Merge #16887: Abstract out some of the descriptor Span-parsing helpers

bb36372b8f2bd675313ae8553ceb61f28c2c1afd test: add unit tests for Span-parsing helpers (Sebastian Falbesoner)
5e69aeec3f2a0fafd5e591b7222716f00145761d Add documenting comments to spanparsing.h (Pieter Wuille)
230d43fdbc41b356700b0d8a6984d69e00279ade Abstract out some of the descriptor Span-parsing helpers (Pieter Wuille)

Pull request description:

  As suggested here: https://github.com/bitcoin/bitcoin/pull/16800#issuecomment-531605482.

  This moves the Span parsing functions out of the descriptor module, making them more easily usable for other parsers (in particular, in preparation for miniscript parsing).

ACKs for top commit:
  MarcoFalke:
    ACK bb36372b8f2bd675313ae8553ceb61f28c2c1afd

Tree-SHA512: b5c5c11a9bc3f0a1c2c4cfa22755654ecfb8d4b69da0dc1fb9f04e1556dc0f6ffd87ad153600963279ac465d587d7971b53d240ced802d12693682411ac73deb

5 files changed, 252 insertions, 57 deletions

diff --git a/src/Makefile.am b/src/Makefile.am
index 84254e45d1..eec84122ae 100644
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -208,6 +208,7 @@ BITCOIN_CORE_H = \
   util/bytevectorhash.h \
   util/error.h \
   util/fees.h \
+  util/spanparsing.h \
   util/system.h \
   util/macros.h \
   util/memory.h \
@@ -505,6 +506,7 @@ libbitcoin_util_a_SOURCES = \
   util/moneystr.cpp \
   util/rbf.cpp \
   util/threadnames.cpp \
+  util/spanparsing.cpp \
   util/strencodings.cpp \
   util/string.cpp \
   util/time.cpp \
diff --git a/src/script/descriptor.cpp b/src/script/descriptor.cpp
index b223349eb1..ed1bd4cda9 100644
--- a/src/script/descriptor.cpp
+++ b/src/script/descriptor.cpp
@@ -11,6 +11,7 @@
 
 #include <span.h>
 #include <util/bip32.h>
+#include <util/spanparsing.h>
 #include <util/system.h>
 #include <util/strencodings.h>
 
@@ -640,63 +641,6 @@ enum class ParseScriptContext {
     P2WSH,
 };
 
