bip93: reject HRPs longer than 83 characters

src/codex32/bech32.py

@@ -88,6 +88,8 @@ def bech32_decode(value: str, spec: _Checksum | None = None) -> tuple[str, list[
         raise MissingSeparator("No separator (1) was found.")
     if separator == 0:
         raise MissingSeparator("The application prefix before 1 is missing.")
+    if separator > 83:
+        raise InvalidLength(f"human-readable part exceeds 83 characters ({separator})")
     lowered = value.lower()
     hrp = lowered[:separator]
     data = _chars_to_u5(lowered[separator + 1 :], separator + 2)

src/codex32/correction.py

@@ -768,6 +768,8 @@ def _validate_context(context: CorrectionContext) -> None:
         _validate_single_case_ascii(context.hrp)
         if context.hrp.lower() != context.hrp:
             raise ValueError("hrp must be a normalized application prefix")
+        if len(context.hrp) > 83:
+            raise ValueError(f"human-readable part exceeds 83 characters ({len(context.hrp)})")
         length = context.expected_length
         if length is not None:
             if isinstance(length, bool) or not isinstance(length, int):

tests/data/bip93_vectors.py

@@ -79,6 +79,31 @@
     "codex32_peev": "cl10peevst6cqh0wu7p5ssjyf4z4ez42ks9jlt3zneju9uuypr2hddak6tlqsjhsks4laxts8q",
 }
 
+# BIP-93 generalized-HRP draft, BenWestgate/bips PR #2 @ 01374bf.
+BIP93_GENERAL_HRP = {
+    "share_a": "test_vector12spana320zyxwvutsrqpnmlkjhgfedca320zyxwvutsrqpnmlkjhgfedca304ppsqh4l7v3dh",
+    "share_c": "test_vector12spancacdefghjklmnpqrstuvwxyz023acdefghjklmnpqrstuvwxyz023jxmjy7q9xl7d3ul",
+    "derived_d": "test_vector12spandll4f8jlh4e5vdvuldlfxu2jhdnll4f8jlh4e5vdvuldlfxu2jhdnx3pe3yqzdl7p30y",
+    "secret_s": "test_vector12spans6xqguzttxkeqnjsjzv4jv3nz5k6xqguzttxkeqnjsjzv4jv3nz5kdf37v5qm2l7l3az",
+    "valid_83": (
+        "a83characterlongcodex32humanreadablepartforsecretsharewiththedigit1andthelettersbio"
+        "10keyss9mfjjk25y05e3nq"
+    ),
+    "invalid_gap_regular": (
+        "test_vector12spans6xqguzttxkeqnjsjzv4jv3nz5k6xqguzttxkeqnjsjzv4jv3nz5kp0f0tlr2cxh0t"
+    ),
+    "invalid_gap_long": (
+        "test_vector12spans6xqguzttxkeqnjsjzv4jv3nz5k6xqguzttxkeqnjsjzv4jv3nz5c2m7mxmk8s5qecp"
+    ),
+    "invalid_uppercase_hrp_checksum": (
+        "CL10PEEVST6CQH0WU7P5SSJYF4Z4EZ42KS9JLT3ZNEJU9UUYPR2HDDAK6TLQS3J5AYX4Y08079"
+    ),
+    "invalid_84": (
+        "a84characterslongcodex32humanreadablepartforsecretsharewiththedigit1andthelettersbio"
+        "10keyss9hqxu3pxsnkdkth"
+    ),
+}
+
 
