peerplays_migrated/libraries/chain/protocol/confidential.cpp
Daniel Larimer be65c277c9 Implement and Test Confidential Transactions
This checkin contains a fully functional confidential transaction
integration with passing unit tests.
2015-07-22 14:45:13 -04:00

122 lines
4 KiB
C++

#include <graphene/chain/protocol/confidential.hpp>
#include <graphene/chain/confidential_evaluator.hpp>
#include <graphene/chain/database.hpp>
namespace graphene { namespace chain {
void transfer_to_blind_operation::validate()const
{
FC_ASSERT( fee.amount >= 0 );
FC_ASSERT( amount.amount > 0 );
vector<commitment_type> in;
vector<commitment_type> out(outputs.size());
int64_t net_public = amount.amount.value;
for( uint32_t i = 0; i < out.size(); ++i )
{
out[i] = outputs[i].commitment;
/// require all outputs to be sorted prevents duplicates AND prevents implementations
/// from accidentally leaking information by how they arrange commitments.
if( i > 0 ) FC_ASSERT( out[i-1] < out[i], "all outputs must be sorted by commitment id" );
FC_ASSERT( !outputs[i].owner.is_impossible() );
}
FC_ASSERT( out.size(), "there must be at least one output" );
auto public_c = fc::ecc::blind(blinding_factor,net_public);
FC_ASSERT( fc::ecc::verify_sum( {public_c}, out, 0 ), "", ("net_public",net_public) );
if( outputs.size() > 1 )
{
for( auto out : outputs )
{
auto info = fc::ecc::range_get_info( out.range_proof );
FC_ASSERT( info.max_value <= GRAPHENE_MAX_SHARE_SUPPLY );
}
}
}
share_type transfer_to_blind_operation::calculate_fee( const fee_parameters_type& k )const
{
return k.fee + outputs.size() * k.price_per_output + calculate_data_fee( fc::raw::pack_size(*this), k.price_per_kb );
}
void transfer_from_blind_operation::validate()const
{
FC_ASSERT( amount.amount > 0 );
FC_ASSERT( fee.amount >= 0 );
FC_ASSERT( inputs.size() > 0 );
FC_ASSERT( amount.asset_id == fee.asset_id );
vector<commitment_type> in(inputs.size());
vector<commitment_type> out;
int64_t net_public = fee.amount.value + amount.amount.value;
out.push_back( fc::ecc::blind( blinding_factor, net_public ) );
for( uint32_t i = 0; i < in.size(); ++i )
{
in[i] = inputs[i].commitment;
/// by requiring all inputs to be sorted we also prevent duplicate commitments on the input
if( i > 0 ) FC_ASSERT( in[i-1] < in[i], "all inputs must be sorted by commitment id" );
}
FC_ASSERT( in.size(), "there must be at least one input" );
FC_ASSERT( fc::ecc::verify_sum( in, out, 0 ) );
}
/**
* If fee_payer = temp_account_id, then the fee is paid by the surplus balance of inputs-outputs and
* 100% of the fee goes to the network.
*/
account_id_type blind_transfer_operation::fee_payer()const
{
return GRAPHENE_TEMP_ACCOUNT;
}
/**
* This method can be computationally intensive because it verifies that input commitments - output commitments add up to 0
*/
void blind_transfer_operation::validate()const
{ try {
vector<commitment_type> in(inputs.size());
vector<commitment_type> out(outputs.size());
int64_t net_public = fee.amount.value;//from_amount.value - to_amount.value;
for( uint32_t i = 0; i < in.size(); ++i )
{
in[i] = inputs[i].commitment;
/// by requiring all inputs to be sorted we also prevent duplicate commitments on the input
if( i > 0 ) FC_ASSERT( in[i-1] < in[i] );
}
for( uint32_t i = 0; i < out.size(); ++i )
{
out[i] = outputs[i].commitment;
if( i > 0 ) FC_ASSERT( out[i-1] < out[i] );
FC_ASSERT( !outputs[i].owner.is_impossible() );
}
FC_ASSERT( in.size(), "there must be at least one input" );
FC_ASSERT( fc::ecc::verify_sum( in, out, net_public ), "", ("net_public", net_public) );
if( outputs.size() > 1 )
{
for( auto out : outputs )
{
auto info = fc::ecc::range_get_info( out.range_proof );
FC_ASSERT( info.max_value <= GRAPHENE_MAX_SHARE_SUPPLY );
}
}
} FC_CAPTURE_AND_RETHROW( (*this) ) }
share_type blind_transfer_operation::calculate_fee( const fee_parameters_type& k )const
{
return k.fee + outputs.size() * k.price_per_output + calculate_data_fee( fc::raw::pack_size(*this), k.price_per_kb );;
}
} } // graphene::chain