Python JSON duplicate names: reject ambiguous object fields

Operation contract

A signed instruction should not let two parsers disagree about the tenant field. object_pairs_hook receives ordered pairs for each object; the hook rejects a repeated name before producing a dict. This is parser-level validation. It does not authenticate the payload, enforce its complete schema, or reject a different spelling of the same logical identifier after case normalization.

Failure and ownership boundary

The default decoder accepts duplicate names and keeps a later value. Applying a strict hook to every object is simple but costs a set per object. Set a byte limit before parsing untrusted input. Keep the original bytes for signature verification when the protocol signs bytes; reserializing a parsed dict is not equivalent to those original bytes.

Working program

import json

def unique_names(pairs):
    result = {}
    for name, value in pairs:
        if name in result:
            raise ValueError("duplicate name: " + name)
        result[name] = value
    return result

payload = '{"tenant":"north-47","tenant":"south-47"}'
try:
    json.loads(payload, object_pairs_hook=unique_names)
except ValueError as error:
    print(str(error))

Output

duplicate name: tenant

Costs and limits

The hook visits each key once and retains the decoded object. A duplicate check is O(keys) expected time per object with O(keys) extra membership state.

Common Mistakes

• Do not rely on a default decoder to reject duplicate names.
• A parsed object is not the byte sequence that a signature covered.
• Case folding or Unicode normalization may need a separate field-name rule.

Connected lessons

• Python JSON validation: reject duplicate members and non-integer amounts
• Python JSON nonfinite numbers: reject NaN on input and output
• Python HMAC envelopes: authenticate exact bytes and separate replay policy

Test this boundary.