以太坊CPP代码分析（二）

其他 2018-05-06 22:44:32 阅读次数: 11

1.基本概念

SHA-3哈希加密

      Ethereum 代码里哈希（hash)无处不在，许许多多的类型对象通过给定的哈希算法，可以得到一个哈希值。注意，算法中所使用的哈希函数是不可逆的，即对于h = hash(x), 仅仅通过哈希运算的结果h 无法作逆运算得到输入x。哈希值在数学上的唯一性使得它可以用作某个对象的全局唯一标识符。

    Ethereum 中用到的哈希函数全部采用SHA-3(Secure Hash Algorithm 3，wikipedia)。SHA-3在2015年8月由美国标准技术协会(NIST)正式发布，作为Secure Hash Algorithm家族的最新一代标准，它相比于SHA-2和SHA-1，采用了完全不同的设计思路，性能也比较好。需要注意的是，SHA-2目前并没有出现被成功攻克的案例，SHA-3也没有要立即取代SHA-2的趋势，NIST只是考虑到SHA-1有过被攻克的案例，未雨绸缪的征选了采用全新结构和思路的SHA-3来作为一种最新的SHA方案。

RLP编码

    RLP（Recursive Length Prefix）编码，其定义可见wiki，它可以将一个任意嵌套的字节数组([]byte)，编码成一个“展平”无嵌套的[]byte。1 byte取值范围0x00 ～ 0xff，可以表示任意字符，所以[]byte可以线性的表示任意的数据。最简单比如一个字符串，如果每个字符用ASCII码的二进制表示，整个字符串就变成一个[]byte。 RLP 编码其实提供了一种序列化的编码方法，无论输入是何种嵌套形式的元素或数组，编码输出形式都是[]byte。RLP是可逆的，它提供了互逆的编码、解码方法。

    Ethereum 中具体使用的哈希算法，就是对某个类型对象的RLP编码值做了SHA3哈希运算，可称为RLP Hash。 Ethereum 在底层存储中特意选择了专门存储和读取[k, v] 键值对的第三方数据库，[k, v] 中的v 就是某个结构体对象的RLP编码值([]byte)，k大多数情况就是v的RLP编码后的SHA-3哈希值。


2. Block和Header

Block（区块）是Ethereum的核心数据结构之一。所有账户的相关活动，以交易(Transaction)的格式存储，每个Block有一个交易对象的列表；每个交易的执行结果，由一个Receipt对象与其包含的一组Log对象记录；所有交易执行完后生成的Receipt列表，存储在Block中(经过压缩加密)。不同Block之间，通过前向指针ParentHash一个一个串联起来成为一个单向链表，BlockChain 结构体管理着这个链表。

以太坊的主要数据结构


2.1BlockHeader

BlockHeader是Block的核心,Header的成员变量全都很重要，值得细细理解:

  • m_parentHash：指向父区块(parentBlock)的指针。除了创世块(Genesis Block)外，每个区块有且只有一个父区块。
  • m_author：挖掘出这个区块的作者地址。在每次执行交易时系统会给与一定补偿的Ether，这笔金额就是发给这个地址的。
  • m_sha3Uncles:Block结构体的成员uncles的RLP哈希值。uncles是一个Header数组，它的存在，颇具匠心。
  • m_stateRoot：StateDB中的“state Trie”的根节点的RLP哈希值。Block中，每个账户以stateObject对象表示，账户以Address为唯一标示，其信息在相关交易(Transaction)的执行中被修改。所有账户对象可以逐个插入一个Merkle-PatricaTrie(MPT)结构里，形成“state Trie”。
  • m_transactionsRoot: Block中 “tx Trie”的根节点的RLP哈希值。Block的成员变量transactions中所有的tx对象，被逐个插入一个MPT结构，形成“tx Trie”。
  • m_receiptsRoot：Block中的 "Receipt Trie”的根节点的RLP哈希值。Block的所有Transaction执行完后会生成一个Receipt数组，这个数组中的所有Receipt被逐个插入一个MPT结构中，形成"Receipt Trie"。
  • m_logBloom：Bloom过滤器(Filter)，用来快速判断一个参数Log对象是否存在于一组已知的Log集合中。
  • m_difficulty：区块的难度。Block的Difficulty由共识算法基于parentBlock的Time和Difficulty计算得出，它会应用在区块的‘挖掘’阶段。
  • m_number：区块的序号。Block的Number等于其父区块Number +1。
  • m_timestamp：区块“应该”被创建的时间。由共识算法确定，一般来说，要么等于parentBlock.Time + 10s，要么等于当前系统时间。
  • m_gasLimit：区块内所有Gas消耗的理论上限。该数值在区块创建时设置，与父区块有关。具体来说，根据父区块的GasUsed同GasLimit * 2/3的大小关系来计算得出。
  • m_gasUsed：区块内所有Transaction执行时所实际消耗的Gas总和。
  • Nonce：一个64bit的哈希数，它被应用在区块的"挖掘"阶段，并且在使用中会被修改。在C++版本无nonce

