This took quite a bit longer than I expected, but I have finally
finished my BIP 21 URI parser. I am attaching several patches for your
review. If these look good, I will submit a pull request.
The first patch, 0001-Implement-a-bitcoin-URI-parser, adds the parser
to libwallet. The other two patches tighten up the base58 decoding in
libbitcoin.
I have given my best effort towards following the libbitcoin style. My
background is in writing embedded C code for tiny micro-controllers,
but I have tried to leave those habits behind and code in an idiomatic
C++11 manner. Hopefully the code is clean enough for your standards.
-William
From edd0420c81991a7984adb317e46d0b9f11705b0a Mon Sep 17 00:00:00 2001
From: William Swanson <swansontec@???>
Date: Tue, 11 Mar 2014 00:52:16 -0700
Subject: [PATCH] Implement a bitcoin URI parser
This parser follows the BIP 21 standard as closely as possible.
---
AUTHORS | 3 +
examples/.gitignore | 2 +
examples/Makefile | 9 +-
examples/uri.cpp | 105 ++++++++++++++++++++++
include/wallet/Makefile.am | 3 +-
include/wallet/uri.hpp | 72 +++++++++++++++
include/wallet/wallet.hpp | 1 +
src/Makefile.am | 3 +-
src/uri.cpp | 212 +++++++++++++++++++++++++++++++++++++++++++++
9 files changed, 405 insertions(+), 5 deletions(-)
create mode 100644 examples/.gitignore
create mode 100644 examples/uri.cpp
create mode 100644 include/wallet/uri.hpp
create mode 100644 src/uri.cpp
diff --git a/AUTHORS b/AUTHORS
index e9fa421..c9ca425 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -13,6 +13,9 @@ Authors:
* Denis Roio (jaromil)
Hosting.
+* William Swanson
+ Development. Added URI code.
+
Thanks:
Kamil Domański (kdomanski)
diff --git a/examples/.gitignore b/examples/.gitignore
new file mode 100644
index 0000000..c91bfb6
--- /dev/null
+++ b/examples/.gitignore
@@ -0,0 +1,2 @@
+determ
+uri
diff --git a/examples/Makefile b/examples/Makefile
index 17c56de..2dcd229 100644
--- a/examples/Makefile
+++ b/examples/Makefile
@@ -3,14 +3,17 @@ LIBS=$(shell pkg-config --libs libbitcoin libwallet)
default: all
-determ.o: determ.cpp
+.cpp.o:
$(CXX) -o $@ -c $< $(CXXFLAGS)
determ: determ.o
$(CXX) -o $@ $< $(LIBS)
-all: determ
+uri: uri.o
+ $(CXX) -o $@ $< $(LIBS)
+
+all: determ uri
clean:
- rm -f determ
+ rm -f determ uri
rm -f *.o
diff --git a/examples/uri.cpp b/examples/uri.cpp
new file mode 100644
index 0000000..e5d881f
--- /dev/null
+++ b/examples/uri.cpp
@@ -0,0 +1,105 @@
+/*
+ * Copyright (c) 2011-2013 libwallet developers (see AUTHORS)
+ *
+ * This file is part of libwallet.
+ *
+ * libwallet is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License with
+ * additional permissions to the one published by the Free Software
+ * Foundation, either version 3 of the License, or (at your option)
+ * any later version. For more information see LICENSE.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+/*
+ Demonstration of URI utilities.
