From bc4f57f53d03e2243cd5a41c87ac3976faf811fc Mon Sep 17 00:00:00 2001
From: Luke Dashjr <luke-jr+git@utopios.org>
Date: Wed, 19 Jun 2024 21:01:37 +0000
Subject: [PATCH] bisecting...

---
 .cirrus.yml               |  3 ++-
 src/script/script.cpp     | 50 +++++++++++++++++++++++++++++++++++++++
 src/script/script.h       |  2 ++
 src/test/script_tests.cpp | 29 +++++++++++++++++++++++
 4 files changed, 83 insertions(+), 1 deletion(-)

diff --git a/.cirrus.yml b/.cirrus.yml
index 6628d0bbbbb0f0..02db3e4c6d58e5 100644
--- a/.cirrus.yml
+++ b/.cirrus.yml
@@ -48,7 +48,7 @@ env:  # Global defaults
 
 # https://cirrus-ci.org/guide/tips-and-tricks/#sharing-configuration-between-tasks
 filter_template: &FILTER_TEMPLATE
-  skip: $CIRRUS_REPO_FULL_NAME == "bitcoin-core/gui" && $CIRRUS_PR == ""  # No need to run on the read-only mirror, unless it is a PR. https://cirrus-ci.org/guide/writing-tasks/#conditional-task-execution
+  skip: true
   stateful: false  # https://cirrus-ci.org/guide/writing-tasks/#stateful-tasks
 
 base_template: &BASE_TEMPLATE
@@ -184,6 +184,7 @@ task:
 task:
   name: 'multiprocess, i686, DEBUG'
   << : *GLOBAL_TASK_TEMPLATE
+  skip: false
   persistent_worker:
     labels:
       type: medium
diff --git a/src/script/script.cpp b/src/script/script.cpp
index 80e8d26bcfb978..a3d13b9845a39a 100644
--- a/src/script/script.cpp
+++ b/src/script/script.cpp
@@ -286,6 +286,56 @@ bool CScript::HasValidOps() const
     return true;
 }
 
+std::pair<size_t, size_t> CScript::DatacarrierBytes() const
+{
+    size_t counted{0};
+    opcodetype opcode, last_opcode{OP_INVALIDOPCODE};
+    std::vector<unsigned char> push_data;
+    unsigned int inside_noop{0}, inside_conditional{0};
+    CScript::const_iterator opcode_it = begin(), data_began = begin();
+    for (CScript::const_iterator it = begin(); it < end(); last_opcode = opcode) {
+        opcode_it = it;
+        if (!GetOp(it, opcode, push_data)) {
+            // Invalid scripts are necessarily all data
+            return {0, size()};
+        }
+
+        if (opcode == OP_IF || opcode == OP_NOTIF) {
+            ++inside_conditional;
+        } else if (opcode == OP_ENDIF) {
+            if (!inside_conditional) return {0, size()};  // invalid
+            --inside_conditional;
+        } else if (opcode == OP_RETURN && !inside_conditional) {
+            // unconditional OP_RETURN is unspendable
+            return {size(), 0};
+        }
+
+        // Match OP_FALSE OP_IF
+        if (inside_noop) {
+            switch (opcode) {
+            case OP_IF: case OP_NOTIF:
+                ++inside_noop;
+                break;
+            case OP_ENDIF:
+                if (0 == --inside_noop) {
+                    counted += it - data_began + 1;
+                }
+                break;
+            default: /* do nothing */;
+            }
+        } else if (opcode == OP_IF && last_opcode == OP_FALSE) {
+            inside_noop = 1;
+            data_began = opcode_it;
+        // Match <data> OP_DROP
+        } else if (opcode <= OP_PUSHDATA4) {
+            data_began = opcode_it;
+        } else if (opcode == OP_DROP && last_opcode <= OP_PUSHDATA4) {
+            counted += it - data_began;
+        }
+    }
+    return {0, counted};
+}
+
 bool GetScriptOp(CScriptBase::const_iterator& pc, CScriptBase::const_iterator end, opcodetype& opcodeRet, std::vector<unsigned char>* pvchRet)
 {
     opcodeRet = OP_INVALIDOPCODE;
diff --git a/src/script/script.h b/src/script/script.h
index 66d63fae89e12e..dbbed45c6004df 100644
--- a/src/script/script.h
+++ b/src/script/script.h
@@ -554,6 +554,8 @@ class CScript : public CScriptBase
         return (size() > 0 && *begin() == OP_RETURN) || (size() > MAX_SCRIPT_SIZE);
     }
 
+    std::pair<size_t, size_t> DatacarrierBytes() const;
+
     void clear()
     {
         // The default prevector::clear() does not release memory
diff --git a/src/test/script_tests.cpp b/src/test/script_tests.cpp
index 1f674408b2c667..5682a95ee852f9 100644
--- a/src/test/script_tests.cpp
+++ b/src/test/script_tests.cpp
@@ -1464,6 +1464,35 @@ BOOST_AUTO_TEST_CASE(script_HasValidOps)
     BOOST_CHECK(!script.HasValidOps());
 }
 
+static std::string DatacarrierBytesStr(const CScript &script) {
+    auto dcb = script.DatacarrierBytes();
+    return strprintf("%s+%s", dcb.first, dcb.second);
+}
+
+BOOST_AUTO_TEST_CASE(script_DataCarrierBytes)
+{
+    using zeros = std::vector<unsigned char>;
+
+    // empty script
+    BOOST_CHECK_EQUAL("0+0", DatacarrierBytesStr(CScript()));
+    // series of pushes are not data
+    BOOST_CHECK_EQUAL("0+0", DatacarrierBytesStr(CScript() << OP_0 << OP_0 << OP_0));
+    // unspendable if first op is OP_RETURN, then length(1), zeros(11)
+    BOOST_CHECK_EQUAL("13+0", DatacarrierBytesStr(CScript() << OP_RETURN << zeros(11)));
+    // invalid script (no data following PUSHDATA) makes it all data
+    BOOST_CHECK_EQUAL("0+2", DatacarrierBytesStr(CScript() << OP_0 << OP_PUSHDATA4));
+    // no data here
+    BOOST_CHECK_EQUAL("0+0", DatacarrierBytesStr(CScript() << OP_TRUE << OP_IF << OP_ENDIF));
+    // specific data pattern, entire script is data
+    BOOST_CHECK_EQUAL("0+4", DatacarrierBytesStr(CScript() << OP_FALSE << OP_IF << OP_7 << OP_ENDIF));
+    // consecutive data
+    BOOST_CHECK_EQUAL("0+6", DatacarrierBytesStr(CScript() << OP_FALSE << OP_IF << OP_ENDIF << OP_FALSE << OP_IF << OP_ENDIF));
+    // nested data (all is data)
+    BOOST_CHECK_EQUAL("0+6", DatacarrierBytesStr(CScript() << OP_FALSE << OP_IF << OP_TRUE << OP_IF << OP_ENDIF << OP_ENDIF));
+    // pushing then immediately dropping is data: length(1), zero(11), OP_DROP
+    BOOST_CHECK_EQUAL("0+13", DatacarrierBytesStr(CScript() << zeros(11) << OP_DROP));
+}
+
 static CMutableTransaction TxFromHex(const std::string& str)
 {
     CMutableTransaction tx;