  代码中写道：To determine the header hash without the nonce (for sealing), the method hash(WithoutNonce) is
 * provided.

class BlockHeader
{
	friend class BlockChain;
public:
	static const unsigned BasicFields = 13;

	BlockHeader();
	explicit BlockHeader(bytesConstRef _data, BlockDataType _bdt = BlockData, h256 const& _hashWith = h256());
	explicit BlockHeader(bytes const& _data, BlockDataType _bdt = BlockData, h256 const& _hashWith = h256()): BlockHeader(&_data, _bdt, _hashWith) {}
	BlockHeader(BlockHeader const& _other);
	BlockHeader& operator=(BlockHeader const& _other);

	static h256 headerHashFromBlock(bytes const& _block) { return headerHashFromBlock(&_block); }
	static h256 headerHashFromBlock(bytesConstRef _block);
	static RLP extractHeader(bytesConstRef _block);

	explicit operator bool() const { return m_timestamp != Invalid256; }

	bool operator==(BlockHeader const& _cmp) const
	{
		return m_parentHash == _cmp.parentHash() &&
			m_sha3Uncles == _cmp.sha3Uncles() &&
			m_author == _cmp.author() &&
			m_stateRoot == _cmp.stateRoot() &&
			m_transactionsRoot == _cmp.transactionsRoot() &&
			m_receiptsRoot == _cmp.receiptsRoot() &&
			m_logBloom == _cmp.logBloom() &&
			m_difficulty == _cmp.difficulty() &&
			m_number == _cmp.number() &&
			m_gasLimit == _cmp.gasLimit() &&
			m_gasUsed == _cmp.gasUsed() &&
			m_timestamp == _cmp.timestamp() &&
			m_extraData == _cmp.extraData();
	}
	bool operator!=(BlockHeader const& _cmp) const { return !operator==(_cmp); }

	void clear();
	void noteDirty() const { Guard l(m_hashLock); m_hashWithout = m_hash = h256(); }
	void populateFromParent(BlockHeader const& parent);

	// TODO: pull out into abstract class Verifier.
	void verify(Strictness _s = CheckEverything, BlockHeader const& _parent = BlockHeader(), bytesConstRef _block = bytesConstRef()) const;
	void verify(Strictness _s, bytesConstRef _block) const { verify(_s, BlockHeader(), _block); }

	h256 hash(IncludeSeal _i = WithSeal) const;
	void streamRLP(RLPStream& _s, IncludeSeal _i = WithSeal) const;

	void setParentHash(h256 const& _v) { m_parentHash = _v; noteDirty(); }
	void setSha3Uncles(h256 const& _v) { m_sha3Uncles = _v; noteDirty(); }
	void setTimestamp(int64_t _v) { m_timestamp = _v; noteDirty(); }
	void setAuthor(Address const& _v) { m_author = _v; noteDirty(); }
	void setRoots(h256 const& _t, h256 const& _r, h256 const& _u, h256 const& _s) { m_transactionsRoot = _t; m_receiptsRoot = _r; m_stateRoot = _s; m_sha3Uncles = _u; noteDirty(); }
	void setGasUsed(u256 const& _v) { m_gasUsed = _v; noteDirty(); }
	void setNumber(int64_t _v) { m_number = _v; noteDirty(); }
	void setGasLimit(u256 const& _v) { m_gasLimit = _v; noteDirty(); }
	void setExtraData(bytes const& _v) { m_extraData = _v; noteDirty(); }
	void setLogBloom(LogBloom const& _v) { m_logBloom = _v; noteDirty(); }
	void setDifficulty(u256 const& _v) { m_difficulty = _v; noteDirty(); }
	template <class T> void setSeal(unsigned _offset, T const& _value) { Guard l(m_sealLock); if (m_seal.size() <= _offset) m_seal.resize(_offset + 1); m_seal[_offset] = rlp(_value); noteDirty(); }
	template <class T> void setSeal(T const& _value) { setSeal(0, _value); }

