Cannot paste unicode into the REPL

[tannewt]

It should echo back unicode characters just fine even if they don't show on the display's terminal.

[DavePutz]

For UTF-8 characters the fix seems simple: In lib/mp-readline/readline.c replace line 209
} else if (32 <= c && c <= 126) {
with
} else if (32 <= c ) {

This allows REPL to handle at least the UTF-8 characters:

   >>> s='27° C'
   >>> print(s)
   27° C

But there seem to be some other issues:

   >>> s='27° C'
   >>> s
   '27 C'       // no degree symbol
   >>> print (len(s))
   4              //should be 5 if the degree symbol is included

And, of course, UTF-16 characters aren't covered.

Is the simple patch worth a PR, or do the other issues need to be addressed as well?

Note that CPython (tested on 3.6.8) throws an error in REPL:

`   >>> s='27° C'`
  File "<stdin>", line 0

    ^
SyntaxError: 'utf-8' codec can't decode byte 0xb0 in position 5: invalid start byte

[tannewt]

Thanks for taking a look at this @DavePutz! I think a small simple patch that improves the current situation would be great. For the other issues we can file separate issues and/or PRs.

I wouldn't worry about UTF-16. UTF-8 is most common.

[dhalbert]

Fixed by #2631.