ci(rust): check MSRV in CI and fix clippy (lancedb#3054)

Author: wjones127

.cargo/config.toml

@@ -9,24 +9,6 @@ debug = true
 codegen-units = 16
 lto = "thin"
 
-[target.'cfg(all())']
-rustflags = [
-    "-Wclippy::all",
-    "-Wclippy::style",
-    "-Wclippy::fallible_impl_from",
-    "-Wclippy::manual_let_else",
-    "-Wclippy::redundant_pub_crate",
-    "-Wclippy::string_add_assign",
-    "-Wclippy::string_add",
-    "-Wclippy::string_lit_as_bytes",
-    "-Wclippy::string_to_string",
-    "-Wclippy::use_self",
-    "-Dclippy::cargo",
-    "-Dclippy::dbg_macro",
-    # not too much we can do to avoid multiple crate versions
-    "-Aclippy::multiple-crate-versions",
-]
-
 [target.x86_64-unknown-linux-gnu]
 rustflags = ["-C", "target-cpu=haswell", "-C", "target-feature=+avx2,+fma,+f16c"]
 

.github/workflows/rust.yml

@@ -38,8 +38,8 @@ jobs:
           - nightly
     env:
       # Need up-to-date compilers for kernels
-      CC: clang-18
-      CXX: clang-18
+      CC: clang
+      CXX: clang
     steps:
       - uses: actions/checkout@v4
       # pin the toolchain version to avoid surprises
@@ -81,7 +81,7 @@ jobs:
       - name: Run tests
         if: ${{ matrix.toolchain == 'stable' }}
         run: |
-          cargo llvm-cov --workspace --codecov --output-path coverage.codecov --features dynamodb,tensorflow,dynamodb_tests,cli
+          cargo llvm-cov --workspace --codecov --output-path coverage.codecov --all-features
       - name: Run tests (nightly)
         if: ${{ matrix.toolchain != 'stable' }}
         run: |
@@ -127,6 +127,10 @@ jobs:
   clippy_and_benchmark:
     runs-on: ubuntu-24.04
     timeout-minutes: 30
+    env:
+      # Need up-to-date compilers for kernels
+      CC: clang
+      CXX: clang
     steps:
       - uses: actions/checkout@v4
       - uses: Swatinem/rust-cache@v2
@@ -137,7 +141,9 @@ jobs:
           sudo apt update
           sudo apt install -y protobuf-compiler libssl-dev
       - name: Run clippy
-        run: cargo clippy --features cli,dynamodb,tensorflow,dynamodb_tests --tests --benches -- -D warnings
+        run: |
+          cargo clippy --version
+          cargo clippy --all-features --tests --benches -- -D warnings
       - name: Build benchmarks
         run: cargo build --benches
   mac-build:
@@ -196,3 +202,28 @@ jobs:
         run: |
           cargo build --tests --benches --all-features --workspace
           cargo test
+  msrv:
+    # Check the minimum supported Rust version
+    name: MSRV Check - Rust v${{ matrix.msrv }}
+    runs-on: ubuntu-24.04
+    strategy:
+      matrix:
+        msrv: ["1.78.0"] # This should match up with rust-version in Cargo.toml
+    env:
+      # Need up-to-date compilers for kernels
+      CC: clang
+      CXX: clang
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: true
+      - name: Install dependencies
+        run: |
+          sudo apt update
+          sudo apt install -y protobuf-compiler libssl-dev
+      - name: Install ${{ matrix.msrv }}
+        uses: dtolnay/rust-toolchain@master
+        with:
+          toolchain: ${{ matrix.msrv }}
+      - name: cargo +${{ matrix.msrv }} check
+        run: cargo check --workspace --tests --benches --all-features

