diff options
| author | Adam Soltys <[email protected]> | 2026-06-29 22:53:17 -0700 |
|---|---|---|
| committer | Adam Soltys <[email protected]> | 2026-06-29 22:53:17 -0700 |
| commit | feef48565032edf8d5d5e3e531db43076d08fe51 (patch) | |
| tree | 326ac97d235db59555fbd55c9140749807b1bee7 /src/api.js | |
| parent | 052e9c714092d8f4a5a9bb29bb85277630c1926d (diff) | |
| download | coinosv3-feef48565032edf8d5d5e3e531db43076d08fe51.tar.gz coinosv3-feef48565032edf8d5d5e3e531db43076d08fe51.zip | |
Wallet: don't understate 'amount sent' for multi-input/swap-funding txs
_txSummary computed a tx's spend only from inputs whose prevout address is on the
scanned frontier (addrMap). A coin spent from an address not currently in that
set — e.g. a fresh reverse-swap claim — was dropped from the total, so the net
showed e.g. -4,768 instead of -50,794, and the scan overwrote applySentTx's
correct net with it. Now: applySentTx counts every input it spent (falling back
to the witness-utxo value if a coin was dropped from the utxo set) and marks the
tx 'local'; the rescan preserves a local tx's net/fee instead of recomputing.
Funds were always correct on-chain — this was a display-only undercount.
Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
Diffstat (limited to 'src/api.js')
0 files changed, 0 insertions, 0 deletions
