There is currently a Public Review, number 405.
http://www.unicode.org/review/pri405/
It is about the following document.
http://www.unicode.org/reports/tr51/tr51-17.html
The issue of screen readers is mentioned in the document.
I have thought of a possible solution.
However I am not expert on many of the details of what is allowed and
what is not allowed in Unicode text, so I am posting the idea here so
that depending upon any discussion that takes place, I might send in the
idea as a formal response to the Public Review, or send in a modified
form based on advice provided, or just abandon the idea as unworkable.
Here is the basic idea as I suggest it at the moment, please endorse,
reject, discuss, improve the idea as you think best.
Decide what text, in any Unicode characters that you wish in any
language you choose, is to be the text that the screen reader speaks.
Save that text as a UTF-8 byte sequence.
Encode that text in its UTF-8 form to produce a text string twice as
long as that UTF-8 string such that, byte by byte, each UTF-8 byte is
encoded as two hexadecimal "digits" each in the range 0..9, A..F and
then use the tag version of each of those characters.
Add a U+0020 SPACE character at the front as the base character and add
a cancel tag character at the end.
Include that string in the document after the QID emoji character.
With my limited knowledge of the intricacies of Unicode it seems to me
that that might well solve the problem.
Screen reader software could decode the tag characters into a string and
try to speak it out.
Other software would just ignore the tag characters and display the
space character.
William Overington
Wednesday 25 September 2019