	h256 const& parentHash() const { return m_parentHash; }
	h256 const& sha3Uncles() const { return m_sha3Uncles; }
	bool hasUncles() const { return m_sha3Uncles != EmptyListSHA3; }
	int64_t timestamp() const { return m_timestamp; }
	Address const& author() const { return m_author; }
	h256 const& stateRoot() const { return m_stateRoot; }
	h256 const& transactionsRoot() const { return m_transactionsRoot; }
	h256 const& receiptsRoot() const { return m_receiptsRoot; }
	u256 const& gasUsed() const { return m_gasUsed; }
	int64_t number() const { return m_number; }
	u256 const& gasLimit() const { return m_gasLimit; }
	bytes const& extraData() const { return m_extraData; }
	LogBloom const& logBloom() const { return m_logBloom; }
	u256 const& difficulty() const { return m_difficulty; }
	template <class T> T seal(unsigned _offset = 0) const { T ret; Guard l(m_sealLock); if (_offset < m_seal.size()) ret = RLP(m_seal[_offset]).convert<T>(RLP::VeryStrict); return ret; }

private:
	void populate(RLP const& _header);
	void streamRLPFields(RLPStream& _s) const;
	std::vector<bytes> seal() const
	{
		Guard l(m_sealLock);
		return m_seal;
	}
	h256 hashRawRead() const
	{
		Guard l(m_hashLock);
		return m_hash;
	}
	h256 hashWithoutRawRead() const
	{
		Guard l(m_hashLock);
		return m_hashWithout;
	}

	h256 m_parentHash;
	h256 m_sha3Uncles;
	h256 m_stateRoot;
	h256 m_transactionsRoot;
	h256 m_receiptsRoot;
	LogBloom m_logBloom;
	int64_t m_number = 0;
	u256 m_gasLimit;
	u256 m_gasUsed;
	bytes m_extraData;
	int64_t m_timestamp = -1;

	Address m_author;
	u256 m_difficulty;

	std::vector<bytes> m_seal;		///< Additional (RLP-encoded) header fields.
	mutable Mutex m_sealLock;

	mutable h256 m_hash;			///< (Memoised) SHA3 hash of the block header with seal.
	mutable h256 m_hashWithout;		///< (Memoised) SHA3 hash of the block header without seal.
	mutable Mutex m_hashLock;		///< A lock for both m_hash and m_hashWithout.
};

h256类型

  与hash相关的基本都是h256类型，在以太坊中，h256类型定义如下：

using h256 = FixedHash<32>;

class FixedHash
{
public:
	/// The corresponding arithmetic type.
	using Arith = boost::multiprecision::number<boost::multiprecision::cpp_int_backend<N * 8, N * 8, boost::multiprecision::unsigned_magnitude, boost::multiprecision::unchecked, void>>;

	/// The size of the container.
	enum { size = N };

	/// A dummy flag to avoid accidental construction from pointer.
	enum ConstructFromPointerType { ConstructFromPointer };

	/// Method to convert from a string.
	enum ConstructFromStringType { FromHex, FromBinary };

	/// Method to convert from a string.
	enum ConstructFromHashType { AlignLeft, AlignRight, FailIfDifferent };

	/// Construct an empty hash.
	FixedHash() { m_data.fill(0); }

	/// Construct from another hash, filling with zeroes or cropping as necessary.
	template <unsigned M> explicit FixedHash(FixedHash<M> const& _h, ConstructFromHashType _t = AlignLeft) { m_data.fill(0); unsigned c = std::min(M, N); for (unsigned i = 0; i < c; ++i) m_data[_t == AlignRight ? N - 1 - i : i] = _h[_t == AlignRight ? M - 1 - i : i]; }

