Handling "long" (non-BMP) Unicode characters in autocorrect.txt

Question

Some Unicode characters have a code longer than 4 hex digits. For instance David Rowe was trying to help someone who needed to type U+1133c GRANTHA SIGN NUKTA.

The usual method of \u followed by the digits fails with any code longer than 4 digits. (A Unicode character outside the Basic Multilingual Pane). But it is possible to write this as a string of two 4 digit Unicode characters if converted to UTF-16. So this rule would give the desired character when typing two periods:
..-->\ud804\udf3c

David found the UTF16 equivalent of his desired character from the Branah Unicode converter page
https://www.branah.com/unicode-converter

You can also find it by searching FileFormat.Info for U+1133c, the page on that character gives the UTF16 coding equivalent for it.