Merge bitcoin/bitcoin#32185: coins: replace manual CDBBatch size estimation with LevelDB's native ApproximateSize

e419b0e refactor: Remove manual CDBBatch size estimation (Lőrinc)
8b5e19d refactor: Delegate to LevelDB for CDBBatch size estimation (Lőrinc)
751077c Coins: Add `kHeader` to `CDBBatch::size_estimate` (Lőrinc)

Pull request description:

  ### Summary

  The manual batch size estimation of `CDBBatch` serialized size was [added](bitcoin/bitcoin@e66dbde) when LevelDB [didn't expose this functionality  yet](google/leveldb@69e2bd2).
  The PR refactors the logic to use the native `leveldb::WriteBatch::ApproximateSize()` function, structured in 3 focused commits to incrementally replace the old behavior safely.

  ### Context

  The previous manual size calculation initialized the estimate to 0, instead of LevelDB's header size (containing an 8-byte sequence number followed by a 4-byte count).
  This PR corrects that and transitions to the now-available native LevelDB function for improved accuracy and maintainability.

  ### Approach
  The fix and refactor follow a strangle pattern over three commits:
  * correct the initialization bug in the existing manual calculation, isolating the fix and ensuring the subsequent assertions use the corrected logic;
  * introduce the native `ApproximateSize()` method alongside the corrected manual one, adding assertions to verify their equivalence at runtime;
  * remove the verified manual calculation logic and assertions, leaving only the native method.

ACKs for top commit:
  sipa:
    utACK e419b0e
  TheCharlatan:
    ACK e419b0e
  laanwj:
    Code review ACK e419b0e

Tree-SHA512: a12b973dd480d4ffec4ec89a119bf0b6f73bde4e634329d6e4cc3454b867f2faf3742b78ec4a3b6d98ac4fb28fb2174f44ede42d6c701ed871987a7274560691

Author: ryanofsky

src/dbwrapper.cpp

@@ -158,14 +158,16 @@ struct CDBBatch::WriteBatchImpl {
 
 CDBBatch::CDBBatch(const CDBWrapper& _parent)
     : parent{_parent},
-      m_impl_batch{std::make_unique<CDBBatch::WriteBatchImpl>()} {};
+      m_impl_batch{std::make_unique<CDBBatch::WriteBatchImpl>()}
+{
+    Clear();
+};
 
 CDBBatch::~CDBBatch() = default;
 
 void CDBBatch::Clear()
 {
     m_impl_batch->batch.Clear();
-    size_estimate = 0;
 }
 
 void CDBBatch::WriteImpl(std::span<const std::byte> key, DataStream& ssValue)
@@ -174,26 +176,17 @@ void CDBBatch::WriteImpl(std::span<const std::byte> key, DataStream& ssValue)
     ssValue.Xor(dbwrapper_private::GetObfuscateKey(parent));
     leveldb::Slice slValue(CharCast(ssValue.data()), ssValue.size());
     m_impl_batch->batch.Put(slKey, slValue);
-    // LevelDB serializes writes as:
-    // - byte: header
-    // - varint: key length (1 byte up to 127B, 2 bytes up to 16383B, ...)
-    // - byte[]: key
-    // - varint: value length
-    // - byte[]: value
-    // The formula below assumes the key and value are both less than 16k.
-    size_estimate += 3 + (slKey.size() > 127) + slKey.size() + (slValue.size() > 127) + slValue.size();
 }
 
 void CDBBatch::EraseImpl(std::span<const std::byte> key)
 {
     leveldb::Slice slKey(CharCast(key.data()), key.size());
     m_impl_batch->batch.Delete(slKey);
-    // LevelDB serializes erases as:
-    // - byte: header
-    // - varint: key length
-    // - byte[]: key
-    // The formula below assumes the key is less than 16kB.
-    size_estimate += 2 + (slKey.size() > 127) + slKey.size();
+}
+
+size_t CDBBatch::ApproximateSize() const
+{
+    return m_impl_batch->batch.ApproximateSize();
 }
 
 struct LevelDBContext {

src/dbwrapper.h

@@ -83,8 +83,6 @@ class CDBBatch
     DataStream ssKey{};
     DataStream ssValue{};
 
-    size_t size_estimate{0};
-
     void WriteImpl(std::span<const std::byte> key, DataStream& ssValue);
     void EraseImpl(std::span<const std::byte> key);
 
@@ -117,7 +115,7 @@ class CDBBatch
         ssKey.clear();
     }
 
-    size_t SizeEstimate() const { return size_estimate; }
+    size_t ApproximateSize() const;
 };
 
 class CDBIterator

src/txdb.cpp

@@ -119,16 +119,19 @@ bool CCoinsViewDB::BatchWrite(CoinsViewCacheCursor& cursor, const uint256 &hashB
     for (auto it{cursor.Begin()}; it != cursor.End();) {
         if (it->second.IsDirty()) {
             CoinEntry entry(&it->first);
-            if (it->second.coin.IsSpent())
+            if (it->second.coin.IsSpent()) {
                 batch.Erase(entry);
-            else
+            } else {
                 batch.Write(entry, it->second.coin);
+            }
+
             changed++;
         }
         count++;
         it = cursor.NextAndMaybeErase(*it);
-        if (batch.SizeEstimate() > m_options.batch_write_bytes) {
-            LogDebug(BCLog::COINDB, "Writing partial batch of %.2f MiB\n", batch.SizeEstimate() * (1.0 / 1048576.0));
+        if (batch.ApproximateSize() > m_options.batch_write_bytes) {
+            LogDebug(BCLog::COINDB, "Writing partial batch of %.2f MiB\n", batch.ApproximateSize() * (1.0 / 1048576.0));
+
             m_db->WriteBatch(batch);
             batch.Clear();
             if (m_options.simulate_crash_ratio) {
@@ -145,7 +148,7 @@ bool CCoinsViewDB::BatchWrite(CoinsViewCacheCursor& cursor, const uint256 &hashB
     batch.Erase(DB_HEAD_BLOCKS);
     batch.Write(DB_BEST_BLOCK, hashBlock);
 
-    LogDebug(BCLog::COINDB, "Writing final batch of %.2f MiB\n", batch.SizeEstimate() * (1.0 / 1048576.0));
+    LogDebug(BCLog::COINDB, "Writing final batch of %.2f MiB\n", batch.ApproximateSize() * (1.0 / 1048576.0));
     bool ret = m_db->WriteBatch(batch);
     LogDebug(BCLog::COINDB, "Committed %u changed transaction outputs (out of %u) to coin database...\n", (unsigned int)changed, (unsigned int)count);
     return ret;