Cargo.toml

@@ -163,3 +163,21 @@ zstd = "0.13.1"
 opt-level = 3
 debug = true
 strip = false
+
+[workspace.lints.clippy]
+all = { level = "deny", priority = -1 }
+style = { level = "deny", priority = -1 }
+cargo = { level = "deny", priority = -1 }
+fallible_impl_from = "deny"
+manual_let_else = "deny"
+redundant_pub_crate = "deny"
+string_add_assign = "deny"
+string_add = "deny"
+string_lit_as_bytes = "deny"
+string_to_string = "deny"
+use_self = "deny"
+dbg_macro = "deny"
+trait_duplication_in_bounds = "deny"
+redundant_clone = "deny"
+# not too much we can do to avoid multiple crate versions
+multiple-crate-versions = "allow"

rust/lance-arrow/Cargo.toml

@@ -26,3 +26,6 @@ rand.workspace = true
 
 [target.'cfg(target_arch = "wasm32")'.dependencies]
 getrandom = { version = "0.2", features = ["js"] }
+
+[lints]
+workspace = true

rust/lance-arrow/src/deepcopy.rs

@@ -56,5 +56,5 @@ pub fn deep_copy_batch(batch: &RecordBatch) -> crate::Result<RecordBatch> {
         .iter()
         .map(|array| deep_copy_array(array))
         .collect::<Vec<_>>();
-    RecordBatch::try_new(batch.schema().clone(), arrays)
+    RecordBatch::try_new(batch.schema(), arrays)
 }

rust/lance-arrow/src/lib.rs

@@ -475,7 +475,7 @@ impl RecordBatchExt for RecordBatch {
             .position(|f| f.name() == name)
             .ok_or_else(|| ArrowError::SchemaError(format!("Field {} does not exist", name)))?;
         columns[field_i] = column;
-        Self::try_new(self.schema().clone(), columns)
+        Self::try_new(self.schema(), columns)
     }
 
     fn column_by_qualified_name(&self, name: &str) -> Option<&ArrayRef> {
@@ -637,7 +637,7 @@ pub fn interleave_batches(
     let first_batch = batches.first().ok_or_else(|| {
         ArrowError::InvalidArgumentError("Cannot interleave zero RecordBatches".to_string())
     })?;
-    let schema = first_batch.schema().clone();
+    let schema = first_batch.schema();
     let num_columns = first_batch.num_columns();
     let mut columns = Vec::with_capacity(num_columns);
     let mut chunks = Vec::with_capacity(batches.len());
@@ -785,7 +785,7 @@ mod tests {
             .with_metadata(metadata.clone().into()),
         );
         let batch = RecordBatch::try_new(
-            schema.clone(),
+            schema,
             vec![
                 Arc::new(Int32Array::from_iter_values(0..20)),
                 Arc::new(StringArray::from_iter_values(
@@ -830,7 +830,7 @@ mod tests {
         // Sub schema
         let sub_schema = Schema::new(vec![Field::new("a", DataType::Int32, true)]);
         let sub_projected = batch.project_by_schema(&sub_schema).unwrap();
-        let expected_schema = Arc::new(sub_schema.with_metadata(metadata.clone().into()));
+        let expected_schema = Arc::new(sub_schema.with_metadata(metadata.into()));
         assert_eq!(
             sub_projected,
             RecordBatch::try_new(

rust/lance-core/Cargo.toml

@@ -54,3 +54,6 @@ proptest.workspace = true
 
 [features]
 datafusion = ["datafusion-common", "datafusion-sql"]
+
+[lints]
+workspace = true

rust/lance-core/src/utils/mask.rs

@@ -901,7 +901,7 @@ mod tests {
             right_rows in proptest::collection::vec(0..u64::MAX, 0..1000),
         ) {
             let mut left = RowIdTreeMap::default();
-            for fragment in left_full_fragments.clone() {
+            for fragment in left_full_fragments {
                 left.insert_fragment(fragment);
             }
             left.extend(left_rows.iter().copied());
@@ -925,7 +925,7 @@ mod tests {
             left_rows in proptest::collection::vec(0..u64::MAX, 0..1000),
         ) {
             let mut left = RowIdTreeMap::default();
-            for fragment in left_full_fragments.clone() {
+            for fragment in left_full_fragments {
                 left.insert_fragment(fragment);
             }
             left.extend(left_rows.iter().copied());

