BlockSci v0.6 not recognizing ZCash t-addr addresses

[jredondo]

Greetings.
I'm using BlockSci AMI with BlockSci v0.6 (commit 304e1ff7eb670e5fff161688e26e698c56f21e28).
It seems I have parsed successfully the ZCash blockchain since parser finished with the expected message and the data directory seems filled as expected.
Nonetheless, it seems BlockSci is not able to interpret ZCash t-addr addresses.
I have this small code for testing:

#include <blocksci/blocksci.hpp>
#include <blocksci/chain/algorithms.hpp>
#include <blocksci/chain/transaction.hpp>
#include <blocksci/chain/block.hpp>
#include <blocksci/chain/blockchain.hpp>
#include <blocksci/script.hpp>
#include <blocksci/address/address.hpp>
#include <blocksci/chain/transaction_range.hpp>

#include <range/v3/all.hpp>

#include <numeric>
#include <chrono>
#include <unordered_map>
#include <map>
#include <fstream>
#include <iostream>
#include <set>
#include <functional>
#include <string>

using namespace blocksci;

int main(int argc, const char * argv[]) {
    Blockchain chain(argv[1]);
    auto begin = std::chrono::steady_clock::now();

    auto address = getAddressFromString(argv[2], chain.getAccess());
    if(address == ranges::nullopt) {
        std::cout << "Null address: " << argv[2]  << std::endl ;
        return -1;
    }

    std::cout << "Address: " << (*address).toString() << std::endl ;

    //std::cout << "Address: " << address->toString() << std::endl ;
    std::cout << "Address balance: " << (*address).calculateBalance(std::atoi(argv[3])) << std::endl ;

    auto endTime = std::chrono::steady_clock::now();
    double timeMSecs = std::chrono::duration_cast<std::chrono::microseconds>(endTime - begin).count();
    std::cout << "time: " << timeMSecs << std::endl ;
}

After running it like this (ZCash):

user@server:...$ ./test_balance ../../blocksci-http-server/conf/conf_zch.txt t1aZvxRLCGVeMPFXvqfnBgHVEbi4c6g8MVa 1000000
Null address: t1aZvxRLCGVeMPFXvqfnBgHVEbi4c6g8MVa
user@server:...$ 

And after running like (Litecoin):

user@server:...$ ./test_balance ../../blocksci-http-server/conf/conf_ltc.txt ltc1qxhpwa9u8dwtpegj8gkay4ewju7py09uawhp28p 10000000
Address: WitnessPubkeyAddress(ltc1qxhpwa9u8dwtpegj8gkay4ewju7py09uawhp28p)
Address balance: 1017909332001
time: 17651
user@server:...$ 

Here are the config files used up there:
ZCash:

{
    "chainConfig": {
        "coinName": "zcash",
        "dataDirectory": "/home/ubuntu/zcash",
        "pubkeyPrefix": [
            28,
            184
        ],
        "scriptPrefix": [
            28,
            189
        ],
        "segwitActivationHeight": 2147483647,
        "segwitPrefix": "NONE"
    },
    "parser": {
        "maxBlockNum": -6,
        "rpc": {
            "address": "127.0.0.1",
            "password": "test",
            "port": 8232,
            "username": "test"
        }
    },
    "version": 5
}

Litecoin:

{
    "chainConfig": {
        "coinName": "litecoin",
        "dataDirectory": "/home/ubuntu/litecoin",
        "pubkeyPrefix": [
            48
        ],
        "scriptPrefix": [
            50
        ],
        "segwitActivationHeight": 1201536,
        "segwitPrefix": "ltc"
    },
    "parser": {
        "disk": {
            "blockMagic": 3686187259,
            "coinDirectory": "/home/ubuntu/.litecoin",
            "hashFuncName": "doubleSha256"
        },
        "maxBlockNum": -6
    },
    "version": 5
}

Please let me know if there is anything else I could provide.
Thanks in advance.

[jredondo]

Using gdb I have realized that at this point the variable type is blocksci::AddressType::NONSTANDARD and thus addressNum never gets anything different than ranges::nullopt.

[jredondo]

I have implemented this small patch and it seems to work as expected using:

auto address = getAddressFromString_2(argv[2], chain.getAccess());

