Code simplification: inline CTxInWitness inside CTxIn #8452

1. 
   sipa commented at 2:23 am on August 4, 2016: member

   Based on #8451.

   This is a code change that was suggested early on, but not included in the segwit PR to not invalidate review. It turns out to be simpler than what I had expected, and actually shortens the code.

   Instead of a structure where all CTransaction’s witness data was encapsulated under CTransaction::wit, it is now inlined inside CTransaction::vin[x]. This also reduces memory usage for witness-bearing transactions (as no separate vector for all CTxInWitnesses is needed) and serialization/deserialization time as a result. For non-witness transactions with more than 1 input, it increases memory usage.

   I think it’s neat that it changes the verbose and often-repeated tx.wit.vtxinwit.size() > i ? &tx.wit.vtxinwit[i].scriptWitness : NULL pattern in VerifyScript calls to just &tx.vin[i].scriptWitness.

   This should have no effect on behaviour.

2. Get rid of the const field in CTransaction 232529fbf1
3. 
   NicolasDorier commented at 2:38 am on August 4, 2016: contributor

   wow I want this one! Actually, this is the approach I am using already in NBitcoin, will test soon.
4. 
   dcousens commented at 2:52 am on August 4, 2016: contributor

   concept ACK
5. sipa force-pushed on Aug 4, 2016
6. Move CTxInWitness inside CTxIn 9f5e40d059
7. in src/primitives/transaction.h: in bb6e238662 outdated

   298-};
   299-
   300 struct CMutableTransaction;
   301 
   302+template<typename TxType>
   303+bool HasWitness(const TxType& tx) {
   

   ---

   ---

   

   NicolasDorier commented at 3:18 am on August 4, 2016:

   nit: This is probably better suited as a member method of CTransaction.
8. sipa force-pushed on Aug 4, 2016
9. jonasschnelli added the label Refactoring on Aug 4, 2016
10. 
    NicolasDorier commented at 6:46 am on August 14, 2016: contributor

    utACK the commit 9f5e40d05970b6b82eb5315073c31f70cde9c0d8. I reviewed 232529fbf1af2261085e45407611c32f08d5f6f3 as well, but I don’t know enough in C++ to understand the subtleties of the change.
11. 
    sipa commented at 11:39 am on August 25, 2016: member

    Closing in favor of #8589. If #8580 is rejected I could reopen.
12. sipa closed this on Aug 25, 2016

    ---

13. MarcoFalke locked this on Sep 8, 2021

Contributors
sipa NicolasDorier dcousens

Labels
Refactoring