Ligatures aren't correctly read #14740
Description
Steps to reproduce:
Open the following html in either Firefox or Chrome or Edge:
<html>
<body>
<span>fit and fix the refix</span>
</body>
</html>Actual behavior:
It's read as "t and x the rex".
Expected behavior:
It should be read as "fit and fix the refix".
The entity fi corresponds to the f and i ligature char and it's just omitted when read hence I suppose that the text should be normalized (with NFKC) probably before being read.
For example, when opening https://en.wikipedia.org/wiki/Ligature_(writing) some the ligatures are omitted by the reader when the text is read.
More generally, I do wonder if similarly to the copy event (see https://developer.mozilla.org/en-US/docs/Web/API/Element/copy_event) we should have a read event which will allow to transform the text before it'll be read by the reader.
NVDA version:
2022.4
Windows version:
11