Compact proof utilities in sp_trie. (#8574)

* validation extension in sp_io

* need paths

* arc impl

* missing host function in executor

* io to pkdot

* decode function.

* encode primitive.

* trailing tab

* multiple patch

* fix child trie logic

* restore master versionning

* bench compact proof size

* trie-db 22.3 is needed

* line width

* split line

* fixes for bench (additional root may not be needed as original issue was
with empty proof).

* revert compact from block size calculation.

* New error type for compression.

* Adding test (incomplete (failing)).
Also lacking real proof checking (no good primitives in sp-trie crate).

* There is currently no proof recording utility in sp_trie, removing
test.

* small test of child root in proof without a child proof.

* remove empty test.

* remove non compact proof size

* Missing revert.

* proof method to encode decode.

client/db/src/bench.rs

@@ -117,6 +117,7 @@ pub struct BenchmarkingState<B: BlockT> {
 	read_write_tracker: RefCell<ReadWriteTracker>,
 	whitelist: RefCell<Vec<TrackedStorageKey>>,
 	proof_recorder: Option<ProofRecorder<B::Hash>>,
+	proof_recorder_root: Cell<B::Hash>,
 }
 
 impl<B: BlockT> BenchmarkingState<B> {
@@ -129,7 +130,7 @@ impl<B: BlockT> BenchmarkingState<B> {
 		let mut state = BenchmarkingState {
 			state: RefCell::new(None),
 			db: Cell::new(None),
-			root: Cell::new(root),
+			root: Cell::new(root.clone()),
 			genesis: Default::default(),
 			genesis_root: Default::default(),
 			record: Default::default(),
@@ -139,6 +140,7 @@ impl<B: BlockT> BenchmarkingState<B> {
 			read_write_tracker: Default::default(),
 			whitelist: Default::default(),
 			proof_recorder: record_proof.then(Default::default),
+			proof_recorder_root: Cell::new(root.clone()),
 		};
 
 		state.add_whitelist_to_tracker();
@@ -166,7 +168,10 @@ impl<B: BlockT> BenchmarkingState<B> {
 			None => Arc::new(kvdb_memorydb::create(1)),
 		};
 		self.db.set(Some(db.clone()));
-		self.proof_recorder.as_ref().map(|r| r.reset());
+		if let Some(recorder) = &self.proof_recorder {
+			recorder.reset();
+			self.proof_recorder_root.set(self.root.get());
+		}
 		let storage_db = Arc::new(StorageDb::<B> {
 			db,
 			proof_recorder: self.proof_recorder.clone(),
@@ -516,7 +521,27 @@ impl<B: BlockT> StateBackend<HashFor<B>> for BenchmarkingState<B> {
 	}
 
 	fn proof_size(&self) -> Option<u32> {
-		self.proof_recorder.as_ref().map(|recorder| recorder.estimate_encoded_size() as u32)
+		self.proof_recorder.as_ref().map(|recorder| {
+			let proof_size = recorder.estimate_encoded_size() as u32;
+			let proof = recorder.to_storage_proof();
+			let proof_recorder_root = self.proof_recorder_root.get();
+			if proof_recorder_root == Default::default() || proof_size == 1 {
+				// empty trie
+				proof_size
+			} else {
+				if let Some(size) = proof.encoded_compact_size::<HashFor<B>>(proof_recorder_root) {
+					size as u32
+				} else {
+					panic!(
+						"proof rec root {:?}, root {:?}, genesis {:?}, rec_len {:?}",
+						self.proof_recorder_root.get(),
+						self.root.get(),
+						self.genesis_root,
+						proof_size,
+					);
+				}
+			}
+		})
 	}
 }
 