+*/
+#include <bitcoin/bitcoin.hpp>
+#include <wallet/wallet.hpp>
+#include <iostream>
+
+void test_uri_parse(std::string uri)
+{
+ std::cout << "parse URI: \"" << uri << "\"" << std::endl;
+ class parse_handler: public libwallet::uri_parse_handler
+ {
+ virtual void got_address(std::string address)
+ {
+ std::cout << " got address: \"" << address << "\"" << std::endl;
+ }
+ virtual void got_param(std::string key, std::string value)
+ {
+ std::cout << " got parameter: \"" << key << "\" = \"" <<
+ value << "\"" << std::endl;
+ }
+ } handler;
+ if (libwallet::uri_parse(uri, handler))
+ std::cout << " ok" << std::endl;
+ else
+ std::cout << " error" << std::endl;
+}
+
+void test_uri_decode(std::string uri)
+{
+ libwallet::decoded_uri out = libwallet::uri_decode(uri);
+ std::cout << "decode URI: \"" << uri << "\"" << std::endl;
+ if (!out.valid)
+ std::cout << " invalid" << std::endl;
+ if (out.has_address)
+ std::cout << " address: " << out.address.encoded() << std::endl;
+ if (out.has_amount)
+ std::cout << " amount: " << out.amount << std::endl;
+ if (out.has_label)
+ std::cout << " label: \"" << out.label << "\"" << std::endl;
+ if (out.has_message)
+ std::cout << " message: \"" << out.message << "\"" << std::endl;
+ if (out.has_r)
+ std::cout << " r: \"" << out.r << "\"" << std::endl;
+}
+
+int main()
+{
+ test_uri_parse("bitcoin:113Pfw4sFqN1T5kXUnKbqZHMJHN9oyjtgD?label=test");
+ test_uri_parse("bitcoin:");
+ test_uri_parse("bitcorn:");
+ test_uri_parse("BITCOIN:?");
+ test_uri_parse("Bitcoin:?&");
+ test_uri_parse("bitcOin:&");
+ test_uri_parse("bitcoin:?x=y");
+ test_uri_parse("bitcoin:?x=");
+ test_uri_parse("bitcoin:?=y");
+ test_uri_parse("bitcoin:?=");
+ test_uri_parse("bitcoin:?x");
+ test_uri_parse("bitcoin:19z88");
+ test_uri_parse("bitcoin:19l88");
+ test_uri_parse("bitcoin:19z88?x=http://www.example.com?purchase%3Dshoes");
+ test_uri_parse("bitcoin:19z88?name=%E3%83%95"); // UTF-8
+ test_uri_parse("bitcoin:19z88?name=%3");
+ test_uri_parse("bitcoin:19z88?name=%3G");
+ test_uri_parse("bitcoin:19z88?name=%3f");
+ test_uri_parse("bitcoin:%31");
+
+ std::cout << "================================" << std::endl;
+
+ test_uri_decode("bitcoin:113Pfw4sFqN1T5kXUnKbqZHMJHN9oyjtgD");
+ test_uri_decode("bitcoin:19z88");
+ test_uri_decode("bitcoin:?=");
+ test_uri_decode("bitcoin:?amount=4.2");
+ test_uri_decode("bitcoin:?amount=.");
+ test_uri_decode("bitcoin:?amount=4.2.4");
+ test_uri_decode("bitcoin:?amount=foo");
+ test_uri_decode("bitcoin:?label=Bob");
+ test_uri_decode("bitcoin:?message=Hi%20Alice");
+ test_uri_decode("bitcoin:?r=http://www.example.com?purchase%3Dshoes");
+ test_uri_decode("bitcoin:?foo=ignore");
+ test_uri_decode("bitcoin:?req-foo=die");
+
+ return 0;
+}
+
diff --git a/include/wallet/Makefile.am b/include/wallet/Makefile.am
index 748eed4..94eca86 100644
--- a/include/wallet/Makefile.am
+++ b/include/wallet/Makefile.am
@@ -4,5 +4,6 @@ wallet_include_HEADERS = \
key_formats.hpp \
transaction.hpp \
deterministic_wallet.hpp \
- mnemonic.hpp
+ mnemonic.hpp \
+ uri.hpp
diff --git a/include/wallet/uri.hpp b/include/wallet/uri.hpp
new file mode 100644
index 0000000..959d02b
--- /dev/null
+++ b/include/wallet/uri.hpp
@@ -0,0 +1,72 @@
+/*
+ * Copyright (c) 2011-2013 libwallet developers (see AUTHORS)
+ *
+ * This file is part of libwallet.