rust/lance-datafusion/Cargo.toml

@@ -37,3 +37,6 @@ lance-datagen.workspace = true
 
 [features]
 substrait = ["dep:datafusion-substrait"]
+
+[lints]
+workspace = true

rust/lance-datafusion/src/chunker.rs

@@ -180,7 +180,7 @@ pub fn chunk_concat_stream(
     stream: SendableRecordBatchStream,
     chunk_size: usize,
 ) -> SendableRecordBatchStream {
-    let schema = stream.schema().clone();
+    let schema = stream.schema();
     let schema_copy = schema.clone();
     let chunked = chunk_stream(stream, chunk_size);
     let chunk_concat = chunked

rust/lance-datafusion/src/exec.rs

@@ -50,7 +50,7 @@ pub struct OneShotExec {
 impl OneShotExec {
     /// Create a new instance from a given stream
     pub fn new(stream: SendableRecordBatchStream) -> Self {
-        let schema = stream.schema().clone();
+        let schema = stream.schema();
         Self {
             stream: Mutex::new(Some(stream)),
             schema: schema.clone(),
@@ -272,7 +272,7 @@ struct OneShotPartitionStream {
 
 impl OneShotPartitionStream {
     fn new(data: SendableRecordBatchStream) -> Self {
-        let schema = data.schema().clone();
+        let schema = data.schema();
         Self {
             data: Arc::new(Mutex::new(Some(data))),
             schema,
@@ -298,7 +298,7 @@ impl SessionContextExt for SessionContext {
         &self,
         data: SendableRecordBatchStream,
     ) -> datafusion::common::Result<DataFrame> {
-        let schema = data.schema().clone();
+        let schema = data.schema();
         let part_stream = Arc::new(OneShotPartitionStream::new(data));
         let provider = StreamingTable::try_new(schema, vec![part_stream])?;
         self.read_table(Arc::new(provider))

rust/lance-datafusion/src/planner.rs

@@ -608,7 +608,7 @@ impl Planner {
                             ))?;
                         }
                     }
-                    Field::new("item", data_type.clone(), true)
+                    Field::new("item", data_type, true)
                 } else {
                     Field::new("item", ArrowDataType::Null, true)
                 };
@@ -825,7 +825,7 @@ impl Planner {
         let simplifier =
             datafusion::optimizer::simplify_expressions::ExprSimplifier::new(simplify_context);
 
-        let expr = simplifier.simplify(expr.clone())?;
+        let expr = simplifier.simplify(expr)?;
         let expr = simplifier.coerce(expr, &df_schema)?;
 
         Ok(expr)
@@ -1009,7 +1009,7 @@ mod tests {
             ),
         ]));
 
-        let planner = Planner::new(schema.clone());
+        let planner = Planner::new(schema);
 
         fn assert_column_eq(planner: &Planner, expr: &str, expected: &Expr) {
             let expr = planner.parse_filter(&format!("{expr} = 'val'")).unwrap();
@@ -1082,7 +1082,7 @@ mod tests {
             true,
         )]));
 
-        let planner = Planner::new(schema.clone());
+        let planner = Planner::new(schema);
 
         let expected = array_element(col("l"), lit(0_i64));
         let expr = planner.parse_expr("l[0]").unwrap();
@@ -1132,7 +1132,7 @@ mod tests {
     fn test_negative_array_expressions() {
         let schema = Arc::new(Schema::new(vec![Field::new("x", DataType::Int64, false)]));
 
-        let planner = Planner::new(schema.clone());
+        let planner = Planner::new(schema);
 
         let expected = Expr::Literal(ScalarValue::List(Arc::new(
             ListArray::from_iter_primitive::<Float64Type, _, _>(vec![Some(

rust/lance-datagen/Cargo.toml

@@ -32,3 +32,6 @@ bench = false
 [[bench]]
 name = "array_gen"
 harness = false
+
+[lints]
+workspace = true

