|
1 |
| -from base64 import urlsafe_b64encode, urlsafe_b64decode |
| 1 | +from typing import Iterator, List |
2 | 2 |
|
3 | 3 | __all__ = ["base64", "unbase64"]
|
4 | 4 |
|
| 5 | +Base64String = str |
5 | 6 |
|
6 |
| -def base64(s: str) -> str: |
7 |
| - """ "Encode the string s using Base64.""" |
8 |
| - return urlsafe_b64encode(s.encode("utf-8")).decode("utf-8") |
| 7 | + |
| 8 | +def base64(s: str) -> Base64String: |
| 9 | + """Encode the string s using Base64.""" |
| 10 | + if isinstance(s, (bytearray, bytes)): |
| 11 | + s = s.decode("unicode") # handle encoded string gracefully |
| 12 | + |
| 13 | + unicode_list = list(str_to_unicode_seq(s)) |
| 14 | + length = len(unicode_list) |
| 15 | + rest = length % 3 |
| 16 | + result: List[str] = [] |
| 17 | + extend = result.extend |
| 18 | + |
| 19 | + for i in range(0, length - rest, 3): |
| 20 | + a, b, c = unicode_list[i : i + 3] |
| 21 | + result.extend( |
| 22 | + ( |
| 23 | + first_6_bits(a), |
| 24 | + last_2_bits_and_first_4_bits(a, b), |
| 25 | + last_4_bits_and_first_2_bits(b, c), |
| 26 | + last_6_bits(c), |
| 27 | + ) |
| 28 | + ) |
| 29 | + |
| 30 | + if rest == 1: |
| 31 | + a = unicode_list[-1] |
| 32 | + extend((first_6_bits(a), last_2_bits_and_first_4_bits(a, 0), "==")) |
| 33 | + elif rest == 2: |
| 34 | + a, b = unicode_list[-2:] |
| 35 | + extend( |
| 36 | + ( |
| 37 | + first_6_bits(a), |
| 38 | + last_2_bits_and_first_4_bits(a, b), |
| 39 | + last_4_bits_and_first_2_bits(b, 0), |
| 40 | + "=", |
| 41 | + ) |
| 42 | + ) |
| 43 | + |
| 44 | + return "".join(result) |
| 45 | + |
| 46 | + |
| 47 | +def first_6_bits(a: int) -> str: |
| 48 | + return to_base_64_char(a >> 2 & 0x3F) |
| 49 | + |
| 50 | + |
| 51 | +def last_2_bits_and_first_4_bits(a: int, b: int) -> str: |
| 52 | + return to_base_64_char((a << 4 | b >> 4) & 0x3F) |
| 53 | + |
| 54 | + |
| 55 | +def last_4_bits_and_first_2_bits(b: int, c: int) -> str: |
| 56 | + return to_base_64_char((b << 2 | c >> 6) & 0x3F) |
| 57 | + |
| 58 | + |
| 59 | +def last_6_bits(c: int) -> str: |
| 60 | + return to_base_64_char(c & 0x3F) |
9 | 61 |
|
10 | 62 |
|
11 | 63 | def unbase64(s: str) -> str:
|
12 |
| - """ "Decode the string s using Base64.""" |
13 |
| - return urlsafe_b64decode(s).decode("utf-8") |
| 64 | + """Decode the string s using Base64.""" |
| 65 | + if isinstance(s, (bytearray, bytes)): |
| 66 | + s = s.decode("ascii") # handle encoded string gracefully |
| 67 | + |
| 68 | + unicode_list: List[int] = [] |
| 69 | + extend = unicode_list.extend |
| 70 | + length = len(s) |
| 71 | + |
| 72 | + for i in range(0, length, 4): |
| 73 | + try: |
| 74 | + a, b, c, d = [from_base_64_char(char) for char in s[i : i + 4]] |
| 75 | + except (KeyError, ValueError): |
| 76 | + return "" # for compatibility |
| 77 | + bitmap_24 = a << 18 | b << 12 | c << 6 | d |
| 78 | + extend((bitmap_24 >> 16 & 0xFF, bitmap_24 >> 8 & 0xFF, bitmap_24 & 0xFF)) |
| 79 | + |
| 80 | + i = length - 1 |
| 81 | + while i > 0 and s[i] == "=": |
| 82 | + i -= 1 |
| 83 | + unicode_list.pop() |
| 84 | + |
| 85 | + return "".join(unicode_list_to_str(unicode_list)) |
| 86 | + |
| 87 | + |
| 88 | +b64_characters = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/" |
| 89 | + |
| 90 | +b64_character_map = {c: i for i, c in enumerate(b64_characters)} |
| 91 | + |
| 92 | + |
| 93 | +def to_base_64_char(bit_map_6: int) -> str: |
| 94 | + return b64_characters[bit_map_6] |
| 95 | + |
| 96 | + |
| 97 | +def from_base_64_char(base_64_char: str) -> int: |
| 98 | + return 0 if base_64_char == "=" else b64_character_map[base_64_char] |
| 99 | + |
| 100 | + |
| 101 | +def str_to_unicode_seq(s: str) -> Iterator[int]: |
| 102 | + for utf_char in s: |
| 103 | + code = ord(utf_char) |
| 104 | + if code < 0x80: |
| 105 | + yield code |
| 106 | + elif code < 0x800: |
| 107 | + yield 0xC0 | code >> 6 |
| 108 | + yield 0x80 | code & 0x3F |
| 109 | + elif code < 0x10000: |
| 110 | + yield 0xE0 | code >> 12 |
| 111 | + yield 0x80 | code >> 6 & 0x3F |
| 112 | + yield 0x80 | code & 0x3F |
| 113 | + else: |
| 114 | + yield 0xF0 | code >> 18 |
| 115 | + yield 0x80 | code >> 12 & 0x3F |
| 116 | + yield 0x80 | code >> 6 & 0x3F |
| 117 | + yield 0x80 | code & 0x3F |
| 118 | + |
| 119 | + |
| 120 | +def unicode_list_to_str(s: List[int]) -> Iterator[str]: |
| 121 | + s.reverse() |
| 122 | + next_code = s.pop |
| 123 | + while s: |
| 124 | + a = next_code() |
| 125 | + if a & 0x80 == 0: |
| 126 | + yield chr(a) |
| 127 | + continue |
| 128 | + b = next_code() |
| 129 | + if a & 0xE0 == 0xC0: |
| 130 | + yield chr((a & 0x1F) << 6 | b & 0x3F) |
| 131 | + continue |
| 132 | + c = next_code() |
| 133 | + if a & 0xF0 == 0xE0: |
| 134 | + yield chr((a & 0x0F) << 12 | (b & 0x3F) << 6 | c & 0x3F) |
| 135 | + continue |
| 136 | + d = next_code() |
| 137 | + yield chr((a & 0x07) << 18 | (b & 0x3F) << 12 | (c & 0x3F) << 6 | d & 0x3F) |
0 commit comments