writing text for epub3

[robertmchicago]

I am writing a reflowable epub 3. I have learned a lot, but I am still learning.
I know that we do this:
use &lt; instead of <
use &gt; instead of >
use &amp; instead of &
My question: Are there other things on the keyboard that we don't put directly from the keyboard?
Thanks a lot

[JSWolf]

Originally Posted by robertmchicago

I am writing a reflowable epub 3. I have learned a lot, but I am still learning.
I know that we do this:
use &lt; instead of <
use &gt; instead of >
use &amp; instead of &
My question: Are there other things on the keyboard that we don't put directly from the keyboard?
Thanks a lot

If this ePub 3 is going to be for sale, make sure you make it as backwards compatible with ePub2 as you can.

[Turtle91]

Originally Posted by robertmchicago

I am writing a reflowable epub 3. I have learned a lot, but I am still learning.
I know that we do this:
use &lt; instead of <
use &gt; instead of >
use &amp; instead of &
My question: Are there other things on the keyboard that we don't put directly from the keyboard?
Thanks a lot

You really don’t need to worry about all those nitnoid details when you use the right tools.

I would actually recommend using an authoring program to write the book, or, at least, a word processor. Then use SaveAs to save as html format. Then you can import that html into an ePub editor like Sigil, or Calibre Editor, to do all the epubifying.

[Slevin#7]

Originally Posted by robertmchicago

My question: Are there other things on the keyboard that we don't put directly from the keyboard?
Thanks a lot

No, these three are the only ones, considered special characters, thus they have to be masked.

Since you are writing for EPUB3 you deal with XHTML5 and UTF-8, which means, you can safely use and should use all other characters directly for your main text. There are some exceptions when you wanna put special chars into tag-property values like for 'aria-label', but I would strongly encourage you to not use special chars in these values.

Also, as mentioned before, it will make life easier when you use a proper EPUB tool like Sigil (it's free), which can help you alot with all this tideous stuff.

[robertmchicago]

Originally Posted by JSWolf

If this ePub 3 is going to be for sale, make sure you make it as backwards compatible with ePub2 as you can.

Yes, this epub 3 will be for sale, I will try backward compatibility if I can—thank you for the reminder.

[robertmchicago]

Originally Posted by Turtle91

You really don’t need to worry about all those nitnoid details when you use the right tools.

I would actually recommend using an authoring program to write the book, or, at least, a word processor. Then use SaveAs to save as html format. Then you can import that html into an ePub editor like Sigil, or Calibre Editor, to do all the epubifying.

I am learning Sigil. I already have tried : Word - html filtered - Sigil, but it needed a lot of cleaning, so decided to use only Sigil.

[robertmchicago]

Originally Posted by Slevin#7

No, these three are the only ones, considered special characters, thus they have to be masked.

Since you are writing for EPUB3 you deal with XHTML5 and UTF-8, which means, you can safely use and should use all other characters directly for your main text. There are some exceptions when you wanna put special chars into tag-property values like for 'aria-label', but I would strongly encourage you to not use special chars in these values.

Also, as mentioned before, it will make life easier when you use a proper EPUB tool like Sigil (it's free), which can help you alot with all this tideous stuff.

That really helps! I am already learning Sigil. I’ll remember not to use these in tag values.
But are you sure that keyboard keys like % * - _ = + ' " ; : don't need to be changed? For example we use these in tags too: = "

[Slevin#7]

Originally Posted by robertmchicago

But are you sure that keyboard keys like % * - _ = + ' " ; : don't need to be changed?"

Yes, I am sure. The parser won't catch them as functional characters, it will only seek for < and > which are part of HTML tags itself, and the & char which will get interpreted as the beginning of an unicode expression. So there is no need to worry. In fact, when coding for EPUB3 Sigil will swap your unicode or HTML entities with the actual chars, since this is the way to go.

Also, without additional information, the parser wouldn't even be able to assign the required chars when using HTML entities, since HTML5 doesn't require these additional informations anymore, thus having no assigning table in hand out of the box.

[DNSB]

Just to toss in my 5 cents worth, in an ePub3 files, you must use &amp;, &lt;, &gt;, &apos; and &quot; and not &, <, >, ' and ".

[Slevin#7]

Originally Posted by DNSB

Just to toss in my 5 cents worth, in an ePub3 files, you must use &amp;, &lt;, &gt;, &apos; and &quot; and not &, <, >, ' and ".

&apos; and &quot; are only required when used inside the tags, otherwise the parser will treat them as tag content and will not apply any special meaning to them.

The XML (XHTML) parser will go through the document char by char, it will recoginze a < as an opening sign of either an opening- or closing tag and will then change its mode to treat the following chars as part of the tag, until it hits the > char and will then know that the tag itself has ended. Within an openening and a closing tag, everything will be treated as tag-value or tag-content. The parser will then only seek for the next < char to find the next tag. XML (XHTML) is all about tags, there is no further special programming language involved in the meaning of Turing. The only other special sign which will be treated by the parser as functional is the &, since as mentioned before, it is the hook for the beginning of unicode codes (respectively HTML entities).

[Slevin#7]

And to get a little more into detail: Your five codes are the only ones HTML5 knows by default, since the <, > and & are mainly relevant for the content itself and the other two ' and " have to be used inside the tags since the property-values must be wrapped within them (either ' or "). All other entities like &copy; and so on couldn't even get rendered since no assigning table is provided so far. You'd have to assign these codes manually to help the parser which unicode to use for them. This is one of the many benefits of XHTML5 and the required UTF-8 format.

[Slevin#7]

My last 2 cents in addition: ' and " are only recognized by the parser as functional characters inside a tag.

When the parser parses a tag itself, the composition of a tag has to follow special rules. Chars like = assign a value to a property, where the property must be known to the parser and the value has to be wrapped inside ' or ". So The only other two chars which may have to be masked are these two quotes - in case you wanna put such a character into your value (and of course inside a tag value you have to mask the other special chars as well). Since when you put such a quote inside your value, the parser would interpret " or ' as the end of the value and this is not what you want.

Usually in HTML it is best practice to use " to wrap your values, which means, you could even use ' inside the value without the requirement of masking this char. The parser takes the opening quote " and only looks for the same quote to close the value. It works the other way around as well, opening with ' and using " inside the values. But as mentioned before, you should avoid all these special character inside your values at all, at least as far as you can. There may be some cases like metadata for titles where you have to put them into your values, but most of the time there is no need for that.

[RbnJrg]

Originally Posted by robertmchicago

Yes, this epub 3 will be for sale, I will try backward compatibility if I can.

You can and you must (since your epub is for sale). And that is going to be a "funny" task (especially if you use advanced css3 properties to style your epub).

[robertmchicago]

Originally Posted by Slevin#7

My last 2 cents in addition: ' and " are only recognized by the parser as functional characters inside a tag.

When the parser parses a tag itself, the composition of a tag has to follow

Thank you so much for all these clear explanations. Everything you said makes perfect sense. I really appreciate you sharing your knowledge. It’s helping me a lot and making things much easier for me. Thanks again!

[robertmchicago]

Originally Posted by RbnJrg

You can and you must (since your epub is for sale). And that is going to be a "funny" task (especially if you use advanced css3 properties to style your epub).

Thanks for the advice. My learning process depends on: Sigil, PageEdit, Sigil Guide (which is epub 3), PageEdit Guide (which is epub 2), w3schools.com, and this very helpful MobileRead forum. With all this, I’m sure I’ll soon be able to make my epub 3 files backward compatible with epub 2, for example, by using the NCX file and avoiding things that don’t work in EPUB 2. Thanks again for pointing me in the right direction.