style: fix nightly clippy (#17178)

Signed-off-by: xxchan <[email protected]>

src/batch/src/executor/join/hash_join.rs

@@ -54,7 +54,7 @@ use crate::task::{BatchTaskContext, ShutdownToken};
 /// 2. Iterate over the probe side (i.e. left table) and compute the hash value of each row.
 ///    Then find the matched build side row for each probe side row in the hash map.
 /// 3. Concatenate the matched pair of probe side row and build side row into a single row and push
-/// it into the data chunk builder.
+///    it into the data chunk builder.
 /// 4. Yield chunks from the builder.
 pub struct HashJoinExecutor<K> {
     /// Join type e.g. inner, left outer, ...
@@ -1669,7 +1669,7 @@ impl<K: HashKey> HashJoinExecutor<K> {
     /// | 4 | 3 | 3 | - |
     ///
     /// 3. Remove duplicate rows with NULL build side. This is done by setting the visibility bitmap
-    /// of the chunk.
+    ///    of the chunk.
     ///
     /// | offset | v1 | v2 | v3 |
     /// |---|---|---|---|

src/batch/src/executor/test_utils.rs

@@ -75,7 +75,7 @@ pub fn gen_sorted_data(
         let mut array_builder = DataType::Int64.create_array_builder(batch_size);
 
         for _ in 0..batch_size {
-            array_builder.append(&data_gen.generate_datum(0));
+            array_builder.append(data_gen.generate_datum(0));
         }
 
         let array = array_builder.finish();
@@ -102,7 +102,7 @@ pub fn gen_projected_data(
         let mut array_builder = DataType::Int64.create_array_builder(batch_size);
 
         for j in 0..batch_size {
-            array_builder.append(&data_gen.generate_datum(((i + 1) * (j + 1)) as u64));
+            array_builder.append(data_gen.generate_datum(((i + 1) * (j + 1)) as u64));
         }
 
         let chunk = DataChunk::new(vec![array_builder.finish().into()], batch_size);

src/common/metrics/src/relabeled_metric.rs

@@ -24,8 +24,9 @@ use crate::{
 /// - when `metric_level` <= `relabel_threshold`, they behaves exactly the same as their inner
 ///   metric.
 /// - when `metric_level` > `relabel_threshold`, all their input label values are rewrite to "" when
-/// calling `with_label_values`. That's means the metric vec is aggregated into a single metric.
-
+///   calling `with_label_values`. That's means the metric vec is aggregated into a single metric.
+///
+///
 /// These wrapper classes add a `metric_level` field to corresponding metric.
 /// We could have use one single struct to represent all `MetricVec<T: MetricVecBuilder>`, rather
 /// than specializing them one by one. However, that's undoable because prometheus crate doesn't

src/common/src/array/data_chunk.rs

@@ -804,7 +804,7 @@ impl DataChunkTestExt for DataChunk {
                 let arr = col;
                 let mut builder = arr.create_builder(n * 2);
                 for v in arr.iter() {
-                    builder.append(&v.to_owned_datum());
+                    builder.append(v.to_owned_datum());
                     builder.append_null();
                 }
 

src/common/src/cache.rs

@@ -1033,6 +1033,7 @@ mod tests {
 
     pub struct Block {
         pub offset: u64,
+        #[allow(dead_code)]
         pub sst: u64,
     }
 

src/common/src/session_config/search_path.rs

@@ -23,12 +23,12 @@ pub const USER_NAME_WILD_CARD: &str = "\"$user\"";
 /// see <https://www.postgresql.org/docs/14/runtime-config-client.html#GUC-SEARCH-PATH>
 ///
 /// 1. when we `select` or `drop` object and don't give a specified schema, it will search the
-/// object from the valid items in schema `rw_catalog`, `pg_catalog` and `search_path`. If schema
-/// `rw_catalog` and `pg_catalog` are not in `search_path`, we will search them firstly. If they're
-/// in `search_path`, we will follow the order in `search_path`.
+///    object from the valid items in schema `rw_catalog`, `pg_catalog` and `search_path`. If schema
+///    `rw_catalog` and `pg_catalog` are not in `search_path`, we will search them firstly. If they're
+///    in `search_path`, we will follow the order in `search_path`.
 ///
 /// 2. when we `create` a `source` or `mv` and don't give a specified schema, it will use the first
-/// valid schema in `search_path`.
+///    valid schema in `search_path`.
 ///
 /// 3. when we `create` a `index` or `sink`, it will use the schema of the associated table.
 #[derive(Clone, Debug, PartialEq)]

