fix: p2pkh_lock accepts 20-byte hash regardless of string encoding (#389)

Callers passing `"\x00" * 20` (UTF-8 encoding) were routed to the
Base58Check decode path, which failed on null bytes. A 20-byte string
can never be a valid Base58Check address (minimum ~34 chars), so
bytesize alone is sufficient to disambiguate.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: sgbett

gem/bsv-sdk/lib/bsv/script/script.rb

@@ -171,8 +171,11 @@ def self.p2pkh_lock(pubkey_hash_or_address)
       # Resolve a pubkey_hash argument that may be a raw binary hash or a
       # Base58Check address string.
       def self.resolve_pubkey_hash(arg)
-        # A 20-byte ASCII-8BIT string is treated as a raw binary hash.
-        return arg if arg.encoding == Encoding::ASCII_8BIT && arg.bytesize == 20
+        # A 20-byte string is treated as a raw binary hash regardless of
+        # encoding — callers commonly pass `"\x00" * 20` which is UTF-8.
+        # A valid Base58Check address is always longer than 20 characters
+        # (25 raw bytes → ~34 Base58 characters), so this is unambiguous.
+        return arg.b if arg.bytesize == 20
 
         # Otherwise treat as a Base58Check address string.
         decoded = BSV::Primitives::Base58.check_decode(arg, prefix_length: 1)