solana/nft_bridge: allow invalid utf-8 metadata

Change-Id: If2fba5d394b8afab34ed0f12714851d0121a2924

solana/modules/nft_bridge/Cargo.lock

@@ -264,6 +264,17 @@ version = "0.3.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "476e9cd489f9e121e02ffa6014a8ef220ecb15c05ed23fc34cca13925dc283fb"
 
+[[package]]
+name = "bstr"
+version = "0.2.16"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "90682c8d613ad3373e66de8c6411e0ae2ab2571e879d2efbf73558cc66f21279"
+dependencies = [
+ "lazy_static",
+ "memchr",
+ "regex-automata",
+]
+
 [[package]]
 name = "bumpalo"
 version = "3.7.0"
@@ -1645,6 +1656,7 @@ version = "0.1.0"
 dependencies = [
  "borsh",
  "bridge",
+ "bstr",
  "byteorder",
  "hex",
  "hex-literal",
@@ -2199,6 +2211,12 @@ dependencies = [
  "regex-syntax",
 ]
 
+[[package]]
+name = "regex-automata"
+version = "0.1.10"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6c230d73fb8d8c1b9c0b3135c5142a8acee3a0558fb8db5cf1cb65f8d7862132"
+
 [[package]]
 name = "regex-syntax"
 version = "0.6.25"

solana/modules/nft_bridge/program/Cargo.toml

@@ -19,6 +19,7 @@ default = []
 [dependencies]
 bridge = { path = "../../../bridge/program", features = ["no-entrypoint", "cpi"] }
 borsh = "0.8.1"
+bstr = "0.2.16"
 byteorder = "1.4.3"
 rocksalt = { path = "../../../solitaire/rocksalt" }
 solitaire = { path = "../../../solitaire/program" }

solana/modules/nft_bridge/program/src/api/complete_transfer.rs

@@ -215,6 +215,8 @@ pub fn complete_wrapped(
     accs: &mut CompleteWrapped,
     data: CompleteWrappedData,
 ) -> Result<()> {
+    use bstr::ByteSlice;
+
     // Verify the chain registration
     let derivation_data: EndpointDerivationData = (&*accs).into();
     accs.chain_registration
@@ -261,10 +263,12 @@ pub fn complete_wrapped(
             },
         )?;
 
-        let mut name = accs.vaa.name.clone();
-        name.truncate(32);
-        let mut symbol = accs.vaa.symbol.clone();
+        let name = accs.vaa.name.clone();
+        let mut symbol: Vec<u8> = accs.vaa.symbol.clone().as_bytes().to_vec();
         symbol.truncate(10);
+        let mut symbol: Vec<char> = symbol.chars().collect();
+        symbol.retain(|&c| c != '\u{FFFD}');
+        let symbol: String = symbol.iter().collect();
 
         let spl_token_metadata_ix = spl_token_metadata::instruction::create_metadata_accounts(
             spl_token_metadata::id(),

solana/modules/nft_bridge/program/src/messages.rs

@@ -67,6 +67,7 @@ pub struct PayloadTransfer {
 
 impl DeserializePayload for PayloadTransfer {
     fn deserialize(buf: &mut &[u8]) -> Result<Self, SolitaireError> {
+        use bstr::ByteSlice;
         let mut v = Cursor::new(buf);
 
         if v.read_u8()? != 1 {
@@ -78,17 +79,23 @@ impl DeserializePayload for PayloadTransfer {
 
         let token_chain = v.read_u16::<BigEndian>()?;
 
-        let mut symbol_data: [u8; 32] = [0; 32];
+        // We may receive invalid UTF-8 over the bridge, especially if truncated. To compensate for
+        // this we rely on the bstr libraries ability to parse invalid UTF-8, and strip out the
+        // "Invalid Unicode Codepoint" (FFFD) characters. This becomes the canonical representation
+        // on Solana.
+        let mut symbol_data = vec![0u8; 32];
         v.read_exact(&mut symbol_data)?;
-        let mut symbol = String::from_utf8(symbol_data.to_vec())
-            .map_err::<SolitaireError, _>(|_| TokenBridgeError::InvalidUTF8String.into())?;
-        symbol = symbol.chars().filter(|c| c != &'\0').collect();
+        symbol_data.retain(|&c| c != 0);
+        let mut symbol: Vec<char> = symbol_data.chars().collect();
+        symbol.retain(|&c| c != '\u{FFFD}');
+        let symbol: String = symbol.iter().collect();
 
-        let mut name_data: [u8; 32] = [0; 32];
+        let mut name_data = vec![0u8; 32];
         v.read_exact(&mut name_data)?;
-        let mut name = String::from_utf8(name_data.to_vec())
-            .map_err::<SolitaireError, _>(|_| TokenBridgeError::InvalidUTF8String.into())?;
-        name = name.chars().filter(|c| c != &'\0').collect();
+        name_data.retain(|&c| c != 0);
+        let mut name: Vec<char> = name_data.chars().collect();
+        name.retain(|&c| c != '\u{FFFD}');
+        let name: String = name.iter().collect();
 
         let mut id_data: [u8; 32] = [0; 32];
         v.read_exact(&mut id_data)?;