src/common/src/util/chunk_coalesce.rs

@@ -459,13 +459,13 @@ mod tests {
 
         let mut left_array_builder = DataType::Int32.create_array_builder(5);
         for v in [1, 2, 3, 4, 5] {
-            left_array_builder.append(&Some(ScalarImpl::Int32(v)));
+            left_array_builder.append(Some(ScalarImpl::Int32(v)));
         }
         let left_arrays = [left_array_builder.finish()];
 
         let mut right_array_builder = DataType::Int64.create_array_builder(5);
         for v in [5, 4, 3, 2, 1] {
-            right_array_builder.append(&Some(ScalarImpl::Int64(v)));
+            right_array_builder.append(Some(ScalarImpl::Int64(v)));
         }
         let right_arrays = [right_array_builder.finish()];
 

src/common/src/util/memcmp_encoding.rs

@@ -543,7 +543,7 @@ mod tests {
         use rand::seq::SliceRandom;
 
         fn serialize(f: F32) -> MemcmpEncoded {
-            encode_value(&Some(ScalarImpl::from(f)), OrderType::default()).unwrap()
+            encode_value(Some(ScalarImpl::from(f)), OrderType::default()).unwrap()
         }
 
         fn deserialize(data: MemcmpEncoded) -> F32 {

src/connector/src/schema/schema_registry/util.rs

@@ -150,6 +150,7 @@ pub struct Subject {
 #[derive(Debug, Deserialize)]
 pub struct SchemaReference {
     /// The name of the reference.
+    #[allow(dead_code)]
     pub name: String,
     /// The subject that the referenced schema belongs to
     pub subject: String,

src/connector/src/sink/dynamodb.rs

@@ -13,7 +13,6 @@
 // limitations under the License.
 
 use std::collections::{BTreeMap, HashMap, HashSet};
-use std::usize;
 
 use anyhow::{anyhow, Context};
 use aws_sdk_dynamodb as dynamodb;

src/connector/src/sink/encoder/mod.rs

@@ -58,7 +58,7 @@ pub trait RowEncoder {
 /// * an json object
 /// * a protobuf message
 /// * an avro record
-/// into
+///   into
 /// * string (required by kinesis key)
 /// * bytes
 ///

src/connector/src/sink/encoder/proto.rs

@@ -205,6 +205,7 @@ impl MaybeData for () {
 /// * Top level is always a message.
 /// * All message fields can be omitted in proto3.
 /// * All repeated elements must have a value.
+///
 /// So we handle [`ScalarRefImpl`] rather than [`DatumRef`] here.
 impl MaybeData for ScalarRefImpl<'_> {
     type Out = Value;

src/connector/src/sink/formatter/mod.rs

@@ -45,9 +45,9 @@ pub trait SinkFormatter {
     type V;
 
     /// * Key may be None so that messages are partitioned using round-robin.
-    /// For example append-only without `primary_key` (aka `downstream_pk`) set.
+    ///   For example append-only without `primary_key` (aka `downstream_pk`) set.
     /// * Value may be None so that messages with same key are removed during log compaction.
-    /// For example debezium tombstone event.
+    ///   For example debezium tombstone event.
     fn format_chunk(
         &self,
         chunk: &StreamChunk,

src/connector/src/sink/google_pubsub.rs

@@ -13,7 +13,6 @@
 // limitations under the License.
 
