test_unicode.py 文件源码

def main():
    # ????????????Unicode??,????????????
    print("A")
    print("\u0041")
    print("\N{LATIN CAPITAL LETTER A}")

    # ????????encode()????????????????
    print("\u00A1Help!")
    print("\u00A1Help!".encode("utf-8"))
    print(b"\xc2\xa1Help!".decode("utf-8"))

    # unicodedata??,???????????????.???????????Unicode?????.
    # ????unicodedata??,???????????Unicode?????.?????????????????.
    # ??????????????,??????????UTF-8
    data = b"\xd9\x85\xd8\xb1\xd8\xad\xd8\xa8\xd8\xa7\xd8\xa3\xd9\x84\xd8\xa7\xd9\x86"
    print(data.decode("utf-8"))
    # ?ud.name()?????????,??????
    for ch in data.decode("utf-8"):
        print(ord(ch), ud.name(ch))