primitives/state-machine/src/lib.rs

@@ -1402,14 +1402,22 @@ mod tests {
 		}
 	}
 
+	fn test_compact(remote_proof: StorageProof, remote_root: &sp_core::H256) -> StorageProof {
+		let compact_remote_proof = remote_proof.into_compact_proof::<BlakeTwo256>(
+			remote_root.clone(),
+		).unwrap();
+		compact_remote_proof.to_storage_proof::<BlakeTwo256>(Some(remote_root)).unwrap().0
+	}
+
 	#[test]
 	fn prove_read_and_proof_check_works() {
 		let child_info = ChildInfo::new_default(b"sub1");
 		let child_info = &child_info;
 		// fetch read proof from 'remote' full node
 		let remote_backend = trie_backend::tests::test_trie();
-		let remote_root = remote_backend.storage_root(::std::iter::empty()).0;
+		let remote_root = remote_backend.storage_root(std::iter::empty()).0;
 		let remote_proof = prove_read(remote_backend, &[b"value2"]).unwrap();
+		let remote_proof = test_compact(remote_proof, &remote_root);
  		// check proof locally
 		let local_result1 = read_proof_check::<BlakeTwo256, _>(
 			remote_root,
@@ -1429,12 +1437,13 @@ mod tests {
 		assert_eq!(local_result2, false);
 		// on child trie
 		let remote_backend = trie_backend::tests::test_trie();
-		let remote_root = remote_backend.storage_root(::std::iter::empty()).0;
+		let remote_root = remote_backend.storage_root(std::iter::empty()).0;
 		let remote_proof = prove_child_read(
 			remote_backend,
 			child_info,
 			&[b"value3"],
 		).unwrap();
+		let remote_proof = test_compact(remote_proof, &remote_root);
 		let local_result1 = read_child_proof_check::<BlakeTwo256, _>(
 			remote_root,
 			remote_proof.clone(),
@@ -1457,6 +1466,50 @@ mod tests {
 		);
 	}
 
+	#[test]
+	fn compact_multiple_child_trie() {
+		// this root will be queried
+		let child_info1 = ChildInfo::new_default(b"sub1");
+		// this root will not be include in proof
+		let child_info2 = ChildInfo::new_default(b"sub2");
+		// this root will be include in proof
+		let child_info3 = ChildInfo::new_default(b"sub");
+		let mut remote_backend = trie_backend::tests::test_trie();
+		let (remote_root, transaction) = remote_backend.full_storage_root(
+			std::iter::empty(),
+			vec![
+				(&child_info1, vec![
+					(&b"key1"[..], Some(&b"val2"[..])),
+					(&b"key2"[..], Some(&b"val3"[..])),
+				].into_iter()),
+				(&child_info2, vec![
+					(&b"key3"[..], Some(&b"val4"[..])),
+					(&b"key4"[..], Some(&b"val5"[..])),
+				].into_iter()),
+				(&child_info3, vec![
+					(&b"key5"[..], Some(&b"val6"[..])),
+					(&b"key6"[..], Some(&b"val7"[..])),
+				].into_iter()),
+			].into_iter(),
+		);
+		remote_backend.backend_storage_mut().consolidate(transaction);
+		remote_backend.essence.set_root(remote_root.clone());
+		let remote_proof = prove_child_read(
+			remote_backend,
+			&child_info1,
+			&[b"key1"],
+		).unwrap();
+		let remote_proof = test_compact(remote_proof, &remote_root);
+		let local_result1 = read_child_proof_check::<BlakeTwo256, _>(
+			remote_root,
+			remote_proof.clone(),
+			&child_info1,
+			&[b"key1"],
+		).unwrap();
+		assert_eq!(local_result1.len(), 1);
+		assert_eq!(local_result1.get(&b"key1"[..]), Some(&Some(b"val2".to_vec())));
+	}
+
 	#[test]
 	fn child_storage_uuid() {
 

primitives/trie/Cargo.toml

@@ -21,7 +21,7 @@ harness = false
 codec = { package = "parity-scale-codec", version = "2.0.0", default-features = false }
 sp-std = { version = "3.0.0", default-features = false, path = "../std" }
 hash-db = { version = "0.15.2", default-features = false }
-trie-db = { version = "0.22.2", default-features = false }
+trie-db = { version = "0.22.3", default-features = false }
 trie-root = { version = "0.16.0", default-features = false }
 memory-db = { version = "0.26.0", default-features = false }
 sp-core = { version = "3.0.0", default-features = false, path = "../core" }

primitives/trie/src/error.rs

@@ -26,7 +26,7 @@ pub enum Error {
 	/// Bad format.
 	BadFormat,
 	/// Decoding error.
-	Decode(codec::Error)
+	Decode(codec::Error),
 }
 
