The current implementation allows a browser to copy the text, with the math symbols as unicode.
I wonder if it would be beneficial to:
- Have a mode where the original source code could (still) be copied instead.
- Whether the copied code could be formatted to match the actual source. At the moment a lot of new lines get added.
I suspect (1-2) are feasible using different approaches, ie, accessing the raw code depending on selected lines, or by keeping a copy of the DOM editor element in memory.