bugfix: fixed incorrect bytestring encoding PlutusData (#269)

* bugfix: fixed incorrect bytestring encoding for bytestring longer than 64 bytes

* Removed extraneous test encoder

* Removed metadata dummy class

* Added dummy ByteString class

* Added ByteString equality test for bytes

* Updated test reference hash

* Updated byte encoding errors, added ByteString to as valid PlutusData type

* Removed debug print statements

* Fixed mypy error for equality checks on ByteString

Author: theeldermillenial

pycardano/metadata.py

@@ -41,6 +41,7 @@ def _validate_type_and_size(data):
                 if len(data) > self.MAX_ITEM_SIZE:
                     raise InvalidArgumentException(
                         f"The size of {data} exceeds {self.MAX_ITEM_SIZE} bytes."
+                        "Use pycardano.serialization.ByteString for long bytes."
                     )
             elif isinstance(data, str):
                 if len(data.encode("utf-8")) > self.MAX_ITEM_SIZE:

pycardano/plutus.py

@@ -14,11 +14,12 @@
 from nacl.encoding import RawEncoder
 from nacl.hash import blake2b
 
-from pycardano.exception import DeserializeException
+from pycardano.exception import DeserializeException, InvalidArgumentException
 from pycardano.hash import DATUM_HASH_SIZE, SCRIPT_HASH_SIZE, DatumHash, ScriptHash
 from pycardano.nativescript import NativeScript
 from pycardano.serialization import (
     ArrayCBORSerializable,
+    ByteString,
     CBORSerializable,
     DictCBORSerializable,
     IndefiniteList,
@@ -468,6 +469,8 @@ class will reduce the complexity of serialization and deserialization tremendous
         >>> assert test == Test.from_cbor("d87a9f187b43333231ff")
     """
 
+    MAX_BYTES_SIZE = 64
+
     @classproperty
     def CONSTR_ID(cls):
         """
@@ -489,13 +492,20 @@ def CONSTR_ID(cls):
         return getattr(cls, k)
 
     def __post_init__(self):
-        valid_types = (PlutusData, dict, IndefiniteList, int, bytes)
+        valid_types = (PlutusData, dict, IndefiniteList, int, ByteString, bytes)
         for f in fields(self):
             if inspect.isclass(f.type) and not issubclass(f.type, valid_types):
                 raise TypeError(
                     f"Invalid field type: {f.type}. A field in PlutusData should be one of {valid_types}"
                 )
 
+            data = getattr(self, f.name)
+            if isinstance(data, bytes) and len(data) > 64:
+                raise InvalidArgumentException(
+                    f"The size of {data} exceeds {self.MAX_BYTES_SIZE} bytes. "
+                    "Use pycardano.serialization.ByteString for long bytes."
+                )
+
     def to_shallow_primitive(self) -> CBORTag:
         primitives: Primitive = super().to_shallow_primitive()
         if primitives:
@@ -553,6 +563,8 @@ def _dfs(obj):
                 return {"int": obj}
             elif isinstance(obj, bytes):
                 return {"bytes": obj.hex()}
+            elif isinstance(obj, ByteString):
+                return {"bytes": obj.value.hex()}
             elif isinstance(obj, IndefiniteList) or isinstance(obj, list):
                 return {"list": [_dfs(item) for item in obj]}
             elif isinstance(obj, dict):
@@ -667,7 +679,10 @@ def _dfs(obj):
                 elif "int" in obj:
                     return obj["int"]
                 elif "bytes" in obj:
-                    return bytes.fromhex(obj["bytes"])
+                    if len(obj["bytes"]) > 64:
+                        return ByteString(bytes.fromhex(obj["bytes"]))
+                    else:
+                        return bytes.fromhex(obj["bytes"])
                 elif "list" in obj:
                     return IndefiniteList([_dfs(item) for item in obj["list"]])
                 else:

pycardano/serialization.py

@@ -60,6 +60,22 @@ class IndefiniteFrozenList(FrozenList, IndefiniteList):  # type: ignore
     pass
 
 
+@dataclass
+class ByteString:
+    value: bytes
+
+    def __hash__(self):
+        return hash(self.value)
+
+    def __eq__(self, other: object):
+        if isinstance(other, ByteString):
+            return self.value == other.value
+        elif isinstance(other, bytes):
+            return self.value == other
+        else:
+            return False
+
+
 @dataclass
 class RawCBOR:
     """A wrapper class for bytes that represents a CBOR value."""
@@ -160,6 +176,7 @@ def default_encoder(
     assert isinstance(
         value,
         (
+            ByteString,
             CBORSerializable,
             IndefiniteList,
             RawCBOR,
@@ -178,6 +195,15 @@ def default_encoder(
         for item in value:
             encoder.encode(item)
         encoder.write(b"\xff")
+    elif isinstance(value, ByteString):
+        if len(value.value) > 64:
+            encoder.write(b"\x5f")
+            for i in range(0, len(value.value), 64):
+                imax = min(i + 64, len(value.value))
+                encoder.encode(value.value[i:imax])
+            encoder.write(b"\xff")
+        else:
+            encoder.encode(value.value)
     elif isinstance(value, RawCBOR):
         encoder.write(value.cbor)
     elif isinstance(value, FrozenList):

test/pycardano/test_plutus.py

@@ -19,7 +19,7 @@
     RedeemerTag,
     plutus_script_hash,
 )
-from pycardano.serialization import IndefiniteList
+from pycardano.serialization import ByteString, IndefiniteList
 
 
 @dataclass
@@ -396,3 +396,25 @@ class A(PlutusData):
     assert (
         res == res2
     ), "Same class has different default constructor id in two consecutive runs"
+
+
+def test_plutus_data_long_bytes():
+    @dataclass
+    class A(PlutusData):
+        a: ByteString
+
+    quote = (
+        "The line separating good and evil passes ... right through every human heart."
+    )
+
+    quote_hex = (
+        "d866821a8e5890cf9f5f5840546865206c696e652073657061726174696e6720676f6f6420616"
+        "e64206576696c20706173736573202e2e2e207269676874207468726f7567682065766572794d"
+        "2068756d616e2068656172742effff"
+    )
+
+    A_tmp = A(ByteString(quote.encode()))
+
+    assert (
+        A_tmp.to_cbor_hex() == quote_hex
+    ), "Long metadata bytestring is encoded incorrectly."