Kobo and interlinear epub documents

[clairvoyant]

I currently have a Kindle Paperwhite for language learning but find it slow and tedious to look up individual words. So I looked into interlinear texts and found this thread by @Doitsu as well as source code to generate interlinear epub documents through Google translate.

Now I'm considering getting a Kobo Libra H2O but I'm wondering if they support interlinear books. Or does maybe KOReader support it if I install it on Kobo?

[DNSB]

I ran into this a couple of years back and thought it looked interesting. I played with Doitsu's original idea of using <ruby> since I knew that worked with Kobo's webkit based renderer. I converted his original to an epub3 document and using this in combination with Kobo's alternate renderer worked quite happily after sending it to my Kobo using the KoboTouchExtended driver to send as a kepub. I've attached a screenshot from a Kobo Forma.

Edit: Finding that document makes me realize that I really need to do some cleanup.

[clairvoyant]

I appreciate the response @DNSB.

So I went ahead and purchased a Kobo device and wrote a software in Rust to accomplish this. It automatically fetches all the words, translates them and then export a bilingual epub. I then use Calibre with Kobo Touch Extended Driver installed to convert the book into kepub while transferring it to my device.

<ruby> looks great but unfortunately I ran into two issues:

1. Loading a chapter makes the device freeze for about 50 seconds while showing a blank page with the book title visible in the footer. Once the chapter (xhtml file I assume) is loaded, turning pages is as responsive as usual.
2. Clicking on a word freezes the device for a couple of seconds and does nothing. When trying to select multiple words, it freezes and sometimes selects all the words from your word to the beginning of the page.

I assume it's due to how Kobo parse <ruby>. Code example:

<p>
    <ruby>
        Die
        <rt>the</rt>
    </ruby>
    <ruby>
        Unglücksfälle
        <rt>accidents</rt>
    </ruby>
    <ruby>
        des
        <rt>of the</rt>
    </ruby>
    <ruby>
        Studenten
        <rt>students</rt>
    </ruby>
    <ruby>
        Anselmus.
        <rt>*</rt>
    </ruby>
</p>

When converted to kepub by Kepubify for example:

<p>
    <ruby>
        <span class="koboSpan" id="kobo.13.1">Die</span>
        <rt><span class="koboSpan" id="kobo.13.2">the</span></rt>
    </ruby>
    <span class="koboSpan" id="kobo.13.3"> </span>
    <ruby>
        <span class="koboSpan" id="kobo.13.4">Unglücksfälle</span>
        <rt><span class="koboSpan" id="kobo.13.5">accidents</span></rt>
    </ruby>
    <span class="koboSpan" id="kobo.13.6"> </span>
    <ruby>
        <span class="koboSpan" id="kobo.13.7">des</span>
        <rt><span class="koboSpan" id="kobo.13.8">of the</span></rt>
    </ruby>
    <span class="koboSpan" id="kobo.13.9"> </span>
    <ruby>
        <span class="koboSpan" id="kobo.13.10">Studenten</span>
        <rt><span class="koboSpan" id="kobo.13.11">students</span></rt>
    </ruby>
    <span class="koboSpan" id="kobo.13.12"> </span>
    <ruby>
        <span class="koboSpan" id="kobo.13.13">Anselmus.</span>
        <rt>*</rt>
    </ruby>
</p>

To debug, I tried changing <ruby> to a similar format with <span> and relying on CSS to get the same style. It seems to have solved both the issues but results in Kobo ignoring the connection and sometimes have page breaks between a word and its translation. Playing around with "break-inside: avoid;" doesn't seem to make any difference. <ruby> doesn't seem to have this issue.

Is there anything else I could try to fix the issues?