	/// Convert from the corresponding arithmetic type.
	FixedHash(Arith const& _arith) { toBigEndian(_arith, m_data); }

	/// Convert from unsigned
	explicit FixedHash(unsigned _u) { toBigEndian(_u, m_data); }

	/// Explicitly construct, copying from a byte array.
	explicit FixedHash(bytes const& _b, ConstructFromHashType _t = FailIfDifferent) { if (_b.size() == N) memcpy(m_data.data(), _b.data(), std::min<unsigned>(_b.size(), N)); else { m_data.fill(0); if (_t != FailIfDifferent) { auto c = std::min<unsigned>(_b.size(), N); for (unsigned i = 0; i < c; ++i) m_data[_t == AlignRight ? N - 1 - i : i] = _b[_t == AlignRight ? _b.size() - 1 - i : i]; } } }

	/// Explicitly construct, copying from a byte array.
	explicit FixedHash(bytesConstRef _b, ConstructFromHashType _t = FailIfDifferent) { if (_b.size() == N) memcpy(m_data.data(), _b.data(), std::min<unsigned>(_b.size(), N)); else { m_data.fill(0); if (_t != FailIfDifferent) { auto c = std::min<unsigned>(_b.size(), N); for (unsigned i = 0; i < c; ++i) m_data[_t == AlignRight ? N - 1 - i : i] = _b[_t == AlignRight ? _b.size() - 1 - i : i]; } } }

	/// Explicitly construct, copying from a bytes in memory with given pointer.
	explicit FixedHash(byte const* _bs, ConstructFromPointerType) { memcpy(m_data.data(), _bs, N); }

	/// Explicitly construct, copying from a  string.
	explicit FixedHash(std::string const& _s, ConstructFromStringType _t = FromHex, ConstructFromHashType _ht = FailIfDifferent): FixedHash(_t == FromHex ? fromHex(_s, WhenError::Throw) : dev::asBytes(_s), _ht) {}

	/// Convert to arithmetic type.
	operator Arith() const { return fromBigEndian<Arith>(m_data); }

	/// @returns true iff this is the empty hash.
	explicit operator bool() const { return std::any_of(m_data.begin(), m_data.end(), [](byte _b) { return _b != 0; }); }

	// The obvious comparison operators.
	bool operator==(FixedHash const& _c) const { return m_data == _c.m_data; }
	bool operator!=(FixedHash const& _c) const { return m_data != _c.m_data; }
	bool operator<(FixedHash const& _c) const { for (unsigned i = 0; i < N; ++i) if (m_data[i] < _c.m_data[i]) return true; else if (m_data[i] > _c.m_data[i]) return false; return false; }
	bool operator>=(FixedHash const& _c) const { return !operator<(_c); }
	bool operator<=(FixedHash const& _c) const { return operator==(_c) || operator<(_c); }
	bool operator>(FixedHash const& _c) const { return !operator<=(_c); }

	// The obvious binary operators.
	FixedHash& operator^=(FixedHash const& _c) { for (unsigned i = 0; i < N; ++i) m_data[i] ^= _c.m_data[i]; return *this; }
	FixedHash operator^(FixedHash const& _c) const { return FixedHash(*this) ^= _c; }
	FixedHash& operator|=(FixedHash const& _c) { for (unsigned i = 0; i < N; ++i) m_data[i] |= _c.m_data[i]; return *this; }
	FixedHash operator|(FixedHash const& _c) const { return FixedHash(*this) |= _c; }
	FixedHash& operator&=(FixedHash const& _c) { for (unsigned i = 0; i < N; ++i) m_data[i] &= _c.m_data[i]; return *this; }
	FixedHash operator&(FixedHash const& _c) const { return FixedHash(*this) &= _c; }
	FixedHash operator~() const { FixedHash ret; for (unsigned i = 0; i < N; ++i) ret[i] = ~m_data[i]; return ret; }

	// Big-endian increment.
	FixedHash& operator++() { for (unsigned i = size; i > 0 && !++m_data[--i]; ) {} return *this; }

