(based on #5181 as it would otherwise conflict and I'm lazy.)
Adds RPC calls to generate (and verify) merkle blocks which, though verify is less useful for Bitcoin Core, just making them more accessible so that (hopefully) they become useful for people to exchange around would be very nice.
Full Disclosure: This is required to sidechain-related work.
Oh, also a simple flag to getblock so you can get only the header. That one is not related to sidechains, but I needed it while debugging.
3 | @@ -4,15 +4,17 @@ 4 | 5 | #include "base58.h" 6 | #include "clientversion.h" 7 | +#include "core/block.h" // for MAX_BLOCK_SIZE 8 | #include "core/transaction.h" 9 | #include "core_io.h" 10 | +#include "coins.h"
coins should be before core ;)
13 | #include "script/script.h"
14 | #include "script/sign.h"
15 | #include "ui_interface.h" // for _(...)
16 | #include "univalue/univalue.h"
17 | #include "util.h"
18 | +#include "utilstrencodings.h"
utilstrencodings should be below utilmoneystr
0 | @@ -0,0 +1,177 @@ 1 | +// Copyright (c) 2009-2010 Satoshi Nakamoto 2 | +// Copyright (c) 2009-2014 The Bitcoin developers 3 | +// Distributed under the MIT software license, see the accompanying 4 | +// file COPYING or http://www.opensource.org/licenses/mit-license.php. 5 | + 6 | +#include "merkleblock.h" 7 | + 8 | +#include "hash.h"
core before hash
62 | + 63 | +uint256 CPartialMerkleTree::CalcHash(int height, unsigned int pos, const std::vector<uint256> &vTxid) { 64 | + if (height == 0) { 65 | + // hash at height 0 is the txids themself 66 | + return vTxid[pos]; 67 | + } else {
Woulnd't need an else block, right?
That is #5181 and is code-move-only, so I dont want to change it.
Right, but most of the time these nits or improvement suggestions are not catched by anyone later on anyway ;).
0 | @@ -0,0 +1,150 @@ 1 | +// Copyright (c) 2009-2010 Satoshi Nakamoto 2 | +// Copyright (c) 2009-2014 The Bitcoin developers 3 | +// Distributed under the MIT software license, see the accompanying 4 | +// file COPYING or http://www.opensource.org/licenses/mit-license.php. 5 | + 6 | +#ifndef BITCOIN_MERKLEBLOCK_H 7 | +#define BITCOIN_MERKLEBLOCK_H 8 | + 9 | +#include "serialize.h"
Same here, ordering.
193 | @@ -193,6 +194,112 @@ Value getrawtransaction(const Array& params, bool fHelp) 194 | return result; 195 | } 196 | 197 | +Value gettxoutproof(const Array& params, bool fHelp) 198 | +{ 199 | + if (fHelp || (params.size() != 1 && params.size() != 2)) 200 | + throw runtime_error( 201 | + "gettxoutproof \"txid\" ( block hash )\n"
This looks like we require 2 arguments here, block and hash, but its only 1 blockhash, so I would remove the space between the two.
Done.
Fixed all the nits that we're related to header-include (I believe we removed that from our coding standards?)
@TheBlueMatt You as a core dev, I really would love to see you still following that rules :).
Fixed include ordering for includes related to this pull (others are related to #5181)
Rebased on top of the now-merged #5181, also now support multiple transactions in one proof.
Concept ACK, but this new functionality needs a test in qa/rpc-tests.
@laanwj done.
258 | "\nIf verbose is false, returns a string that is serialized, hex-encoded data for block 'hash'.\n"
259 | "If verbose is true, returns an Object with information about block <hash>.\n"
260 | + "If header only is true (and verbose is false), limits the output to only the block header.\n"
261 | "\nArguments:\n"
262 | "1. \"hash\" (string, required) The block hash\n"
263 | "2. verbose (boolean, optional, default=true) true for a json object, false for the hex encoded data\n"
headeronly (boolean) would need to be mentioned in the arguments here.
BTW: what about adding a getblockheader call, instead of adding a boolean argument? I'd slightly prefer that for readability/simplicity. It would only result in minimal code duplication.
ping @thebluematt
Rebased, fixed an indentation error, removed the blockheader-only option (meh, people can just substring(0, 160) anyway...)
Rebased. Review?
Tested ACK.
utACK