 VALID_CODEX32 = [
     "A12UEL5LLGCHJ4UJCQVHG",
@@ -93,7 +118,6 @@
 VALID_CODEX32_LONG = [
     "A12UEL5LQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQQV3RR8ZLCK96GTC3",
     "a12uel5lqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqv3rr8zlck96gtc3",
-    "a1002characterlonghumanreadablepartthatcontainsthenumber1,theexcludedcharactersbio,andeveryus-asciicharacterin[33-126]!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~15ttgtscr3gvktxamm8mzt",
     "abcdef12l7aum6echk45nj3s0wdvt2fg8x9yrzpql7aum6echk45nj3s0wdvt2fg8x9yrzpql7aum6echk45nj3s0wdvt2fg8x9yrzpqp9evrmhc52umqew",
     "1177777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777777fn0jxg9gc35xwa8",
     "split13checkupstagehandshakeupstreamerranterredcaperredscatteredsusurrantplunderedqsp5ws8r2klm66l",
@@ -117,8 +141,6 @@
     + "12fauxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx836hdd09mhkhkhx",  # HRP character out of range
     "\x80"
     + "12fauxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxlhf5ywnkmk4r3tc",  # HRP character out of range
-    # overall max length exceeded
-    "a1003characterslonghumanreadablepartthatcontainsthenumber1,theexcludedcharactersbio,andeveryus-asciicharacterin[33-126]!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~!\"#$%&'()*+,-./0123456789:;<=>?@[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~12fauxxru38cppmlpu0t6l",
     "y12bfauxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxt3y5fewy4gnw2hs",  # Invalid data character
     "lt12ifauxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxjxd0ehq868vm3zl",  # Invalid data character
     "in12fauxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxvgegljrsvs5w9q",  # Too short checksum

tests/test_correction_bch.py

@@ -217,6 +217,7 @@ def test_public_context_constrains_length_prefix_and_used_indices() -> None:
         CorrectionContext(Profile.MS, excluded_indices=["a"]),  # type: ignore[arg-type]
         CorrectionContext(Profile.MS, excluded_indices=("s",)),
         CorrectionContext(Profile.MS, excluded_indices=("a", "A")),
+        CorrectionContext("z" * 84, expected_length=106),
     ),
 )
 def test_malformed_public_context_is_rejected(context: CorrectionContext) -> None:

tests/test_generic_hrp.py

@@ -7,6 +7,7 @@
 
 import pytest
 from _codex32_oracle import oracle_encode
+from data.bip93_vectors import BIP93_GENERAL_HRP
 
 from codex32 import (
     CorrectionContext,
@@ -18,7 +19,7 @@
     recover_secret,
 )
 from codex32.cli import main, ms_main
-from codex32.errors import MismatchedHrp, MismatchedProfile
+from codex32.errors import InvalidChecksum, InvalidLength, MismatchedHrp, MismatchedProfile
 from tools._wallet_test_vectors import STUB_FINGERPRINT
 
 UNKNOWN = {
@@ -178,3 +179,33 @@ def fingerprint(_secret: object) -> bytes:
     status, output, error = _invoke(main, ["share", "d", "--plain"], basis)
     assert (status, output.strip(), error) == (0, UNKNOWN["short"]["D"], "")
     assert _invoke(ms_main, ["share", "d", "--plain"], basis)[0] == 2
+
+
+def test_bip93_generalized_hrp_share_vectors() -> None:
+    share_a = parse_codex32(BIP93_GENERAL_HRP["share_a"])
+    share_c = parse_codex32(BIP93_GENERAL_HRP["share_c"])
+    assert share_a.hrp == share_c.hrp == "test_vector"
+    assert recover_secret([share_a, share_c]).text == BIP93_GENERAL_HRP["secret_s"]
+    assert derive_share([share_a, share_c], "d").text == BIP93_GENERAL_HRP["derived_d"]
+
+
+def test_bip93_generalized_hrp_boundary_vectors() -> None:
+    valid = BIP93_GENERAL_HRP["valid_83"]
+    invalid = BIP93_GENERAL_HRP["invalid_84"]
+    artifact = parse_codex32(valid)
+
+    assert len(artifact.hrp) == 83
+    assert artifact.hrp == valid.rsplit("1", 1)[0]
+    with pytest.raises(InvalidLength):
+        parse_codex32(invalid)
+
+
+@pytest.mark.parametrize("key", ("invalid_gap_regular", "invalid_gap_long"))
+def test_bip93_generalized_hrp_gap_vectors_are_invalid(key: str) -> None:
+    with pytest.raises(InvalidLength):
+        parse_codex32(BIP93_GENERAL_HRP[key])
+
+
+def test_bip93_generalized_hrp_checksum_uses_lowercase_hrp() -> None:
+    with pytest.raises(InvalidChecksum):
+        parse_codex32(BIP93_GENERAL_HRP["invalid_uppercase_hrp_checksum"])