07-13-2013, 05:45 PM

Quote:

Originally Posted by Notjohn

I wish I had a mnemonic so I would remember where that word came from!

Yes, "declaration" must be the term I was looking for. I just use a generic <!DOCTYPE HTML>. Odd that Microsoft's ^R would pass muster, and the named entity not, when both are in the same file. I'd be less surprised if it were the other way around.

Speaking of declarations, does MathML work in any e-book format?

Well, it meant mindful, back when it was "mnēmōn" in Greek, if that helps.

With regard to your second:

Even the big fancy iBooks volume on Algebra uses all images of the equations and forumlae. Maybe ePUB3 does, but I admit: I knoweth not if it does.

Hitch

Jellby · 07-14-2013, 03:08 AM

If I remember correctly, the DOCTYPE is not needed in ePub. But if it's present it must be correct (which implies you cannot have SVG code, for instance). It may be that Sigil wants it, but according to the spec, files without a DOCTYPE should be XHTML anyway (which normal named entities).

The above is valid only for HTML-like files, of course, you cannot use named entities in XML files, such as the OPF or NCX, you must use numerical entities or the actual characters there. This means that you'd better not use named entities in titles, I guess (maybe Sigil could automatically convert entities to characters when creating the TOC, if it doesn't already).

ventdefolie · 07-14-2013, 10:35 AM

Quote:

Originally Posted by Notjohn

I am revising a book published a few years ago when I still trusted Word's export to html, but when I open it in Sigil, I get an error at the first named entity, which was an – and when I fixed that, it choked at the very next one, with many more waiting down the line.

I had the same problem and solved it by inserting the doctype in the Xhtml files:

Code:

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">

i.e Search : <?xml version="1.0" encoding="UTF-8" standalone="no" ?>
Replace : <?xml version="1.0" encoding="UTF-8" standalone="no" ?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">

After that Sigil doesn't complain any more !

Notjohn · 07-14-2013, 10:36 AM

Quote:

Originally Posted by Hitch

Well, it meant mindful, back when it was "mnēmōn" in Greek, if that helps.

I have always regretted that I didn't take at least enough Greek to be able to read the letters and pronounce the words!

I can handle some Cyrillic because I once noticed that RESTAURANT was spelled PECTOPAH.

DaleDe · 07-14-2013, 10:47 AM

Quote:

Originally Posted by Notjohn

I wish I had a mnemonic so I would remember where that word came from!

Yes, "declaration" must be the term I was looking for. I just use a generic <!DOCTYPE HTML>. Odd that Microsoft's ^R would pass muster, and the named entity not, when both are in the same file. I'd be less surprised if it were the other way around.

Speaking of declarations, does MathML work in any e-book format?

MathML is supported in ePub 3. See the wiki for more on this.

st_albert · 07-14-2013, 11:12 PM

Quote:

Originally Posted by Notjohn

I have always regretted that I didn't take at least enough Greek to be able to read the letters and pronounce the words!

I can handle some Cyrillic because I once noticed that RESTAURANT was spelled PECTOPAH.

Heh. Having studied English, Spanish, Russian, Latin, and German, and having lightly perused a Greek language textbook, my opinion is: you can learn the Greek alphabet ( and you should. If you can read cyrillic, you're half way there.) but NEVER regret not learning the (classical) Greek language. Had I tried it, my brain would probably have melted. My "classical" education falters at that point.

No flames please; this is just MHO. ::grin, duck, run::

Albert

07-13-2013, 09:27 AM	#1
Notjohn mostly an observer Posts: 1,515 Karma: 987654 Join Date: Dec 2012 Device: Kindle	Sigil doesn't like named entities I am revising a book published a few years ago when I still trusted Word's export to html, but when I open it in Sigil, I get an error at the first named entity, which was an – and when I fixed that, it choked at the very next one, with many more waiting down the line. Since Sigil has never before complained about my use of named entities (I prefer them to numerical entities because I can read them as I go), I'm guessing that this is because Microsoft has its own peculiar way of rendering html entities, which show on my text editor as ^R, ^V, and so forth. Is there an easy fix for this, such as by telling Sigil there are two entity modes in use? Mixing Word export and named entities has never caused me a problem in uploading html to Amazon or Barnes & Noble for conversion. Indeed, the very html file I have opened is already on sale as a Kindle and an epub book.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
can sigil convert html entities to regular punctuation?	Gregg Bell	Sigil	31	07-01-2013 09:29 AM
Epub no support for some named entities?	Points	ePub	25	11-19-2012 06:42 PM
named anchors fail validation in sigil	lkasdorf	Sigil	2	04-20-2012 10:55 PM
decimal entities in ePub instead of character entities	epub4ever	Calibre	4	04-20-2012 02:27 AM
Named entities or not?	alecE	ePub	17	07-21-2009 12:24 PM

07-13-2013, 10:46 AM	#2
theducks Well trained by Cats Posts: 29,782 Karma: 54830978 Join Date: Aug 2009 Location: The Central Coast of California Device: Kobo Libra2,Kobo Aura2v1, K4NT(Fixed: New Bat.), Galaxy Tab A	I use them all the time in Sigil (recently used &bul; ° 10 ° &bul; as a chapter # decoration) Something else is wrong. codepage?

07-13-2013, 11:42 AM	#3
Notjohn mostly an observer Posts: 1,515 Karma: 987654 Join Date: Dec 2012 Device: Kindle	Perhaps Sigil wants to see two codepages, one for Word-generated html, the other for named entities? I ran the Word doc through word2cleanhtml.com online, and now all is fine, with the minor cavil that all the entities (Word's and mine) have become numerical entities. In any event, Sigil has no problem opening the book.

07-13-2013, 11:48 AM	#4
theducks Well trained by Cats Posts: 29,782 Karma: 54830978 Join Date: Aug 2009 Location: The Central Coast of California Device: Kobo Libra2,Kobo Aura2v1, K4NT(Fixed: New Bat.), Galaxy Tab A	NJ did you think to Validate one of the failing files before feeding it to Sigil? That might give a clue.

07-13-2013, 12:50 PM	#5
DiapDealer Grand Sorcerer Posts: 27,546 Karma: 193191846 Join Date: Jan 2010 Device: Nexus 7, Kindle Fire HD	Gotta be the declaration. Named entities work fine, for the most part, but not unless the doctype and all the namespace declarations allow for them.

07-13-2013, 05:08 PM	#8
Notjohn mostly an observer Posts: 1,515 Karma: 987654 Join Date: Dec 2012 Device: Kindle	I wish I had a mnemonic so I would remember where that word came from! Yes, "declaration" must be the term I was looking for. I just use a generic <!DOCTYPE HTML>. Odd that Microsoft's ^R would pass muster, and the named entity not, when both are in the same file. I'd be less surprised if it were the other way around. Speaking of declarations, does MathML work in any e-book format?

07-14-2013, 03:08 AM	#10
Jellby frumious Bandersnatch Posts: 7,515 Karma: 18512745 Join Date: Jan 2008 Location: Spaniard in Sweden Device: Cybook Orizon, Kobo Aura	If I remember correctly, the DOCTYPE is not needed in ePub. But if it's present it must be correct (which implies you cannot have SVG code, for instance). It may be that Sigil wants it, but according to the spec, files without a DOCTYPE should be XHTML anyway (which normal named entities). The above is valid only for HTML-like files, of course, you cannot use named entities in XML files, such as the OPF or NCX, you must use numerical entities or the actual characters there. This means that you'd better not use named entities in titles, I guess (maybe Sigil could automatically convert entities to characters when creating the TOC, if it doesn't already).