assumeutxo: Add documentation on dumptxoutset serialization format

src/rpc/blockchain.cpp

@@ -2677,6 +2677,13 @@ UniValue CreateUTXOSnapshot(
     unsigned int iter{0};
     std::vector<std::pair<uint32_t, Coin>> coins;
 
+    // To reduce space the serialization format of the snapshot avoids
+    // dublication of tx hashes. The code takes advantage of the guarantee by
+    // leveldb that keys are lexicographically sorted.
+    // In the coins vector we collect all coins that belong to a certain tx hash
+    // (key.hash) and when we have them all (key.hash != last_hash) we write
+    // them to file using the below lamda function.
+    // See also https://github.com/bitcoin/bitcoin/issues/25675
     auto write_coins_to_file = [&](AutoFile& afile, const uint256& last_hash, const std::vector<std::pair<uint32_t, Coin>>& coins) {
         afile << last_hash;
         WriteCompactSize(afile, coins.size());

src/validation.h

@@ -883,6 +883,12 @@ class ChainstateManager
     CBlockIndex* m_best_invalid GUARDED_BY(::cs_main){nullptr};
 
     //! Internal helper for ActivateSnapshot().
+    //!
+    //! De-serialization of a snapshot that is created with
+    //! CreateUTXOSnapshot() in rpc/blockchain.cpp.
+    //! To reduce space the serialization format of the snapshot avoids
+    //! dublication of tx hashes. The code takes advantage of the guarantee by
+    //! leveldb that keys are lexicographically sorted.
     [[nodiscard]] bool PopulateAndValidateSnapshot(
         Chainstate& snapshot_chainstate,
         AutoFile& coins_file,