util: move spanparsing.h to script/parsing.h

Move miniscript / descriptor script parsing functions out of util library so
they are not a dependency of the kernel.

There are no changes to code or behavior.

src/script/descriptor.cpp

@@ -8,6 +8,7 @@
 #include <key_io.h>
 #include <pubkey.h>
 #include <script/miniscript.h>
+#include <script/parsing.h>
 #include <script/script.h>
 #include <script/signingprovider.h>
 #include <script/solver.h>
@@ -17,7 +18,6 @@
 #include <span.h>
 #include <util/bip32.h>
 #include <util/check.h>
-#include <util/spanparsing.h>
 #include <util/strencodings.h>
 #include <util/vector.h>
 
@@ -1350,8 +1350,6 @@ enum class ParseScriptContext {
 /** Parse a public key that excludes origin information. */
 std::unique_ptr<PubkeyProvider> ParsePubkeyInner(uint32_t key_exp_index, const Span<const char>& sp, ParseScriptContext ctx, FlatSigningProvider& out, bool& apostrophe, std::string& error)
 {
-    using namespace spanparsing;
-
     bool permit_uncompressed = ctx == ParseScriptContext::TOP || ctx == ParseScriptContext::P2SH;
     auto split = Split(sp, '/');
     std::string str(split[0].begin(), split[0].end());
@@ -1424,8 +1422,6 @@ std::unique_ptr<PubkeyProvider> ParsePubkeyInner(uint32_t key_exp_index, const S
 /** Parse a public key including origin information (if enabled). */
 std::unique_ptr<PubkeyProvider> ParsePubkey(uint32_t key_exp_index, const Span<const char>& sp, ParseScriptContext ctx, 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";
@@ -1589,7 +1585,7 @@ struct KeyParser {
 // NOLINTNEXTLINE(misc-no-recursion)
 std::unique_ptr<DescriptorImpl> ParseScript(uint32_t& key_exp_index, Span<const char>& sp, ParseScriptContext ctx, FlatSigningProvider& out, std::string& error)
 {
-    using namespace spanparsing;
+    using namespace script;
 
     auto expr = Expr(sp);
     if (Func("pk", expr)) {
@@ -2038,8 +2034,6 @@ 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";

src/script/miniscript.h

@@ -18,10 +18,10 @@
 
 #include <policy/policy.h>
 #include <primitives/transaction.h>
+#include <script/parsing.h>
 #include <script/script.h>
 #include <span.h>
 #include <util/check.h>
-#include <util/spanparsing.h>
 #include <util/strencodings.h>
 #include <util/string.h>
 #include <util/vector.h>
@@ -1764,7 +1764,7 @@ void BuildBack(const MiniscriptContext script_ctx, Fragment nt, std::vector<Node
 template<typename Key, typename Ctx>
 inline NodeRef<Key> Parse(Span<const char> in, const Ctx& ctx)
 {
-    using namespace spanparsing;
+    using namespace script;
 
     // Account for the minimum script size for all parsed fragments so far. It "borrows" 1
     // script byte from all leaf nodes, counting it instead whenever a space for a recursive

src/script/parsing.cpp

@@ -0,0 +1,52 @@
+// Copyright (c) 2018-2022 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 <script/parsing.h>
+
+#include <span.h>
+
+#include <algorithm>
+#include <cstddef>
+#include <string>
+
+namespace script {
+
+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 == '(' || *it == '{') {
+            ++level;
+        } else if (level && (*it == ')' || *it == '}')) {
+            --level;
+        } else if (level == 0 && (*it == ')' || *it == '}' || *it == ',')) {
+            break;
+        }
+        ++it;
+    }
+    Span<const char> ret = sp.first(it - sp.begin());
+    sp = sp.subspan(it - sp.begin());
+    return ret;
+}
+
+} // namespace script

src/script/parsing.h

@@ -0,0 +1,40 @@
+// Copyright (c) 2018-2022 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_SCRIPT_PARSING_H
+#define BITCOIN_SCRIPT_PARSING_H
+
+#include <span.h>
+
+#include <string>
+
+namespace script {
+
+/** 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);
+
+} // namespace script
+
+#endif // BITCOIN_SCRIPT_PARSING_H