-/** Parse a constant. If successful, sp is updated to skip the constant and return true. */
-bool Const(const std::string& str, Span<const char>& sp)
-{
-    if ((size_t)sp.size() >= str.size() && std::equal(str.begin(), str.end(), sp.begin())) {
-        sp = sp.subspan(str.size());
-        return true;
-    }
-    return false;
-}
-
-/** Parse a function call. If successful, sp is updated to be the function's argument(s). */
-bool Func(const std::string& str, Span<const char>& sp)
-{
-    if ((size_t)sp.size() >= str.size() + 2 && sp[str.size()] == '(' && sp[sp.size() - 1] == ')' && std::equal(str.begin(), str.end(), sp.begin())) {
-        sp = sp.subspan(str.size() + 1, sp.size() - str.size() - 2);
-        return true;
-    }
-    return false;
-}
-
-/** Return the expression that sp begins with, and update sp to skip it. */
-Span<const char> Expr(Span<const char>& sp)
-{
-    int level = 0;
-    auto it = sp.begin();
-    while (it != sp.end()) {
-        if (*it == '(') {
-            ++level;
-        } else if (level && *it == ')') {
-            --level;
-        } else if (level == 0 && (*it == ')' || *it == ',')) {
-            break;
-        }
-        ++it;
-    }
-    Span<const char> ret = sp.first(it - sp.begin());
-    sp = sp.subspan(it - sp.begin());
-    return ret;
-}
-
-/** Split a string on every instance of sep, returning a vector. */
-std::vector<Span<const char>> Split(const Span<const char>& sp, char sep)
-{
-    std::vector<Span<const char>> ret;
-    auto it = sp.begin();
-    auto start = it;
-    while (it != sp.end()) {
-        if (*it == sep) {
-            ret.emplace_back(start, it);
-            start = it + 1;
-        }
-        ++it;
-    }
-    ret.emplace_back(start, it);
-    return ret;
-}
-
 /** Parse a key path, being passed a split list of elements (the first element is ignored). */
 NODISCARD bool ParseKeyPath(const std::vector<Span<const char>>& split, KeyPath& out, std::string& error)
 {
@@ -723,6 +667,8 @@ NODISCARD bool ParseKeyPath(const std::vector<Span<const char>>& split, KeyPath&
 /** Parse a public key that excludes origin information. */
 std::unique_ptr<PubkeyProvider> ParsePubkeyInner(const Span<const char>& sp, bool permit_uncompressed, FlatSigningProvider& out, std::string& error)
 {
+    using namespace spanparsing;
+
     auto split = Split(sp, '/');
     std::string str(split[0].begin(), split[0].end());
     if (str.size() == 0) {
@@ -782,6 +728,8 @@ std::unique_ptr<PubkeyProvider> ParsePubkeyInner(const Span<const char>& sp, boo
 /** Parse a public key including origin information (if enabled). */
 std::unique_ptr<PubkeyProvider> ParsePubkey(const Span<const char>& sp, bool permit_uncompressed, FlatSigningProvider& out, std::string& error)
 {
+    using namespace spanparsing;
+
     auto origin_split = Split(sp, ']');
     if (origin_split.size() > 2) {
         error = "Multiple ']' characters found for a single pubkey";
@@ -816,6 +764,8 @@ std::unique_ptr<PubkeyProvider> ParsePubkey(const Span<const char>& sp, bool per
 /** Parse a script in a particular context. */
 std::unique_ptr<DescriptorImpl> ParseScript(Span<const char>& sp, ParseScriptContext ctx, FlatSigningProvider& out, std::string& error)
 {
+    using namespace spanparsing;
+
     auto expr = Expr(sp);
     bool sorted_multi = false;
     if (Func("pk", expr)) {
@@ -1012,6 +962,8 @@ std::unique_ptr<DescriptorImpl> InferScript(const CScript& script, ParseScriptCo
 /** Check a descriptor checksum, and update desc to be the checksum-less part. */
 bool CheckChecksum(Span<const char>& sp, bool require_checksum, std::string& error, std::string* out_checksum = nullptr)
 {
+    using namespace spanparsing;
+
     auto check_split = Split(sp, '#');
     if (check_split.size() > 2) {
         error = "Multiple '#' symbols";
diff --git a/src/test/util_tests.cpp b/src/test/util_tests.cpp
index d0cd4b0a03..31a66b6fa9 100644
--- a/src/test/util_tests.cpp
+++ b/src/test/util_tests.cpp
@@ -12,6 +12,7 @@
 #include <util/strencodings.h>
 #include <util/string.h>
 #include <util/time.h>
+#include <util/spanparsing.h>
 
 #include <stdint.h>
 #include <thread>
@@ -1572,4 +1573,127 @@ BOOST_AUTO_TEST_CASE(test_Capitalize)
     BOOST_CHECK_EQUAL(Capitalize("\x00\xfe\xff"), "\x00\xfe\xff");
 }
 
