Math Expressions with Special Letters, "i," and "t" Are Verbalized Incorrectly and Not Navigable

[NV-Codes]

MathCAT Version

MathCAT-0.6.6 (NVDA Add-On)

Description of Issue

It seems that certain sequences of symbols, namely those that include $i$, $t$, and special letters, are not verbalized as separate adjacent characters. Instead, they are verbalized as if they were strings to be interpreted as words, and the individual symbols are not accessible in MathCAT's navigation mode.

Examples

Expression	Rough Pronunciation	Expected Verbalization
$i \omega t$	"it"	"i, omega, t"
$t i \alpha$	"tee"	"t, i, alpha"
$i \hbar t$	"it"	"i, reduced Planck constant, t"
$t \beta \alpha i$	"tee"	"t, beta, alpha, i"

[NSoiffer]

Can you include the MathML that is generating the bad pronunciation? I suspect that the problem lies with the input, but I need to see the input to confirm that.

[NV-Codes]

Sure! The following MathML was obtained by using Ctrl + C on the above expressions. They were originally typed in LaTeX to be converted to MathML by GitHub automatically.

$i \omega t$

<math xmlns='http://www.w3.org/1998/Math/MathML'>
 <mi>iωt</mi>
</math>

$t i \alpha$

<math xmlns='http://www.w3.org/1998/Math/MathML'>
 <mi>tiα</mi>
</math>

$i \hbar t$

<math xmlns='http://www.w3.org/1998/Math/MathML'>
 <mi>iℏt</mi>
</math>

$t \beta \alpha i$

<math xmlns='http://www.w3.org/1998/Math/MathML'>
 <mi>tβαi</mi>
</math>

[NSoiffer]

As I suspected, it is bad MathML. If you look at the contents of the <mi>s, all the characters are in a single <mi>. It appears that git is not good at generating MathML from TeX. Here is what I get if I use MathJax V3 to do the conversion for the first example(i \omega t (right click on the rendered expression and choose "Copy to Clipboard: MathML Code"):

<math xmlns="http://www.w3.org/1998/Math/MathML" display="block">
  <mi>i</mi>
  <mi>&#x3C9;</mi>
  <mi>t</mi>
</math>

Notice that each variable is in its own mi. Multi-character variables such as sin and lim shouldn't be split. I think it would likely result in more errors than good repairs if MathCAT tried to fix these.

Closing because this is not a MathCAT bug.