fix(hashes): evaluate adler32 over bytes to resolve non-ASCII divergence

hashes/adler32.py

@@ -11,20 +11,56 @@
 MOD_ADLER = 65521
 
 
-def adler32(plain_text: str) -> int:
+def adler32(plain_text: str | bytes) -> int:
     """
     Function implements adler-32 hash.
-    Iterates and evaluates a new value for each character
+    Iterates and evaluates a new value for each byte in the data.
+    If a string is passed, it is encoded to UTF-8 bytes per the Adler-32 specification.
 
     >>> adler32('Algorithms')
     363791387
 
     >>> adler32('go adler em all')
     708642122
+
+    >>> adler32('€100')
+    196018850
+
+    >>> adler32(b'Algorithms')
+    363791387
+
+    >>> adler32(b'')
+    1
+
+    >>> adler32('')
+    1
+
+    >>> import zlib
+    >>> adler32('日本語テスト') == zlib.adler32('日本語テスト'.encode())
+    True
+
+    >>> adler32(b'\\x00\\xff\\xfe') == zlib.adler32(b'\\x00\\xff\\xfe')
+    True
+
+    >>> adler32(123)
+    Traceback (most recent call last):
+        ...
+    TypeError: plain_text must be str or bytes-like, not int
     """
+    if not isinstance(plain_text, (str, bytes, bytearray)):
+        msg = f"plain_text must be str or bytes-like, not {type(plain_text).__name__}"
+        raise TypeError(msg)
+
+    data = plain_text.encode("utf-8") if isinstance(plain_text, str) else plain_text
     a = 1
     b = 0
-    for plain_chr in plain_text:
-        a = (a + ord(plain_chr)) % MOD_ADLER
+    for byte in data:
+        a = (a + byte) % MOD_ADLER
         b = (b + a) % MOD_ADLER
     return (b << 16) | a
+
+
+if __name__ == "__main__":
+    import doctest
+
+    doctest.testmod()