 use std::collections::BTreeMap;
-use std::usize;
 
 use anyhow::{anyhow, Context};
 use google_cloud_gax::conn::Environment;

src/connector/src/source/pulsar/topic.rs

@@ -108,6 +108,7 @@ pub fn get_partition_index(topic: &str) -> Result<Option<i32>> {
 /// The short topic name can be:
 /// - `<topic>`
 /// - `<tenant>/<namespace>/<topic>`
+///
 /// The fully qualified topic name can be:
 /// `<domain>://<tenant>/<namespace>/<topic>`
 pub fn parse_topic(topic: &str) -> Result<Topic> {

src/frontend/src/binder/bind_context.rs

@@ -107,6 +107,7 @@ pub enum BindingCteState {
 pub struct RecursiveUnion {
     /// currently this *must* be true,
     /// otherwise binding will fail.
+    #[allow(dead_code)]
     pub all: bool,
     /// lhs part of the `UNION ALL` operator
     pub base: Box<BoundSetExpr>,

src/frontend/src/binder/mod.rs

@@ -261,17 +261,17 @@ impl UdfContext {
 /// following the rules:
 /// 1. At the beginning, it contains the user specified parameters type.
 /// 2. When the binder encounters a parameter, it will record it as unknown(call `record_new_param`)
-/// if it didn't exist in `ParameterTypes`.
+///    if it didn't exist in `ParameterTypes`.
 /// 3. When the binder encounters a cast on parameter, if it's a unknown type, the cast function
-/// will record the target type as infer type for that parameter(call `record_infer_type`). If the
-/// parameter has been inferred, the cast function will act as a normal cast.
+///    will record the target type as infer type for that parameter(call `record_infer_type`). If the
+///    parameter has been inferred, the cast function will act as a normal cast.
 /// 4. After bind finished:
 ///     (a) parameter not in `ParameterTypes` means that the user didn't specify it and it didn't
-/// occur in the query. `export` will return error if there is a kind of
-/// parameter. This rule is compatible with PostgreSQL
+///    occur in the query. `export` will return error if there is a kind of
+///    parameter. This rule is compatible with PostgreSQL
 ///     (b) parameter is None means that it's a unknown type. The user didn't specify it
-/// and we can't infer it in the query. We will treat it as VARCHAR type finally. This rule is
-/// compatible with PostgreSQL.
+///    and we can't infer it in the query. We will treat it as VARCHAR type finally. This rule is
+///    compatible with PostgreSQL.
 ///     (c) parameter is Some means that it's a known type.
 #[derive(Clone, Debug)]
 pub struct ParameterTypes(Arc<RwLock<HashMap<u64, Option<DataType>>>>);

src/frontend/src/catalog/table_catalog.rs

@@ -60,7 +60,7 @@ use crate::user::UserId;
 /// - **Order Key**: the primary key for storage, used to sort and access data.
 ///
 ///   For an MV, the columns in `ORDER BY` clause will be put at the beginning of the order key. And
-/// the remaining columns in pk will follow behind.
+///   the remaining columns in pk will follow behind.
 ///
 ///   If there's no `ORDER BY` clause, the order key will be the same as pk.
 ///

src/frontend/src/optimizer/plan_node/logical_multi_join.rs

@@ -406,7 +406,7 @@ impl LogicalMultiJoin {
     ///        a. a filter with the non eq conditions
     ///        b. a projection which reorders the output column ordering to agree with the
     ///           original ordering of the joins.
-    ///   The filter will then be pushed down by another filter pushdown pass.
+    ///    The filter will then be pushed down by another filter pushdown pass.
     pub(crate) fn heuristic_ordering(&self) -> Result<Vec<usize>> {
         let mut labeller = ConnectedComponentLabeller::new(self.inputs.len());
 
@@ -494,9 +494,9 @@ impl LogicalMultiJoin {
     /// 2. Second, for every isolated node will create connection to every other nodes.
     /// 3. Third, select and merge one node for a iteration, and use a bfs policy for which node the
     ///    selected node merged with.
-    ///   i. The select node mentioned above is the node with least number of relations and the
+    ///    i. The select node mentioned above is the node with least number of relations and the
     ///      lowerst join tree.
-    ///   ii. nodes with a join tree higher than the temporal optimal join tree will be pruned.
+    ///    ii. nodes with a join tree higher than the temporal optimal join tree will be pruned.
     pub fn as_bushy_tree_join(&self) -> Result<PlanRef> {
         let (nodes, condition) = self.get_join_graph()?;
 

