Fix CallDecoder type hints/docs for "None" results

[Pratham6392]

This issue improves SDK correctness at the typing/documentation layer: no protocol change, but it prevents avoidable user crashes caused by misleading non-optional hints on outputs that are intentionally optional at runtime

Issue

CallDecoder decode methods in hypersync/init.py return lists where each element can be either:

• decoded values (list[DecodedSolValue])
• or None when calldata does not match loaded function signatures

However, several Python return annotations/documentation currently imply non-optional nested lists, which misleads users into unsafe indexing.

This is a Python API contract/documentation correctness issue .

User-facing impact

out = decoder.decode_inputs_sync(inputs)
print(out[0][0].val)

This crashes when out[0] is None:
TypeError: 'NoneType' object is not subscriptable

The crash is expected runtime behavior for unmatched selectors; the issue is that current type hints/docstrings don’t clearly communicate this.

Scope of fix

Expected return annotation shape

list[Optional[list[DecodedSolValue]]]

If we add this return type in the function then user gets the idea to handle None condition also .
Example : def decode_inputs_sync(self, inputs: list[str]) -> list[Optional[list[DecodedSolValue]]]:

Update Python wrapper type hints + docstrings in hypersync/init.py for CallDecoder methods so API contract reflects actual behavior.

Methods that should be optional-aware

• decode_inputs
• decode_inputs_sync
• decode_transactions_input
• decode_transactions_input_sync
• decode_traces_input
• decode_traces_input_sync