fuzz-target-payload-codec: the §3.1 target — all five invariants in shared, thin fuzz_target wrapper, 97 committed seeds + gitignore negation + README row; placeholder bin replaced; serde (derive) added to shared

fuzz/gen_fuzz_seeds.py

@@ -24,8 +24,108 @@ def write_seed(target, i, data):
         f.write(data)
 
 
+def payload_codec_seeds():
+    i = 0
+
+    # Valid JSON of every Value variant (numbers incl. bounds/floats,
+    # strings incl. escapes/unicode, arrays, objects), plus whitespace
+    # padding.
+    valid_docs = [
+        b"null",
+        b"true",
+        b"false",
+        b"0",
+        b"-1",
+        b"-9223372036854775808",
+        b"9223372036854775807",
+        b"18446744073709551615",
+        b"18446744073709551616",
+        b"3.5",
+        b"-2.25",
+        b"1e10",
+        b"1e309",
+        b"-0",
+        b'""',
+        b'"hello world"',
+        b'"unicode \\u00e9"',
+        b'"emoji \\ud83d\\ude00"',
+        b'"tab\\tnewline\\nquote\\"slash"',
+        b'"null"',
+        b"[]",
+        b'[1, "two", null, true, {"three": 3}, []]',
+        b"{}",
+        b'{"a": 1}',
+        b'{"a": 1, "nested": {"deep": [true, null, 3.5]}}',
+        b'{"\\u00e9": "value"}',
+        b'  \t\r\n{"spaced": true}\t\n  ',
+    ]
+    for doc in valid_docs:
+        write_seed("payload_codec", i, doc)
+        i += 1
+
+    # Truncations at every prefix length of two representative docs.
+    for doc in [b'{"a": 1, "b": [2, 3]}', b"[true, null, 1.5]"]:
+        for n in range(1, len(doc)):
+            write_seed("payload_codec", i, doc[:n])
+            i += 1
+
+    # Deep nesting: shallow, at/near serde_json's 128 recursion limit,
+    # and well past it. Arrays and objects both.
+    for depth in [8, 127, 128, 129, 256, 1024]:
+        write_seed("payload_codec", i, ("[" * depth + "0" + "]" * depth).encode())
+        i += 1
+    for depth in [8, 200]:
+        write_seed("payload_codec", i, ('{"a":' * depth + "1" + "}" * depth).encode())
+        i += 1
+
+    # Non-UTF-8 bytes inside and outside string positions.
+    for raw in [b"\xff\xfe\xfd", b'{"a": "\xff"}', b"[1, 2\xff]"]:
+        write_seed("payload_codec", i, raw)
+        i += 1
+
+    # Trailing garbage after a valid document.
+    for raw in [
+        b"{} trailing",
+        b"[1]x",
+        b"1 2",
+        b"null null",
+        b'"first""second"',
+    ]:
+        write_seed("payload_codec", i, raw)
+        i += 1
+
+    # Duplicate keys (canonicalization is serde_json's).
+    for raw in [b'{"a": 1, "a": 2}', b'{"o": {"k": 1, "k": 2}, "o": 3}']:
+        write_seed("payload_codec", i, raw)
+        i += 1
+
+    # The Null-decode shape: an empty byte store must never decode to
+    # a silent Null (ADR-023 §1).
+    for raw in [b"", b" ", b"null", b"nullx"]:
+        write_seed("payload_codec", i, raw)
+        i += 1
+
+    # Other loud-fail shapes (typed `Codec` partition coverage).
+    for raw in [
+        b"NaN",
+        b"Infinity",
+        b"+1",
+        b"01",
+        b".5",
+        b"1.",
+        b"--1",
+        b"{,}",
+        b"[1,]",
+        b'{"a"}',
+        b'{"missing": }',
+        b"\x80\x81\x82",
+    ]:
+        write_seed("payload_codec", i, raw)
+        i += 1
+
+
 def main():
-    pass
+    payload_codec_seeds()
 
 
 if __name__ == "__main__":