	/// @returns true if all one-bits in @a _c are set in this object.
	bool contains(FixedHash const& _c) const { return (*this & _c) == _c; }

	/// @returns a particular byte from the hash.
	byte& operator[](unsigned _i) { return m_data[_i]; }
	/// @returns a particular byte from the hash.
	byte operator[](unsigned _i) const { return m_data[_i]; }

	/// @returns an abridged version of the hash as a user-readable hex string.
	std::string abridged() const { return toHex(ref().cropped(0, 4)) + "\342\200\246"; }

	/// @returns a version of the hash as a user-readable hex string that leaves out the middle part.
	std::string abridgedMiddle() const { return toHex(ref().cropped(0, 4)) + "\342\200\246" + toHex(ref().cropped(N - 4)); }

	/// @returns the hash as a user-readable hex string.
	std::string hex() const { return toHex(ref()); }

	/// @returns a mutable byte vector_ref to the object's data.
	bytesRef ref() { return bytesRef(m_data.data(), N); }

	/// @returns a constant byte vector_ref to the object's data.
	bytesConstRef ref() const { return bytesConstRef(m_data.data(), N); }

	/// @returns a mutable byte pointer to the object's data.
	byte* data() { return m_data.data(); }

	/// @returns a constant byte pointer to the object's data.
	byte const* data() const { return m_data.data(); }

	/// @returns begin iterator.
	auto begin() const -> typename std::array<byte, N>::const_iterator { return m_data.begin(); }

	/// @returns end iterator.
	auto end() const -> typename std::array<byte, N>::const_iterator { return m_data.end(); }

	/// @returns a copy of the object's data as a byte vector.
	bytes asBytes() const { return bytes(data(), data() + N); }

	/// @returns a mutable reference to the object's data as an STL array.
	std::array<byte, N>& asArray() { return m_data; }

	/// @returns a constant reference to the object's data as an STL array.
	std::array<byte, N> const& asArray() const { return m_data; }

	/// Populate with random data.
	template <class Engine>
	void randomize(Engine& _eng)
	{
		for (auto& i: m_data)
			i = (uint8_t)std::uniform_int_distribution<uint16_t>(0, 255)(_eng);
	}

	/// @returns a random valued object.
	static FixedHash random() { FixedHash ret; ret.randomize(s_fixedHashEngine); return ret; }

	struct hash
	{
		/// Make a hash of the object's data.
		size_t operator()(FixedHash const& _value) const { return boost::hash_range(_value.m_data.cbegin(), _value.m_data.cend()); }
	};

	template <unsigned P, unsigned M> inline FixedHash& shiftBloom(FixedHash<M> const& _h)
	{
		return (*this |= _h.template bloomPart<P, N>());
	}

	template <unsigned P, unsigned M> inline bool containsBloom(FixedHash<M> const& _h)
	{
		return contains(_h.template bloomPart<P, N>());
	}

	template <unsigned P, unsigned M> inline FixedHash<M> bloomPart() const
	{
		unsigned const c_bloomBits = M * 8;
		unsigned const c_mask = c_bloomBits - 1;
		unsigned const c_bloomBytes = (StaticLog2<c_bloomBits>::result + 7) / 8;

		static_assert((M & (M - 1)) == 0, "M must be power-of-two");
		static_assert(P * c_bloomBytes <= N, "out of range");

		FixedHash<M> ret;
		byte const* p = data();
		for (unsigned i = 0; i < P; ++i)
		{
			unsigned index = 0;
			for (unsigned j = 0; j < c_bloomBytes; ++j, ++p)
				index = (index << 8) | *p;
			index &= c_mask;
			ret[M - 1 - index / 8] |= (1 << (index % 8));
		}
		return ret;
	}

	/// Returns the index of the first bit set to one, or size() * 8 if no bits are set.
	inline unsigned firstBitSet() const
	{
		unsigned ret = 0;
		for (auto d: m_data)
			if (d)
				for (;; ++ret, d <<= 1)
					if (d & 0x80)
						return ret;
					else {}
			else
				ret += 8;
		return ret;
	}

	void clear() { m_data.fill(0); }

private:
	std::array<byte, N> m_data;		///< The binary data.
};

可以看到h256类型，本质是一个字节数组，长度为32个字节

   std::array<byte, N> m_data; ///< The binary data.

