e-82C to e-837 are broken - they are multi-codepoint sequences

[iamcal]

From an email conversation with 杜志刚:

On 3/27/2011 9:23 PM, 杜志刚 wrote:

Please see
https://www.unicode.org/~scherer/emoji4unicode/snapshot/full.html#e-82C (e-82C
~ e-837)
There are BAD multiple code points :

|| num   ||chr|| first       || second   ||
|| e-82C || # || U+0023  || U+20E3 ||
|| e-82E || 1 || U+0031  || U+20E3 ||
|| e-82F || 2 || U+0032  || U+20E3 ||
...
|| e-837 || 0 || U+0030  || U+20E3 ||

The second code is BAD !

Ahhh! Now I understand what's happening here. U+20E3 is 'COMBINING ENCLOSING KEYCAP' and it draws a box around the character that comes before it, so the sequence "U+0032 U+20E3" displays a '2' in a square box, which is what we want. I think I want to modify the parser to use the sequence of 2 unicode codepoints instead of just one or the other.

The problem with using the first one is that it replaces characters where you don't expect it to - it will replace any numbers in text with the numbers-in-boxes emoji, which is almost never what you want (and the reason I changed it to begin with).

--cal

[iamcal]

this was patched on March 28, 2011 - forgot to close it i guess