rust/lance-datagen/src/generator.rs

@@ -805,7 +805,7 @@ impl<T: ByteArrayType> CycleBinaryGenerator<T> {
         Self {
             values,
             lengths,
-            data_type: T::DATA_TYPE.clone(),
+            data_type: T::DATA_TYPE,
             array_type: PhantomData,
             width,
             idx: 0,
@@ -863,7 +863,7 @@ impl<T: ByteArrayType> FixedBinaryGenerator<T> {
     pub fn new(value: Vec<u8>) -> Self {
         Self {
             value,
-            data_type: T::DATA_TYPE.clone(),
+            data_type: T::DATA_TYPE,
             array_type: PhantomData,
         }
     }
@@ -910,7 +910,7 @@ pub struct DictionaryGenerator<K: ArrowDictionaryKeyType> {
 
 impl<K: ArrowDictionaryKeyType> DictionaryGenerator<K> {
     fn new(generator: Box<dyn ArrayGenerator>) -> Self {
-        let key_type = Box::new(K::DATA_TYPE.clone());
+        let key_type = Box::new(K::DATA_TYPE);
         let key_width = key_type
             .primitive_width()
             .expect("dictionary key types should have a known width")
@@ -1352,7 +1352,7 @@ pub mod array {
         let mut values_idx = 0;
         Box::new(
             FnGen::<DataType::Native, PrimitiveArray<DataType>, _>::new_known_size(
-                DataType::DATA_TYPE.clone(),
+                DataType::DATA_TYPE,
                 move |_| {
                     let y = values[values_idx];
                     values_idx = (values_idx + 1) % values.len();
@@ -1377,7 +1377,7 @@ pub mod array {
         let mut x = DataType::Native::default();
         Box::new(
             FnGen::<DataType::Native, PrimitiveArray<DataType>, _>::new_known_size(
-                DataType::DATA_TYPE.clone(),
+                DataType::DATA_TYPE,
                 move |_| {
                     let y = x;
                     x += DataType::Native::ONE;
@@ -1411,7 +1411,7 @@ pub mod array {
         let mut x = start;
         Box::new(
             FnGen::<DataType::Native, PrimitiveArray<DataType>, _>::new_known_size(
-                DataType::DATA_TYPE.clone(),
+                DataType::DATA_TYPE,
                 move |_| {
                     let y = x;
                     x += step;
@@ -1435,7 +1435,7 @@ pub mod array {
     {
         Box::new(
             FnGen::<DataType::Native, PrimitiveArray<DataType>, _>::new_known_size(
-                DataType::DATA_TYPE.clone(),
+                DataType::DATA_TYPE,
                 move |_| value,
                 1,
                 DataType::DATA_TYPE
@@ -1470,7 +1470,7 @@ pub mod array {
     {
         Box::new(
             FnGen::<DataType::Native, PrimitiveArray<DataType>, _>::new_known_size(
-                DataType::DATA_TYPE.clone(),
+                DataType::DATA_TYPE,
                 move |rng| rng.gen(),
                 1,
                 DataType::DATA_TYPE
@@ -1495,7 +1495,7 @@ pub mod array {
     {
         Box::new(
             FnGen::<DataType::Native, PrimitiveArray<DataType>, _>::new_known_size(
-                DataType::DATA_TYPE.clone(),
+                DataType::DATA_TYPE,
                 move |rng| rng.sample(dist.clone()),
                 1,
                 DataType::DATA_TYPE
@@ -1632,7 +1632,7 @@ pub mod array {
         let dist = Uniform::new(start_days, end_days);
 
         Box::new(FnGen::<i32, Date32Array, _>::new_known_size(
-            data_type.clone(),
+            data_type,
             move |rng| dist.sample(rng),
             1,
             DataType::Date32
@@ -1727,7 +1727,7 @@ pub mod array {
         let dist = Uniform::new(start_days, end_days);
 
         Box::new(FnGen::<i64, Date64Array, _>::new_known_size(
-            data_type.clone(),
+            data_type,
             move |rng| (dist.sample(rng)) * MS_PER_DAY,
             1,
             DataType::Date64

rust/lance-encoding-datafusion/Cargo.toml

@@ -45,3 +45,6 @@ prost-build.workspace = true
 
 [target.'cfg(target_os = "linux")'.dev-dependencies]
 pprof = { workspace = true }
+
+[lints]
+workspace = true