+ *
+ * libwallet is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License with
+ * additional permissions to the one published by the Free Software
+ * Foundation, either version 3 of the License, or (at your option)
+ * any later version. For more information see LICENSE.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef LIBWALLET_URI_HPP
+#define LIBWALLET_URI_HPP
+
+#include <bitcoin/address.hpp>
+
+namespace libwallet {
+
+struct uri_parse_handler {
+ virtual void got_address(std::string address) = 0;
+ virtual void got_param(std::string key, std::string value) = 0;
+};
+
+bool uri_parse(const std::string& uri, uri_parse_handler& handler);
+bool uri_validate(const std::string& uri);
+
+/**
+ * A decoded bitcoin URI corresponding to BIP 21 and BIP 72.
+ * All string members are UTF-8.
+ */
+struct decoded_uri
+{
+ bool valid;
+ bool has_address;
+ bool has_amount;
+ bool has_label;
+ bool has_message;
+ bool has_r;
+
+ libbitcoin::payment_address address;
+ uint64_t amount;
+ std::string label;
+ std::string message;
+ std::string r;
+
+ decoded_uri()
+ : valid(true), has_address(false), has_amount(false),
+ has_label(false), has_message(false), has_r(false)
+ {}
+};
+
+decoded_uri uri_decode(const std::string& uri);
+
+/**
+ * Parses a bitcoin amount string.
+ * @return string value, in satoshis, or -1 for failure.
+ */
+uint64_t parse_amount(const std::string& amount);
+
+} // libwallet
+
+#endif
+
diff --git a/include/wallet/wallet.hpp b/include/wallet/wallet.hpp
index ba88126..5829532 100644
--- a/include/wallet/wallet.hpp
+++ b/include/wallet/wallet.hpp
@@ -25,6 +25,7 @@
#include <wallet/mnemonic.hpp>
#include <wallet/key_formats.hpp>
#include <wallet/transaction.hpp>
+#include <wallet/uri.hpp>
#endif
diff --git a/src/Makefile.am b/src/Makefile.am
index 835513c..2225a9c 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -6,7 +6,8 @@ libwallet_la_SOURCES = \
deterministic_wallet.cpp \
mnemonic.cpp \
transaction.cpp \
- key_formats.cpp
+ key_formats.cpp \
+ uri.cpp
libwallet_la_LIBADD = $(libbitcoin_LIBS)
diff --git a/src/uri.cpp b/src/uri.cpp
new file mode 100644
index 0000000..3abce6b
--- /dev/null
+++ b/src/uri.cpp
@@ -0,0 +1,212 @@
+/*
+ * Copyright (c) 2011-2013 libwallet developers (see AUTHORS)
+ *
+ * This file is part of libwallet.
+ *
+ * libwallet is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU Affero General Public License with
+ * additional permissions to the one published by the Free Software
+ * Foundation, either version 3 of the License, or (at your option)
+ * any later version. For more information see LICENSE.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU Affero General Public License for more details.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+#include <wallet/uri.hpp>
+#include <bitcoin/utility/base58.hpp>
+#include <sstream>
+#include <stdlib.h>
+
+namespace libwallet {
+
+static bool is_digit(char c)
+{
+ return '0' <= c && c <= '9';
+}
+static bool is_hex(char c)
+{
+ return is_digit(c) || ('A' <= c && c <= 'F') || ('a' <= c && c <= 'f');
+}
+static bool is_qchar(char c)
+{
+ return
+ ('A' <= c && c <= 'Z') || ('a' <= c && c <= 'z') || is_digit(c) ||
+ '-' == c || '.' == c || '_' == c || '~' == c || // unreserved
+ '!' == c || '$' == c || '\'' == c || '(' == c || ')' == c ||
+ '*' == c || '+' == c || ',' == c || ';' == c || // sub-delims
+ ':' == c || '@' == c || // pchar
+ '/' == c || '?' == c; // query
+}
+
+static unsigned from_hex(char c)
+{
+ return
+ 'A' <= c && c <= 'F' ? 10 + c - 'A' :
+ 'a' <= c && c <= 'f' ? 10 + c - 'a' :
+ c - '0';
+}
+
+/**
+ * Unescapes a URI-encoded string while advancing the iterator.