重载了运算符，封装了转换为16进制等相关的的函数

  std::string hex() const { return toHex(ref()); }

Address 类型

Address m_author; 定义了挖掘出这个区块的作者地址。Address定义如下：

using Address = h160;

using h160 = FixedHash<20>;  

和hash256类似，长度为20个字节，以太坊钱包公钥的地址刚好就是160位

 比如 0xf3bb1c682c48a99f81e62b9a200eb4ccd6b22403，这个钱包地址，40位16进制串，40*4=160位


2.2 Block

区块(Block)是Ethereum的核心结构体之一。在整个区块链(BlockChain)中，一个个Block是以单向链表的形式相互关联起来的。Block中带有一个Header(指针), Header结构体带有Block的所有属性信息，其中的ParentHash 表示该区块的父区块哈希值， 亦即Block之间关联起来的前向指针.

Block的成员如下：


 State m_state; ///< Our state tree, as an OverlayDB DB.


Transactions m_transactions;  ///< The current list of transactions that we've included in the state.

Transaction(简称tx)，是Ethereum里标示一次交易的结构体

using Transactions = std::vector<Transaction>;


TransactionReceipts m_receipts; ///< The corresponding list of transaction receipts.

      using TransactionReceipts = std::vector<TransactionReceipt>;



h256Hash m_transactionSet; ///< The set of transaction hashes that we've included in the state.
State m_precommit; ///< State at the point immediately prior to rewards.


BlockHeader m_previousBlock; ///< The previous block's information.
BlockHeader m_currentBlock; ///< The current block's information.
bytes m_currentBytes; ///< The current block's bytes.
bool m_committedToSeal = false; ///< Have we committed to mine on the present m_currentBlock?


bytes m_currentTxs; ///< The RLP-encoded block of transactions.
bytes m_currentUncles; ///< The RLP-encoded block of uncles.


Address m_author; ///< Our address (i.e. the address to which fees go).


SealEngineFace* m_sealEngine = nullptr; ///< The chain's seal engine.



猜你喜欢

转载自blog.csdn.net/metal1/article/details/80203233
今日推荐
基于大语言模型的开源知识库问答系统 MaxKB GitHub Star 数量突破 5,000 个！
美国拟限制 AI 大模型出口中国和俄罗斯
苹果将与 OpenAI 达成协议，将 ChatGPT 应用于 iPhone
openKylin 社区生态委员会第六次会议圆满召开
阿里云正式发布通义千问 2.5
Python 3.13 发布首个 Beta：实验性自由线程模式和 JIT、改进交互式解释器
Stack Overflow 拿我的代码去训练 AI 大模型，还封了我的账号
Pop!_OS 的 COSMIC 桌面完成 App Store 上架工作
报告：Django 仍然是 74% 开发者的首选
《2024 年一季度互联网投融资运行情况》研究报告
15 年前上了“FFmpeg 耻辱柱”，今天他还得谢谢咱——腾讯QQPlayer一雪前耻？
TIOBE 5 月榜单：Fortran “复活”进入 Top 10
周排行
BPM为企业带来的实际利益
好程序员web前端分享css常用属性缩写
Java文件下载（excel）
css样式的动态添加及显示和隐藏等零碎用法
axios全局配置以及拦截器
使用Logstash来实时同步MySQL和log日志数据到ES
C++获取当前时间（年月日、时分秒、毫秒）
Odoo产品分析 (四) -- 工具板块(11) -- 网站即时聊天(1)
Java环境配置正确，但是java、javac、java -version均返回“不是内部或外部命令，也不是可运行的程序或批处理文件”？
01 官网下载各种CentOS教程（超详细版）
每日归档
更多
2024-05-14(0)
2024-05-13(18)
2024-05-12(0)
2024-05-11(38)
2024-05-10(38)
2024-05-09(35)
2024-05-08(42)
2024-05-07(14)
2024-05-06(40)
2024-05-05(0)

代码天地 © 2018 codetd.com

境外服务器   最新电视剧2022