Nicholas Dudfield 63bedc2d06 refactor(json-tx): split hasBody from empty-body rejection ...

hasBody was conflating 'field is declared' with 'field is non-empty'.
a classical tx could ride along with a spurious empty sfJsonTxBody and
silently fall through to the classical sig path.

now:
* hasBody = sfJsonTxBody is present (pure routing predicate)
* checkSignature / checkStructuralEquivalence reject an empty body up
  front with an explicit 'JsonTxBody is empty.' error string

behaviour for well-formed json-tx and well-formed classical-tx is
unchanged; the classical+empty-body edge case now fails cleanly instead
of silently routing classical.

2026-04-24 21:12:34 +07:00

..

xrpl

refactor(json-tx): split hasBody from empty-body rejection

2026-04-24 21:12:34 +07:00