Cosmos: Implement Binary Round-Trip Fuzzer

Cargo.toml

@@ -88,6 +88,8 @@ version = "1.0.0"
 version = "1.0.0"
 
 [workspace.dependencies]
+arbitrary = { version = "1.4", features = ["derive"] }
+arbitrary-json = "0.1"
 async-lock = "3.4"
 async-stream = { version = "0.3.6" }
 async-trait = "0.1"
@@ -117,6 +119,7 @@ hdrhistogram = "7.5"
 hostname = "0.4"
 hmac = { version = "0.12" }
 include-file = { version = "1.0.0", default-features = false }
+json-canon = "0.1"
 openssl = { version = "0.10.79" }
 opentelemetry = { version = "0.32", features = ["trace"] }
 opentelemetry-appender-tracing = { version = "0.32" }

eng/dict/crates.txt

@@ -1,3 +1,6 @@
+arbitrary
+arbitrary_json
+arbitrary-json
 arc_swap
 arc-swap
 async_lock
@@ -70,8 +73,12 @@ hostname
 http
 include_file
 include-file
+json_canon
+json-canon
 json_patch
 json-patch
+libfuzzer_sys
+libfuzzer-sys
 litemap
 openssl
 opentelemetry

sdk/cosmos/.cspell.json

@@ -33,6 +33,7 @@
     "cabi",
     "canadacentral",
     "canadaeast",
+    "CBOR",
     "cbindgen",
     "CDLL",
     "cdriver",
@@ -52,6 +53,7 @@
     "chinaeast",
     "chinanorth",
     "chokepoint",
+    "canonicalizer",
     "cloneable",
     "codepoint",
     "codepoints",
@@ -172,6 +174,7 @@
     "libazurecosmosdriver",
     "libclang",
     "LIBCLANG",
+    "libfuzzer",
     "libqueryplaninterop",
     "QUERYPLANINTEROP",
     "linearizability",
@@ -256,10 +259,14 @@
     "RAII",
     "readfeed",
     "recompiles",
+    "redecoded",
+    "reencode",
+    "reencoded",
     "refetch",
     "refetched",
     "refetches",
     "refcounted",
+    "reparses",
     "reparsed",
     "Replicaset",
     "reqs",
@@ -284,6 +291,7 @@
     "serviceunavailable",
     "serviceversion",
     "sess",
+    "Signedness",
     "southafricanorth",
     "southafricawest",
     "southcentralus",
@@ -313,11 +321,13 @@
     "testdb",
     "thinclient",
     "threadsafe",
+    "tmin",
     "TOCTOU",
     "Tokio",
     "TOPCOUNT",
     "TPIO",
     "transcoders",
+    "trippable",
     "uaecentral",
     "uaenorth",
     "udfs",
@@ -374,6 +384,35 @@
     "writability",
     "xorshift",
     "xpart",
+    "artifactname",
+    "fsanitize",
+    "logissue",
+    "MSVC",
+    "AFORM",
+    "ANID",
+    "Chenault",
+    "devtestcoll",
+    "ENOENT",
+    "errno",
+    "feff",
+    "Flacco",
+    "logdata",
+    "lossily",
+    "Millett",
+    "modhash",
+    "MSNH",
+    "MUID",
+    "muid",
+    "nametype",
+    "Qsml",
+    "recclass",
+    "reclat",
+    "reclong",
+    "similars",
+    "Unschematized",
+    "vout",
+    "Xpert",
+    "xpert",
     "yxxx"
   ],
   "ignorePaths": [

sdk/cosmos/azure_data_cosmos/Cargo.toml

@@ -52,6 +52,13 @@ azure_data_cosmos_driver = { path = "../azure_data_cosmos_driver", default-featu
 ] }
 azure_identity.workspace = true
 clap.workspace = true
+# Test-only deps for the binary round-trip fuzzer
+# (tests/binary_roundtrip_fuzzer.rs): it generates arbitrary JSON, canonicalizes
+# it (RFC 8785), and hashes the canonical form for differential comparison.
+arbitrary = { workspace = true }
+arbitrary-json = { workspace = true }
+json-canon = { workspace = true }
+sha2 = { workspace = true }
 # Enabled here (dev-only) so the metrics and tracing handler tests can drive
 # in-memory OpenTelemetry exporters. `testing` provides `InMemoryMetricExporter`
 # and `InMemorySpanExporter`.
@@ -144,6 +151,11 @@ name = "binary_encoding"
 path = "tests/binary_encoding.rs"
 required-features = ["key_auth", "control_plane", "fault_injection"]
 
+[[test]]
+name = "binary_roundtrip_fuzzer"
+path = "tests/binary_roundtrip_fuzzer.rs"
+required-features = ["key_auth", "control_plane", "fault_injection"]
+
 [[test]]
 name = "gateway_v2"
 path = "tests/gateway_v2.rs"

sdk/cosmos/azure_data_cosmos/tests/binary_encoding_tests/cosmos_binary_encoding.rs

@@ -56,9 +56,13 @@ fn binary_encoding_options() -> TestOptions {
 }
 
 /// A document covering every JSON value shape the binary encoder emits: literal
-/// and wide integers, an unsigned value beyond `i64::MAX`, a double, booleans,
-/// `null`, unicode/empty strings, nested arrays and objects, and a vector of
-/// objects.
+/// and wide integers, a large unsigned value, a double, booleans, `null`,
+/// unicode/empty strings, nested arrays and objects, and a vector of objects.
+///
+/// Note: `huge` stays at or below `2^53` because the live Cosmos service
+/// normalizes JSON numbers to IEEE-754 doubles. A value beyond `2^53` (e.g.
+/// `u64::MAX`) is echoed back as a `Double` and can no longer be deserialized
+/// into a `u64` field, so it does not round-trip against the real service.
 #[derive(Debug, Clone, Deserialize, Serialize, PartialEq)]
 struct BinaryItem {
     id: String,
@@ -95,11 +99,11 @@ fn sample_item(id: &str, partition_key: &str) -> BinaryItem {
         text: "hello binary".to_owned(),
         unicode: "café ☃ 𝄞 quotes:\" backslash:\\".to_owned(),
         empty: String::new(),
-        small_int: 7,           // literal-int form (0..32)
-        big_int: 9_000_000_000, // Int64 form
-        negative: -1_234_567,   // Int64 form
-        huge: u64::MAX,         // UInt64 form (beyond i64::MAX)
-        ratio: 123.456_789,     // Double form
+        small_int: 7,                // literal-int form (0..32)
+        big_int: 9_000_000_000,      // Int64 form
+        negative: -1_234_567,        // Int64 form
+        huge: 9_007_199_254_740_992, // UInt64 form, exactly f64-representable (2^53)
+        ratio: 123.456_789,          // Double form
         active: true,
         inactive: false,
         maybe: None, // null