Topics

Reading All Unicode Characters with NVDA


Dzhovani
 

Hi all,
Currently, NVDA does not read all unicode symbols it encounters. While there is good support for emois, greek letters and some math symbols are not covered.
1. Can someone give me an idea how the symbol pronunciation dialog works and what is the relationship with the pronunciation level in the synthesizer?
2. Can it be extended to all unicode characters? Are there performance considerations included?
Thanks in advance,
Dzhovani


Quentin Christensen
 

There are currently several issues requesting various unicode characters.  I think the issue might be that there is no standard description for them that we could simply refer to - each one needs to be added by hand.  I haven't looked into it myself so might not be correct - or maybe as you wondered, they could all be added but it would cause a performance impact.  

Here is a link to a list of open issues which came up when I searched for "unicode": https://github.com/nvaccess/nvda/issues?q=is%3Aissue+is%3Aopen+unicode

With the punctuation / symbol pronunciation dialog, yes you can add symbols to that.  Open it, select add, type or paste the symbol, type the text to speak and a comment if desired.  The symbol level is the level the symbol text is read at, vs sending the symbol itself to the synthesizer.  For instance, the comma "," is set to symbol level "all".  There are four levels: "None", "Some", "Most", and "All".  NVDA only reads "comma" at Symbol level "all".  At lower symbol levels (none, some or most), NVDA simply sends the comma itself to the synthesizer, which acknowledges it by a slight pause in reading.  If the symbol level for comma was set to "some", then NVDA would pause at symbol level "none", but read the word "comma" at symbol level some, most and all.

Kind regards

Quentin.


On Thu, Nov 12, 2020 at 7:49 PM Dzhovani <dzhovani.chemishanov@...> wrote:
Hi all,
Currently, NVDA does not read all unicode symbols it encounters. While there is good support for emois, greek letters and some math symbols are not covered.
1. Can someone give me an idea how the symbol pronunciation dialog works and what is the relationship with the pronunciation level in the synthesizer?
2. Can it be extended to all unicode characters? Are there performance considerations included?
Thanks in advance,
Dzhovani



--
Quentin Christensen
Training and Support Manager


Dzhovani
 

Hi,
Thank you for the information. I found this as one of the main discussions on the topic.
https://github.com/nvaccess/nvda/issues/3805
Unfortunately, the greek letters issues is mentioned only ones and not after that.
Regards,
Dzhovani