This is the git diff with commit 6a07f305834c197c1c641eae32bf23da2b31adb3:

user@server:~/jredondo/BlockSci$ git diff
diff --git a/include/blocksci/address/address.hpp b/include/blocksci/address/address.hpp
index af1e6f5..c16e4e2 100644
--- a/include/blocksci/address/address.hpp
+++ b/include/blocksci/address/address.hpp
@@ -110,6 +110,7 @@ namespace blocksci {
     void BLOCKSCI_EXPORT visit(const Address &address, const std::function<bool(const Address &)> &visitFunc);
     
     ranges::optional<Address> BLOCKSCI_EXPORT getAddressFromString(const std::string &addressString, DataAccess &access);
+    ranges::optional<Address> BLOCKSCI_EXPORT getAddressFromString_2(const std::string &addressString, DataAccess &access);
     
     std::vector<Address> BLOCKSCI_EXPORT getAddressesWithPrefix(const std::string &prefix, DataAccess &access);
     
diff --git a/src/address/address.cpp b/src/address/address.cpp
index ebe0b0e..1c663f9 100644
--- a/src/address/address.cpp
+++ b/src/address/address.cpp
@@ -232,6 +232,45 @@ namespace blocksci {
     ScriptBase Address::getBaseScript() const {
         return ScriptBase(*this);
     }
+
+    ranges::optional<Address> getAddressFromString_2(const std::string &addressString, DataAccess &access) {
+        if (addressString.compare(0, access.config.chainConfig.segwitPrefix.size(), access.config.chainConfig.segwitPrefix) == 0) {
+            std::pair<int, std::vector<uint8_t> > decoded = segwit_addr::decode(access.config.chainConfig.segwitPrefix, addressString);
+            if (decoded.first == 0) {
+                if (decoded.second.size() == 20) {
+                    uint160 pubkeyHash(decoded.second.begin(), decoded.second.end());
+                    ranges::optional<uint32_t> addressNum = access.getHashIndex().getPubkeyHashIndex(pubkeyHash);
+                    if (addressNum) {
+                        return Address{*addressNum, AddressType::WITNESS_PUBKEYHASH, access};
+                    }
+                } else if (decoded.second.size() == 32) {
+                    uint256 scriptHash(decoded.second.begin(), decoded.second.end());
+                    ranges::optional<uint32_t> addressNum = access.getHashIndex().getScriptHashIndex(scriptHash);
+                    if (addressNum) {
+                        return Address{*addressNum, AddressType::WITNESS_SCRIPTHASH, access};
+                    }
+                }
+            }
+            return ranges::nullopt;
+        }
+
+        CBitcoinAddress address;
+        assert(access.config.chainConfig.segwitPrefix.size() ==access.config.chainConfig.pubkeyPrefix.size())
+        address.SetString(addressString.c_str(), access.config.chainConfig.pubkeyPrefix.size());
+        uint160 hash;
+        blocksci::AddressType::Enum type;
+        std::tie(hash, type) = address.Get(access.config.chainConfig);
+        ranges::optional<uint32_t> addressNum = ranges::nullopt;
+        if (type == AddressType::Enum::PUBKEYHASH) {
+            addressNum = access.getHashIndex().getPubkeyHashIndex(hash);
+        } else if (type == AddressType::Enum::SCRIPTHASH) {
+            addressNum = access.getHashIndex().getScriptHashIndex(hash);
+        }
+        if (addressNum) {
+            return Address{*addressNum, type, access};
+        } else {
+            return ranges::nullopt;
+        }
+    }
     
     ranges::optional<Address> getAddressFromString(const std::string &addressString, DataAccess &access) {
         if (addressString.compare(0, access.config.chainConfig.segwitPrefix.size(), access.config.chainConfig.segwitPrefix) == 0) {
diff --git a/src/scripts/bitcoin_base58.cpp b/src/scripts/bitcoin_base58.cpp
index 66f4e9d..707e5a1 100755
--- a/src/scripts/bitcoin_base58.cpp
+++ b/src/scripts/bitcoin_base58.cpp
@@ -206,6 +206,12 @@ namespace blocksci {
         return 0;
     }
 
+    CBitcoinAddress::CBitcoinAddress()
+    {
+        vchVersion.clear();
+        vchData.clear();
+    }
+
 
     CBitcoinAddress::CBitcoinAddress(const uint160 &dest, AddressType::Enum type, const ChainConfiguration &config) {
         if (type == AddressType::Enum::PUBKEYHASH || type == AddressType::Enum::PUBKEY || type == AddressType::Enum::MULTISIG_PUBKEY) {
diff --git a/src/scripts/bitcoin_base58.hpp b/src/scripts/bitcoin_base58.hpp
index 41c7c55..4e5144c 100755
--- a/src/scripts/bitcoin_base58.hpp
+++ b/src/scripts/bitcoin_base58.hpp
@@ -104,6 +104,7 @@ namespace blocksci {
      */
     class CBitcoinAddress : public CBase58Data {
     public:
+        CBitcoinAddress();
         CBitcoinAddress(const uint160 &dest, AddressType::Enum type, const ChainConfiguration &config);
         CBitcoinAddress(const uint160 &dest, const std::vector<unsigned char>& version);
         CBitcoinAddress(const std::string& strAddress) { SetString(strAddress); }
user@server:~/jredondo/BlockSci$ 

[jredondo]

Thanks for fixing this.