How to update Kindle Previewer render engine?

[nodymoha]

Hi,

I am kinda new to creating books for kindle, especially KFX format.

First, I am trying ti write an Arabic ebook with special font that has special font character for each number.

However, I am having and issue with how kindle previewer render these characters.

As you can see in this picture.
Lift side: Calibre Reader
Right side: Kindle Previewer 3



Strangely, I have encountered the same issue in the software I am using to edit this book.

Lift side: Sigil 1.9
Right side: Sigil 2.0


Does anyone know how can I fix this issue? Or force Kindle Previewer to use new renderer?

[Quoth]

Is it for your own use or to publish on Amazon via KDP?

[nodymoha]

Originally Posted by Quoth

Is it for your own use or to publish on Amazon via KDP?

It is for me only.

I will not publish it in KDP.

[jhowell]

It can be difficult to get embedded fonts to work properly on the Kindle platform. Sometimes the you need to select Publisher Font in the reader to see them.

You might receive more help if you provide an e-pub file that demonstrates the problem. Perhaps there is a problem with your html/css.

[nodymoha]

Originally Posted by jhowell

Sometimes the you need to select Publisher Font in the reader to see them.

That what I have been doing.

Originally Posted by jhowell

You might receive more help if you provide an e-pub file that demonstrates the problem. Perhaps there is a problem with your html/css.

I have uploaded the demo I have used in the main post.

Already tried to change many css attributes but none of them worked.

What confuse me more is that it doesn't work with Sigil 1.9 but it works with Sigil 2.0.
So, what change between these two version that allowed these characters to be displayed correctly.

[jhowell]

I see the same rendering issue in the Kindle Previewer. I also tried converting it to KFX and viewing it using the Kindle for iOS app and that produced the same results as the Kindle Previewer.

From what I can tell the numeric glyphs in your font have some sort of attribute that indicates they are to be combined into a single glyph when rendered. This is probably something that just isn't supported properly in KFX. I don't have any suggestions of how you might make it work.

[DNSB]

The difference between Sigil 1.9 and Sigil 2.0 is likely due to the different QT version using and how it handles combining glyphs.

Since the Uthman font does seem to have the characters you want in the Arabic Presentation area ( specifically U+FC00 to U+FD1D and U+FD50 to U+FD7D ), it might be worth trying to use those glyphs directly. I generated the hex font names and dumped them into your test epub. Oddly, when I did a mend and prettify, the glyphs displayed in the Sigil code view do not match those glyphs when seen in Preview but do have some semblance to the components of the glyph shown in FontForges Glyph Info => Components.

[nodymoha]

Thank you so much jhowell for your effort in helping me with this problem.

Thank you so much DNSB for this great table of characters.
I wonder how did you get the unicode for these glyphs because I was trying to get them but the font editor I am using says no unicode number for any of these glyphs.

I guess now I will write another python code to replace the numbers by its corresponding glyphs unicode.

[nodymoha]

Thank you jhowell for yor efforts in helping me with this problem.

Thank you DNSB for your great solution.
I wonder how did you get the unicode for these glyphs.
Cause I was trying to find the unicode for these glyphs but my software says there are no unicode number for any of the glyphs.

Now, I guess I will use DNSB modified epub as a reference for the glyphs and replace these glyphs manually with its corresponding unicode number.

[DNSB]

Originally Posted by nodymoha

Thank you so much jhowell for your effort in helping me with this problem.

Thank you so much DNSB for this great table of characters.
I wonder how did you get the unicode for these glyphs because I was trying to get them but the font editor I am using says no unicode number for any of these glyphs.

I guess now I will write another python code to replace the numbers by its corresponding glyphs unicode.

I used an open source project called FontForge to inspect the font. Once I knew where the fonts lived, I generated a list of the hex codes with the &#x prefix and ; suffix appended (stupid repetitive work is why computers were invented. ).

[nodymoha]

Originally Posted by DNSB

I used an open source project called FontForge to inspect the font. Once I knew where the fonts lived,

I guess I might try Font Forge in my coming projects.

Originally Posted by DNSB

I generated a list of the hex codes with the &#x prefix and ; suffix appended (stupid repetitive work is why computers were invented. ).

Exactly, that's why I wrote a simple script to replace these characters in 100+ pages.

Programming is a gift that everyone should have it.