+ * @param i set to one-past the last-read character on return.
+ */
+typedef std::string::const_iterator sci;
+static std::string unescape(sci& i, sci end, bool (*is_valid)(char))
+{
+ auto j = i;
+ size_t count = 0;
+ while (end != i && (is_valid(*i) ||
+ ('%' == *i && 2 < end - i && is_hex(i[1]) && is_hex(i[2]))))
+ {
+ ++count;
+ i += ('%' == *i ? 3 : 1);
+ }
+ std::string out;
+ out.reserve(count);
+ while (j != i)
+ {
+ out.push_back('%' == *j ? from_hex(j[1]) << 4 | from_hex(j[2]) : *j);
+ j += ('%' == *j ? 3 : 1);
+ }
+ return out;
+}
+
+bool uri_parse(const std::string& uri, uri_parse_handler& handler)
+{
+ auto i = uri.begin();
+
+ // URI scheme:
+ const char* lower = "bitcoin:";
+ const char* upper = "BITCOIN:";
+ while (*lower)
+ {
+ if (uri.end() == i || (*lower != *i && *upper != *i))
+ return false;
+ ++lower; ++upper; ++i;
+ }
+
+ // Payment address:
+ std::string address = unescape(i, uri.end(), libbitcoin::is_base58);
+ if (uri.end() != i && '?' != *i)
+ return false;
+ if (!address.empty())
+ handler.got_address(address);
+
+ // Parameters:
+ while (uri.end() != i)
+ {
+ ++i; // Consume '?' or '&'
+ std::string key = unescape(i, uri.end(), is_qchar);
+ std::string value;
+ if (uri.end() != i && '=' == *i)
+ {
+ ++i; // Consume '='
+ if (key.empty())
+ return false;
+ value = unescape(i, uri.end(), is_qchar);
+ }
+ if (uri.end() != i && '&' != *i)
+ return false;
+ if (!key.empty())
+ handler.got_param(key, value);
+ }
+ return true;
+}
+
+bool uri_validate(const std::string& uri)
+{
+ class parse_handler: public uri_parse_handler
+ {
+ virtual void got_address(std::string address)
+ {
+ (void)address;
+ }
+ virtual void got_param(std::string key, std::string value)
+ {
+ (void)key;
+ (void)value;
+ }
+ } handler;
+ return uri_parse(uri, handler);
+}
+
+decoded_uri uri_decode(const std::string& uri)
+{
+ class parse_handler: public uri_parse_handler
+ {
+ public:
+ decoded_uri wip_;
+ virtual void got_address(std::string address)
+ {
+ if (wip_.address.set_encoded(address))
+ wip_.has_address = true;
+ else
+ wip_.valid = false;
+ }
+ virtual void got_param(std::string key, std::string value)
+ {
+ if ("amount" == key)
+ {
+ wip_.amount = parse_amount(value);
+ if (static_cast<uint64_t>(-1) != wip_.amount)
+ wip_.has_amount = true;
+ else
+ wip_.valid = false;
+ }
+ else if ("label" == key)
+ {
+ wip_.label = value;
+ wip_.has_label = true;
+ }
+ else if ("message" == key)
+ {
+ wip_.message = value;
+ wip_.has_message = true;
+ }
+ else if ("r" == key)
+ {
+ wip_.r = value;
+ wip_.has_r = true;
+ }
+ else if (!key.compare(0, 4, "req-"))
+ {
+ wip_.valid = false;
+ }
+ }
+ } handler;
+ if (!uri_parse(uri, handler))
+ handler.wip_.valid = false;
+ return handler.wip_;
+}
+
+/**
+ * Validates an amount string according to the BIP 21 grammar.
