■

When a character's codepoint is beyond U+10000, I should use 32-bit literal.

>>> '\u1f4a9'
'Ὂ9'
>>> '\U0001f4a9'
'💩'

If I try to input '💩' directly in Jupyter console on Windows cmd, it aborts with the error:

Traceback (most recent call last):
  File "C:\ProgramData\Anaconda3\Scripts\jupyter-console-script.py", line 10, in <module>
    sys.exit(main())
  File "C:\ProgramData\Anaconda3\lib\site-packages\jupyter_core\application.py", line 267, in launch_instance
    return super(JupyterApp, cls).launch_instance(argv=argv, **kwargs)
  File "C:\ProgramData\Anaconda3\lib\site-packages\traitlets\config\application.py", line 658, in launch_instance
    app.start()
  File "C:\ProgramData\Anaconda3\lib\site-packages\jupyter_console\app.py", line 155, in start
    self.shell.mainloop()
  File "C:\ProgramData\Anaconda3\lib\site-packages\jupyter_console\ptshell.py", line 508, in mainloop
    self.interact()
  File "C:\ProgramData\Anaconda3\lib\site-packages\jupyter_console\ptshell.py", line 492, in interact
    code = self.prompt_for_code()
  File "C:\ProgramData\Anaconda3\lib\site-packages\jupyter_console\ptshell.py", line 440, in prompt_for_code
    reset_current_buffer=True)
  File "C:\ProgramData\Anaconda3\lib\site-packages\prompt_toolkit\interface.py", line 415, in run
    self.eventloop.run(self.input, self.create_eventloop_callbacks())
  File "C:\ProgramData\Anaconda3\lib\site-packages\prompt_toolkit\eventloop\win32.py", line 80, in run
    for k in keys:
  File "C:\ProgramData\Anaconda3\lib\site-packages\prompt_toolkit\terminal\win32_input.py", line 143, in read
    all_keys = list(self._get_keys(read, input_records))
  File "C:\ProgramData\Anaconda3\lib\site-packages\prompt_toolkit\terminal\win32_input.py", line 186, in _get_keys
    for key_press in self._event_to_key_presses(ev):
  File "C:\ProgramData\Anaconda3\lib\site-packages\prompt_toolkit\terminal\win32_input.py", line 225, in _event_to_key_presses
    ascii_char = u_char.encode('utf-8')
UnicodeEncodeError: 'utf-8' codec can't encode character '\ud83d' in position 0: surrogates not allowed

This is because ev.uChar.UnicodeChar returns one of a surrogate pair, \ud83d, instead of a complete character \U0001F4A9. It depends on Windows KEY_EVENT_RECORD structure (https://docs.microsoft.com/en-us/windows/console/key-event-record-str).

It seems _event_to_key_resses(ev) is called for each element of a surrogate pair. So we should check the key press is one of a surrogate pair and store an element of the pair.

    def _get_keys(self, read, input_records):
        """
        Generator that yields `KeyPress` objects from the input records.
        """
        for i in range(read.value):
            ir = input_records[i]

            # Get the right EventType from the EVENT_RECORD.
            # (For some reason the Windows console application 'cmder'
            # [http://gooseberrycreative.com/cmder/] can return '0' for
            # ir.EventType. -- Just ignore that.)
            if ir.EventType in EventTypes:
                ev = getattr(ir.Event, EventTypes[ir.EventType])

                # Process if this is a key event. (We also have mouse, menu and
                # focus events.)
                if type(ev) == KEY_EVENT_RECORD and ev.KeyDown:
                    for key_press in self._event_to_key_presses(ev):
                        yield key_press

                elif type(ev) == MOUSE_EVENT_RECORD:
                    for key_press in self._handle_mouse(ev):
                        yield key_press