+static std::string SpanToStr(Span<const char>& span)
+{
+    return std::string(span.begin(), span.end());
+}
+
+BOOST_AUTO_TEST_CASE(test_spanparsing)
+{
+    using namespace spanparsing;
+    std::string input;
+    Span<const char> sp;
+    bool success;
+
+    // Const(...): parse a constant, update span to skip it if successful
+    input = "MilkToastHoney";
+    sp = MakeSpan(input);
+    success = Const("", sp); // empty
+    BOOST_CHECK(success);
+    BOOST_CHECK_EQUAL(SpanToStr(sp), "MilkToastHoney");
+
+    success = Const("Milk", sp);
+    BOOST_CHECK(success);
+    BOOST_CHECK_EQUAL(SpanToStr(sp), "ToastHoney");
+
+    success = Const("Bread", sp);
+    BOOST_CHECK(!success);
+
+    success = Const("Toast", sp);
+    BOOST_CHECK(success);
+    BOOST_CHECK_EQUAL(SpanToStr(sp), "Honey");
+
+    success = Const("Honeybadger", sp);
+    BOOST_CHECK(!success);
+
+    success = Const("Honey", sp);
+    BOOST_CHECK(success);
+    BOOST_CHECK_EQUAL(SpanToStr(sp), "");
+
+    // Func(...): parse a function call, update span to argument if successful
+    input = "Foo(Bar(xy,z()))";
+    sp = MakeSpan(input);
+
+    success = Func("FooBar", sp);
+    BOOST_CHECK(!success);
+
+    success = Func("Foo(", sp);
+    BOOST_CHECK(!success);
+
+    success = Func("Foo", sp);
+    BOOST_CHECK(success);
+    BOOST_CHECK_EQUAL(SpanToStr(sp), "Bar(xy,z())");
+
+    success = Func("Bar", sp);
+    BOOST_CHECK(success);
+    BOOST_CHECK_EQUAL(SpanToStr(sp), "xy,z()");
+
+    success = Func("xy", sp);
+    BOOST_CHECK(!success);
+
+    // Expr(...): return expression that span begins with, update span to skip it
+    Span<const char> result;
+
+    input = "(n*(n-1))/2";
+    sp = MakeSpan(input);
+    result = Expr(sp);
+    BOOST_CHECK_EQUAL(SpanToStr(result), "(n*(n-1))/2");
+    BOOST_CHECK_EQUAL(SpanToStr(sp), "");
+
+    input = "foo,bar";
+    sp = MakeSpan(input);
+    result = Expr(sp);
+    BOOST_CHECK_EQUAL(SpanToStr(result), "foo");
+    BOOST_CHECK_EQUAL(SpanToStr(sp), ",bar");
+
+    input = "(aaaaa,bbbbb()),c";
+    sp = MakeSpan(input);
+    result = Expr(sp);
+    BOOST_CHECK_EQUAL(SpanToStr(result), "(aaaaa,bbbbb())");
+    BOOST_CHECK_EQUAL(SpanToStr(sp), ",c");
+
+    input = "xyz)foo";
+    sp = MakeSpan(input);
+    result = Expr(sp);
+    BOOST_CHECK_EQUAL(SpanToStr(result), "xyz");
+    BOOST_CHECK_EQUAL(SpanToStr(sp), ")foo");
+
+    input = "((a),(b),(c)),xxx";
+    sp = MakeSpan(input);
+    result = Expr(sp);
+    BOOST_CHECK_EQUAL(SpanToStr(result), "((a),(b),(c))");
+    BOOST_CHECK_EQUAL(SpanToStr(sp), ",xxx");
+
+    // Split(...): split a string on every instance of sep, return vector
+    std::vector<Span<const char>> results;
+
+    input = "xxx";
+    results = Split(MakeSpan(input), 'x');
+    BOOST_CHECK_EQUAL(results.size(), 4);
+    BOOST_CHECK_EQUAL(SpanToStr(results[0]), "");
+    BOOST_CHECK_EQUAL(SpanToStr(results[1]), "");
+    BOOST_CHECK_EQUAL(SpanToStr(results[2]), "");
+    BOOST_CHECK_EQUAL(SpanToStr(results[3]), "");
+
+    input = "one#two#three";
+    results = Split(MakeSpan(input), '-');
+    BOOST_CHECK_EQUAL(results.size(), 1);
+    BOOST_CHECK_EQUAL(SpanToStr(results[0]), "one#two#three");
+
+    input = "one#two#three";
+    results = Split(MakeSpan(input), '#');
+    BOOST_CHECK_EQUAL(results.size(), 3);
+    BOOST_CHECK_EQUAL(SpanToStr(results[0]), "one");
+    BOOST_CHECK_EQUAL(SpanToStr(results[1]), "two");
+    BOOST_CHECK_EQUAL(SpanToStr(results[2]), "three");
+
+    input = "*foo*bar*";
+    results = Split(MakeSpan(input), '*');
+    BOOST_CHECK_EQUAL(results.size(), 4);
+    BOOST_CHECK_EQUAL(SpanToStr(results[0]), "");
+    BOOST_CHECK_EQUAL(SpanToStr(results[1]), "foo");
+    BOOST_CHECK_EQUAL(SpanToStr(results[2]), "bar");
+    BOOST_CHECK_EQUAL(SpanToStr(results[3]), "");
+}
+
 BOOST_AUTO_TEST_SUITE_END()
