I don't know why, but in Python, code points of frame characters are:
upper_left  = 1066
upper_right = 1111
lower_left  = 1040
lower_right = 1065
horizontal  = 1044
vertical    = 1110
That's really strange, because there are no common codecs encode those characters as such.
Could Admin fix this or at lease give some explanations please?

Also, don't put non-ascii characters in your source code, but use:
  c == chr(1066)
or
  ord(c) == 1066

ord('┌') == 9484
list('┌'.encode('cp437')) == [218]
list('┌'.encode('utf_16_le')) == [12, 37] # 12+37*256 == 9484

and Unicode code point is: U+250C == 9484