 impl From<codec::Error> for Error {

primitives/trie/src/lib.rs

@@ -23,6 +23,7 @@ mod error;
 mod node_header;
 mod node_codec;
 mod storage_proof;
+mod trie_codec;
 mod trie_stream;
 
 use sp_std::{boxed::Box, marker::PhantomData, vec::Vec, borrow::Borrow};
@@ -35,7 +36,7 @@ pub use error::Error;
 pub use trie_stream::TrieStream;
 /// The Substrate format implementation of `NodeCodec`.
 pub use node_codec::NodeCodec;
-pub use storage_proof::StorageProof;
+pub use storage_proof::{StorageProof, CompactProof};
 /// Various re-exports from the `trie-db` crate.
 pub use trie_db::{
 	Trie, TrieMut, DBValue, Recorder, CError, Query, TrieLayout, TrieConfiguration, nibble_ops, TrieDBIterator,
@@ -45,6 +46,9 @@ pub use memory_db::KeyFunction;
 pub use memory_db::prefixed_key;
 /// Various re-exports from the `hash-db` crate.
 pub use hash_db::{HashDB as HashDBT, EMPTY_PREFIX};
+/// Trie codec reexport, mainly child trie support
+/// for trie compact proof.
+pub use trie_codec::{decode_compact, encode_compact, Error as CompactProofError};
 
 #[derive(Default)]
 /// substrate trie layout

primitives/trie/src/storage_proof.rs

@@ -31,6 +31,12 @@ pub struct StorageProof {
 	trie_nodes: Vec<Vec<u8>>,
 }
 
+/// Storage proof in compact form.
+#[derive(Debug, PartialEq, Eq, Clone, Encode, Decode)]
+pub struct CompactProof {
+	pub encoded_nodes: Vec<Vec<u8>>,
+}
+
 impl StorageProof {
 	/// Constructs a storage proof from a subset of encoded trie nodes in a storage backend.
 	pub fn new(trie_nodes: Vec<Vec<u8>>) -> Self {
@@ -79,6 +85,56 @@ impl StorageProof {
 
 		Self { trie_nodes }
 	}
+
+	/// Encode as a compact proof with default
+	/// trie layout.
+	pub fn into_compact_proof<H: Hasher>(
+		self,
+		root: H::Out,
+	) -> Result<CompactProof, crate::CompactProofError<crate::Layout<H>>> {
+		crate::encode_compact::<crate::Layout<H>>(self, root)
+	}
+
+	/// Returns the estimated encoded size of the compact proof.
+	///
+	/// Runing this operation is a slow operation (build the whole compact proof) and should only be
+	/// in non sensitive path.
+	/// Return `None` on error.
+	pub fn encoded_compact_size<H: Hasher>(self, root: H::Out) -> Option<usize> {
+		let compact_proof = self.into_compact_proof::<H>(root);
+		compact_proof.ok().map(|p| p.encoded_size())
+	}
+
+}
+
+impl CompactProof {
+	/// Return an iterator on the compact encoded nodes.
+	pub fn iter_compact_encoded_nodes(&self) -> impl Iterator<Item = &[u8]> {
+		self.encoded_nodes.iter().map(Vec::as_slice)
+	}
+
+	/// Decode to a full storage_proof.
+	///
+	/// Method use a temporary `HashDB`, and `sp_trie::decode_compact`
+	/// is often better.
+	pub fn to_storage_proof<H: Hasher>(
+		&self,
+		expected_root: Option<&H::Out>,
+	) -> Result<(StorageProof, H::Out), crate::CompactProofError<crate::Layout<H>>> {
+		let mut db = crate::MemoryDB::<H>::new(&[]);
+		let root = crate::decode_compact::<crate::Layout<H>, _, _>(
+			&mut db,
+			self.iter_compact_encoded_nodes(),
+			expected_root,
+		)?;
+		Ok((StorageProof::new(db.drain().into_iter().filter_map(|kv|
+			if (kv.1).1 > 0 {
+				Some((kv.1).0)
+			} else {
+				None
+			}
+		).collect()), root))
+	}
 }
 
 /// An iterator over trie nodes constructed from a storage proof. The nodes are not guaranteed to