diff --git a/src/util/spanparsing.cpp b/src/util/spanparsing.cpp
new file mode 100644
index 0000000000..0c8575399a
--- /dev/null
+++ b/src/util/spanparsing.cpp
@@ -0,0 +1,67 @@
+// Copyright (c) 2018 The Bitcoin Core developers
+// Distributed under the MIT software license, see the accompanying
+// file COPYING or http://www.opensource.org/licenses/mit-license.php.
+
+#include <util/spanparsing.h>
+
+#include <span.h>
+
+#include <string>
+#include <vector>
+
+namespace spanparsing {
+
+bool Const(const std::string& str, Span<const char>& sp)
+{
+    if ((size_t)sp.size() >= str.size() && std::equal(str.begin(), str.end(), sp.begin())) {
+        sp = sp.subspan(str.size());
+        return true;
+    }
+    return false;
+}
+
+bool Func(const std::string& str, Span<const char>& sp)
+{
+    if ((size_t)sp.size() >= str.size() + 2 && sp[str.size()] == '(' && sp[sp.size() - 1] == ')' && std::equal(str.begin(), str.end(), sp.begin())) {
+        sp = sp.subspan(str.size() + 1, sp.size() - str.size() - 2);
+        return true;
+    }
+    return false;
+}
+
+Span<const char> Expr(Span<const char>& sp)
+{
+    int level = 0;
+    auto it = sp.begin();
+    while (it != sp.end()) {
+        if (*it == '(') {
+            ++level;
+        } else if (level && *it == ')') {
+            --level;
+        } else if (level == 0 && (*it == ')' || *it == ',')) {
+            break;
+        }
+        ++it;
+    }
+    Span<const char> ret = sp.first(it - sp.begin());
+    sp = sp.subspan(it - sp.begin());
+    return ret;
+}
+
+std::vector<Span<const char>> Split(const Span<const char>& sp, char sep)
+{
+    std::vector<Span<const char>> ret;
+    auto it = sp.begin();
+    auto start = it;
+    while (it != sp.end()) {
+        if (*it == sep) {
+            ret.emplace_back(start, it);
+            start = it + 1;
+        }
+        ++it;
+    }
+    ret.emplace_back(start, it);
+    return ret;
+}
+
+} // namespace spanparsing
diff --git a/src/util/spanparsing.h b/src/util/spanparsing.h
new file mode 100644
index 0000000000..63f54758bd
--- /dev/null
+++ b/src/util/spanparsing.h
@@ -0,0 +1,50 @@
+// Copyright (c) 2018 The Bitcoin Core developers
+// Distributed under the MIT software license, see the accompanying
+// file COPYING or http://www.opensource.org/licenses/mit-license.php.
+
+#ifndef BITCOIN_UTIL_SPANPARSING_H
+#define BITCOIN_UTIL_SPANPARSING_H
+
+#include <span.h>
+
+#include <string>
+#include <vector>
+
+namespace spanparsing {
+
+/** Parse a constant.
+ *
+ * If sp's initial part matches str, sp is updated to skip that part, and true is returned.
+ * Otherwise sp is unmodified and false is returned.
+ */
+bool Const(const std::string& str, Span<const char>& sp);
+
+/** Parse a function call.
+ *
+ * If sp's initial part matches str + "(", and sp ends with ")", sp is updated to be the
+ * section between the braces, and true is returned. Otherwise sp is unmodified and false
+ * is returned.
+ */
+bool Func(const std::string& str, Span<const char>& sp);
+
+/** Extract the expression that sp begins with.
+ *
+ * This function will return the initial part of sp, up to (but not including) the first
+ * comma or closing brace, skipping ones that are surrounded by braces. So for example,
+ * for "foo(bar(1),2),3" the initial part "foo(bar(1),2)" will be returned. sp will be
+ * updated to skip the initial part that is returned.
+ */
+Span<const char> Expr(Span<const char>& sp);
+
+/** Split a string on every instance of sep, returning a vector.
+ *
+ * If sep does not occur in sp, a singleton with the entirety of sp is returned.
+ *
+ * Note that this function does not care about braces, so splitting
+ * "foo(bar(1),2),3) on ',' will return {"foo(bar(1)", "2)", "3)"}.
+ */
+std::vector<Span<const char>> Split(const Span<const char>& sp, char sep);
+
+} // namespace spanparsing
+
+#endif // BITCOIN_UTIL_SPANPARSING_H