Improve canonical ISCC string validation

CHANGELOG.md

@@ -3,6 +3,7 @@
 ## [1.0.8] - Unreleased
 - Import from pydantic v2 lib first
 - Fixed prefix extraction during normalization
+- Improved canonical ISCC string validation
 
 ## [1.0.7] - 2024-01-07
 - Support pydantic v1 & v2

docs/changelog.md

@@ -3,6 +3,7 @@
 ## [1.0.8] - Unreleased
 - Import from pydantic v2 lib first
 - Fixed prefix extraction during normalization
+- Improved canonical ISCC string validation
 
 ## [1.0.7] - 2024-01-07
 - Support pydantic v1 & v2

iscc_core/codec.py

@@ -19,7 +19,7 @@
 def encode_component(mtype, stype, version, bit_length, digest):
     # type: (MainType, SubType, Version, Length, bytes) -> str
     """
-    Encode an ISCC unit inlcuding header and body with standard base32 encoding.
+    Encode an ISCC-UNIT inlcuding header and body with standard base32 encoding.
 
     !!! note
         The `length` value must be the **length in number of bits** for the component.
@@ -67,11 +67,10 @@ def encode_header(mtype, stype, version=0, length=1):
     :rtype: bytes
 
     """
-    # TODO verify that all header params and there combination is valid
     header = bitarray()
     for n in (mtype, stype, version, length):
         header += encode_varnibble(n)
-    # Append zero-padding if required (right side, least significant bits).
+    # Append zero-padding if required (right side, least-significant bits).
     header.fill()
     return header.tobytes()
 
@@ -238,7 +237,7 @@ def encode_length(mtype, length):
         if 0 <= length <= 7:
             return length
         raise ValueError(error)
-    # counter byte lenght encoding
+    # counter byte length encoding
     elif mtype == MT.ID:
         if 64 <= length <= 96:
             return (length - 64) // 8
@@ -503,7 +502,7 @@ def iscc_validate(iscc, strict=True):
     - an ISCC-CODE or ISCC-UNIT
     - encoded with base32 upper without padding
     - has a valid combination of header values
-    - is represented in its canonical URI form
+    - is represented in its canonical form
 
     :param str iscc: ISCC string
     :param bool strict: Raise an exeption if validation fails (default True)
@@ -519,21 +518,42 @@ def iscc_validate(iscc, strict=True):
         else:
             return False
 
+    # Base32 encoding test
+    try:
+        decode_base32(iscc.split(":")[1])
+    except Exception as e:
+        if strict:
+            raise ValueError(e)
+        else:
+            return False
+
     cleaned = iscc_clean(iscc)
+
+    # Prefix test
     prefix = cleaned[:2]
     if prefix not in PREFIXES:
         if strict:
             raise ValueError(f"Header starts with invalid sequence {prefix}")
         else:
             return False
 
+    # Version test
     m, s, v, l, t = decode_header(decode_base32(cleaned))
     if v != 0:
         if strict:
             raise ValueError(f"Unknown version {v} in version header")
         else:
             return False
 
+    # Length test
+    expected_nbyptes = decode_length(m, l).value // 8
+    actual_nbyptes = len(t)
+    if expected_nbyptes != actual_nbyptes:
+        if strict:
+            raise ValueError(f"Header expects {expected_nbyptes} but got {actual_nbyptes} bytes")
+        else:
+            return False
+
     return True
 
 

iscc_core/models.py

@@ -246,7 +246,6 @@ def rnd(cls, mt=None, st=None, bits=64, data=None):
         # Length
         ln_bits = bits or cls.rgen.choice(list(LN)).value
         if mt == MT.ISCC:
-            # TODO fix ramdom ISCC with custom SubType generation
             ln_code = encode_units(units)
         else:
             ln_code = encode_length(mt, bits)
@@ -290,15 +289,6 @@ def mf_base64url(self) -> str:
         """Multiformats base64url encoded."""
         return "u" + encode_base64(self.mc_bytes)
 
-    # TODO: bech32m support
-    # @property
-    # def bech32m(self):
-    #     """Encode as bech32m with hrp `iscc`"""
-    #     data = [bech32.CHARSET.find(c) for c in self.code.lower()]
-    #     return bech32.bech32_encode(
-    #         "iscc", data, bech32.Encoding.BECH32M
-    #     )
-
     def __xor__(self, other) -> int:
         """Use XOR operator for hamming distance calculation."""
         return count_xor(self._body, other._body)

tests/test_code_meta.py

@@ -23,7 +23,6 @@ def test_gen_meta_code_name_only():
 
 
 def test_gen_meta_code_name_and_desc():
-    # TODO handle cases with consecutive newlines and spaces in between
     result = ic.gen_meta_code_v0("Hello  World", "# Some\n\n\n description")
     assert result == {
         "iscc": "ISCC:AAAWN77F72MBZZK3",

tests/test_codec.py

@@ -617,3 +617,23 @@ def test_iscc_code_no_content_code():
 def test_models_Code_rnd_custom_subtype_raises():
     with pytest.raises(ValueError):
         ic.Code.rnd(ic.MT.ISCC, st=ic.ST_ISCC.TEXT)
+
+
+def test_iscc_validate_base_encoding_error():
+    sample = "ISCC:KACT4EBWK27737D2AYCJRAL5Z36G76RFRMO4554RU26HZ4ORJGIVHDI"
+    assert ic.iscc_validate(sample, strict=False) is True
+
+    assert ic.iscc_validate(sample[:-1], strict=False) is False
+
+    with pytest.raises(ValueError) as excinfo:
+        ic.iscc_validate(sample[:-1], strict=True)
+    assert str(excinfo.value) == "Incorrect padding"
+
+
+def test_iscc_validate_bad_length():
+    sample = "ISCC:KACT4EBWK27737D2AYCJRAL5Z36G76RFRMO4554RU26HZ"
+    assert ic.iscc_validate(sample, strict=False) is False
+
+    with pytest.raises(ValueError) as excinfo:
+        ic.iscc_validate(sample, strict=True)
+    assert str(excinfo.value) == "Header expects 32 but got 26 bytes"

tests/test_iscc_code.py

@@ -41,7 +41,6 @@ def test_gen_iscc_code_v0_no_meta():
 def test_gen_iscc_code_v0_no_meta_content():
     icode = ic.gen_iscc_code_v0([DID_128, IID_256])
     assert icode == {"iscc": "ISCC:KUAFVC5DMJJGYKZ4MQV6B6THIBBUG"}
-    # TODO mabye show length for SubType SUM as we now the unit composition.
     # we may also get a ISCC-SUM-V0-256 version
     assert ic.iscc_explain(icode["iscc"]) == "ISCC-SUM-V0-DI-5a8ba362526c2b3c642be0fa67404343"