Add P23 data corruption verification tool. (#4969)

# Description

Add P23 data corruption verification tool.

This allows specifying an external CSV table that contains the expected
p23 Hot Archive corruption data. Then during catchup that covers the
whole range of p23 and the upgrade to p24 the file is used to verify
that:

- Only the entries from the table are ever incorrectly archived, and
that their correct and archived states match those in the table
- Every entry from the table has indeed been incorrectly archived
- The entries that have been marked as restored in the table were indeed
restored with the expected corrupted state
- During the protocol 24 upgrade only the entries from the table that
have never been restored have been updated, and that the update has
brought them back to the correct state

# Checklist
- [ ] Reviewed the
[contributing](https://github.com/stellar/stellar-core/blob/master/CONTRIBUTING.md#submitting-changes)
document
- [ ] Rebased on top of master (no merge commits)
- [ ] Ran `clang-format` v8.0.0 (via `make format` or the Visual Studio
extension)
- [ ] Compiles
- [ ] Ran all tests
- [ ] If change impacts performance, include supporting evidence per the
[performance
document](https://github.com/stellar/stellar-core/blob/master/performance-eval/performance-eval.md)

Author: dmkozh

src/ledger/LedgerManagerImpl.cpp

@@ -2787,6 +2787,15 @@ LedgerManagerImpl::sealLedgerTxnAndTransferEntriesToBucketList(
                     mApp.getBucketManager().addHotArchiveBatch(
                         mApp, lh, evictedState.archivedEntries,
                         restoredEntries);
+                    // Validate evicted entries against Protocol 23 corruption
+                    // data if configured
+                    if (mApp.getProtocol23CorruptionDataVerifier())
+                    {
+                        mApp.getProtocol23CorruptionDataVerifier()
+                            ->verifyArchivalOfCorruptedEntry(evictedState, mApp,
+                                                             lh.ledgerSeq,
+                                                             lh.ledgerVersion);
+                    }
                 }
             }
 
@@ -2803,9 +2812,9 @@ LedgerManagerImpl::sealLedgerTxnAndTransferEntriesToBucketList(
         // Subtle: we snapshot the state size *before* flushing the updated
         // entries into in-memory state (doing that after would be really
         // tricky, as we seal LTX before flushing). So the snapshot taken at
-        // ledger `N` will have the state size for ledger `N - 1`. That doesn't
-        // really change anything for the size accounting, but is important to
-        // maintain as a protocol implementation detail.
+        // ledger `N` will have the state size for ledger `N - 1`. That
+        // doesn't really change anything for the size accounting, but is
+        // important to maintain as a protocol implementation detail.
         mApplyState.getSorobanNetworkConfigForCommit()
             .maybeSnapshotSorobanStateSize(
                 lh.ledgerSeq, mApplyState.getSorobanInMemoryStateSize(), ltx,