+ */
+static bool check_amount(const std::string& amount)
+{
+ auto i = amount.begin();
+ while (amount.end() != i && is_digit(*i))
+ ++i;
+ if (amount.end() != i && '.' == *i)
+ {
+ ++i;
+ while (amount.end() != i && is_digit(*i))
+ ++i;
+ }
+ return amount.end() == i;
+}
+
+uint64_t parse_amount(const std::string& amount)
+{
+ if (!check_amount(amount))
+ return static_cast<uint64_t>(-1);
+ // This code might have numerical problems:
+ return static_cast<uint64_t>(100000000*strtod(amount.c_str(), nullptr));
+}
+
+} // namespace libwallet
--
1.9.0
From 1973ed73de5684ecc078a674c6170bc3ca9df77e Mon Sep 17 00:00:00 2001
From: William Swanson <swansontec@???>
Date: Wed, 5 Mar 2014 16:38:59 -0800
Subject: [PATCH 1/2] Detect wrongly-encoded bitcoin addresses
---
include/bitcoin/utility/base58.hpp | 3 +++
src/address.cpp | 2 ++
src/utility/base58.cpp | 13 ++++++++++++-
3 files changed, 17 insertions(+), 1 deletion(-)
diff --git a/include/bitcoin/utility/base58.hpp b/include/bitcoin/utility/base58.hpp
index 63e67c9..75ac9cd 100644
--- a/include/bitcoin/utility/base58.hpp
+++ b/include/bitcoin/utility/base58.hpp
@@ -24,6 +24,9 @@
namespace libbitcoin {
+bool is_base58(char c);
+bool is_base58(std::string const& text);
+
std::string encode_base58(const data_chunk& unencoded_data);
data_chunk decode_base58(std::string encoded_data);
diff --git a/src/address.cpp b/src/address.cpp
index f71607c..ab52049 100644
--- a/src/address.cpp
+++ b/src/address.cpp
@@ -59,6 +59,8 @@ const short_hash& payment_address::hash() const
bool payment_address::set_encoded(const std::string& encoded_address)
{
+ if (!is_base58(encoded_address))
+ return false;
const data_chunk decoded_address = decode_base58(encoded_address);
// version + 20 bytes short hash + 4 bytes checksum
if (decoded_address.size() != 25)
diff --git a/src/utility/base58.cpp b/src/utility/base58.cpp
index c587e52..f31ae63 100644
--- a/src/utility/base58.cpp
+++ b/src/utility/base58.cpp
@@ -26,7 +26,18 @@ namespace libbitcoin {
// Thanks for all the wonderful bitcoin hackers
-const char* base58_chars = "123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz";
+const char base58_chars[] = "123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz";
+
+bool is_base58(char c)
+{
+ // This works because the base58 characters happen to be in sorted order
+ return std::binary_search(base58_chars, std::end(base58_chars) - 1, c);
+}
+bool is_base58(std::string const& text)
+{
+ return std::all_of(text.begin(), text.end(),
+ [](char c) { return is_base58(c); });
+}
std::string encode_base58(const data_chunk& unencoded_data)
{
--
1.9.0
From 10d3c7e76bd3d9f9eae577f0acb4062f0c0ac7fe Mon Sep 17 00:00:00 2001
From: William Swanson <swansontec@???>
Date: Thu, 6 Mar 2014 16:31:38 -0800
Subject: [PATCH 2/2] Optimize the base58 decoder
The existing version was not using a binary search,
and was potentially allocating memory on each iteration.
---
src/utility/base58.cpp | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/src/utility/base58.cpp b/src/utility/base58.cpp
index f31ae63..a089fb9 100644
--- a/src/utility/base58.cpp
+++ b/src/utility/base58.cpp
@@ -87,8 +87,9 @@ data_chunk decode_base58(std::string encoded_data)
for (const uint8_t current_char: encoded_data)
{
bn *= 58;
- bn += std::string(base58_chars).find(current_char);
- }
+ bn += std::lower_bound(base58_chars, std::end(base58_chars) - 1,
+ current_char) - base58_chars;
+ }
// Get bignum as little endian data
data_chunk temp_data = bn.data();
--
1.9.0
::
[Libbitcoin] [PATCH] Implement a bi…
