Re: [orca-list] Does Orca use SSML tags other than speak and mark?

[Rastislav Kiss <rastislav kish gmail com>]

Hello,
thanks much for your quick response. This state is great, as it means
less work for me. :)

I had been considering SSML for language stuff as well, but I have
eventually refused the idea. The problem is, that SSML doesn't allow
you to switch speech modules, only voices, as a speech engine can't
replace itself with another speech engine.
Thus modules designed primarily for Chinese wouldn't be usable with
espeak for example, or if one wanted for example espeak for normal
usage and sapi bridge for cyrillic, this wouldn't be possible either.
My curent approach is bit more universal in this manner, although it
was noticeably bigger drag to implement. :)

There is a limitation for a speech module in reading html language
tags, which are parsed out before they reach it. But as I generally
hate this feature in all screenreaders anyway, I won't miss it much. :D

best regards

Rastislav
V Sobota,  18. júl 2020 o 01:20 +0200, Samuel Thibault napísal(a):
> Hello,
>
> Rastislav Kiss via orca-list, le sam. 18 juil. 2020 00:59:22 +0200, a
> ecrit:
> > I'd like to ask, does Orca use SSMl tags other than speak and mark
> > when
> > sending messages to speech dispatcher?
>
> You can see that in speechdispatcherfactory.py: only <speak>, <mark>,
> and &lt;, &gt;, &amp; are used currently (and &quot; and &apos; will
> be
> used sooner or later).
>
> > Is this situation likely to change in future?
>
> Possibly. Actually, your very mention of english vs mandarin raises
> the question of multilingualization. Orca should be able to get from
> the application the language of the text, and consequently tell
> speech-dispatcher which synthesis to be used. For text containing
> several languages, the language SSML tag would then be useful.
>
> > I've made my own speech-dispatcher module, which doesn't contain
> > speech
> > synthesis on its own, but which is able to load other sd modules
> > and
> > after receiving text to speak, split it to parts written in various
> > alphabets, including latin and chinese and send the respective
> > parts in
> > right time to right engine according to configuration.
>
> That's interesting, that would probably be useful to some extent.
>
> Samuel