Add code and packaging structure

base122.py

@@ -0,0 +1,106 @@
+"""
+Implementation of Base122 in pure Python.
+It attempts to mirror behaviour of the base64 built-in module.
+
+Its API is not yet complete or final. This is a first version for testing.
+"""
+
+illegals = [0, 10, 13, 34, 38, 92]
+
+def b122encode(raw_data):
+    if isinstance(raw_data, str):
+        raise ValueError("Input should be bytes")
+
+    out_data = []
+    cur_index = 0
+    cur_bit = 0
+
+    def get_7():
+        nonlocal cur_index, cur_bit
+        if cur_index >= len(raw_data):
+            return False
+
+        first_byte = raw_data[cur_index]
+        first_part = ((0b11111110 >> cur_bit) & first_byte) << cur_bit
+        first_part >>= 1
+
+        cur_bit += 7
+        if cur_bit < 8:
+            return first_part
+
+        cur_bit -= 8
+        cur_index += 1
+
+        if cur_index >= len(raw_data):
+            return first_part
+
+        second_byte = raw_data[cur_index]
+        second_part = ((0xFF00 >> cur_bit) & second_byte) & 0xFF
+        second_part >>= 8 - cur_bit
+        return first_part | second_part
+
+    while True:
+        bits = get_7()
+        if bits is False:
+            break
+
+        illegal_index = illegals.index(bits) if bits in illegals else None
+        if illegal_index is not None:
+            next_bits = get_7()
+
+            b1 = 0b11000010
+            b2 = 0b10000000
+            if next_bits is False:
+                b1 |= 7 << 2
+                next_bits = bits
+            else:
+                b1 |= (illegal_index & 0b111) << 2
+
+            first_bit = (next_bits & 0b01000000) > 0
+            b1 |= first_bit
+            b2 |= next_bits & 0b00111111
+            out_data.append(b1)
+            out_data.append(b2)
+        else:
+            out_data.append(bits)
+    return bytes(out_data)
+
+
+def b122decode(base122_data):
+    decoded = []
+    cur_byte = 0
+    bit_of_byte = 0
+
+    def push7(byte):
+        nonlocal cur_byte, bit_of_byte
+        byte <<= 1
+        cur_byte |= (byte >> bit_of_byte)
+        bit_of_byte += 7
+        if bit_of_byte >= 8:
+            decoded.append(cur_byte & 0xFF)
+            bit_of_byte -= 8
+            cur_byte = (byte << (7 - bit_of_byte)) & 0xFF
+
+    for c in base122_data:
+        if c > 127:
+            illegal_index = (c >> 8) & 7  
+            if illegal_index != 7:
+                push7(illegals[illegal_index])
+            push7(c & 127)
+        else:
+            push7(c)
+    return bytes(decoded)
+
+
+if __name__ == "__main__":
+    original_data = b"Hello World!"
+    encoded_data = b122encode(original_data)
+    print(f"Encoded data: {encoded_data}")
+    decoded_data = b122decode(encoded_data)
+    print(f"Decoded data: {decoded_data}")
+
+    original_data = 'Something else 🫨'.encode('utf-8')
+    encoded_data = b122encode(original_data)
+    print(f"Encoded data: {encoded_data.decode('utf8')}")
+    decoded_data = b122decode(encoded_data)
+    print(f"Decoded data: {decoded_data.decode('utf-8')}")