src/frontend/src/optimizer/plan_node/predicate_pushdown.rs

@@ -29,14 +29,14 @@ pub trait PredicatePushdown {
     /// There are three kinds of predicates:
     ///
     /// 1. those can't be pushed down. We just create a `LogicalFilter` for them above the current
-    /// `PlanNode`. i.e.,
+    ///    `PlanNode`. i.e.,
     ///
     ///     ```ignore
     ///     LogicalFilter::create(self.clone().into(), predicate)
     ///     ```
     ///
     /// 2. those can be merged with current `PlanNode` (e.g., `LogicalJoin`). We just merge
-    /// the predicates with the `Condition` of it.
+    ///    the predicates with the `Condition` of it.
     ///
     /// 3. those can be pushed down. We pass them to current `PlanNode`'s input.
     fn predicate_pushdown(

src/frontend/src/optimizer/plan_node/stream_table_scan.rs

@@ -147,6 +147,7 @@ impl StreamTableScan {
     ///        | 1002 | Int64(1) | t                   | 10          |
     ///        | 1003 | Int64(1) | t                   | 10          |
     ///        | 1003 | Int64(1) | t                   | 10          |
+    ///
     /// Eventually we should track progress per vnode, to support scaling with both mview and
     /// the corresponding `no_shuffle_backfill`.
     /// However this is not high priority, since we are working on supporting arrangement backfill,

src/frontend/src/optimizer/rule/logical_filter_expression_simplify_rule.rs

@@ -33,6 +33,7 @@ impl Rule for LogicalFilterExpressionSimplifyRule {
     /// The pattern we aim to optimize, e.g.,
     /// 1. (NOT (e)) OR (e) => True
     /// 2. (NOT (e)) AND (e) => False
+    ///
     /// NOTE: `e` should only contain at most a single column
     /// otherwise we will not conduct the optimization
     fn apply(&self, plan: PlanRef) -> Option<PlanRef> {
@@ -153,6 +154,7 @@ fn check_optimizable_pattern(e1: ExprImpl, e2: ExprImpl) -> (bool, Option<ExprIm
 
 /// 1. True or (...) | (...) or True => True
 /// 2. False and (...) | (...) and False => False
+///
 /// NOTE: the `True` and `False` here not only represent a single `ExprImpl::Literal`
 /// but represent every `ExprImpl` that can be *evaluated* to `ScalarImpl::Bool`
 /// during optimization phase as well

src/frontend/src/scheduler/distributed/stage.rs

@@ -94,7 +94,7 @@ pub enum StageEvent {
         reason: SchedulerError,
     },
     /// All tasks in stage finished.
-    Completed(StageId),
+    Completed(#[allow(dead_code)] StageId),
 }
 
 #[derive(Clone)]

src/meta/src/controller/rename.rs

@@ -170,7 +170,7 @@ impl QueryRewriter<'_> {
     ///
     /// So that we DON'T have to:
     /// 1. rewrite the select and expr part like `schema.table.column`, `table.column`,
-    /// `alias.column` etc.
+    ///    `alias.column` etc.
     /// 2. handle the case that the old name is used as alias.
     /// 3. handle the case that the new name is used as alias.
     fn visit_table_factor(&self, table_factor: &mut TableFactor) {

src/meta/src/hummock/compaction/picker/mod.rs

@@ -97,18 +97,3 @@ pub trait CompactionPicker {
         stats: &mut LocalPickerStatistic,
     ) -> Option<CompactionInput>;
 }
-
-#[derive(Default, Clone, Debug)]
-pub struct PartitionLevelInfo {
-    pub level_id: u32,
-    pub sub_level_id: u64,
-    pub left_idx: usize,
-    pub right_idx: usize,
-    pub total_file_size: u64,
-}
-
-#[derive(Default, Clone, Debug)]
-pub struct LevelPartition {
-    pub sub_levels: Vec<PartitionLevelInfo>,
-    pub total_file_size: u64,
-}

src/meta/src/hummock/compactor_manager.rs

@@ -111,12 +111,12 @@ impl Compactor {
 /// `CompactTaskAssignment`.
 ///
 /// A compact task can be in one of these states:
-/// - 1. Success: an assigned task is reported as success via `CompactStatus::report_compact_task`.
-///   It's the final state.
-/// - 2. Failed: an Failed task is reported as success via `CompactStatus::report_compact_task`.
-///   It's the final state.
-/// - 3. Cancelled: a task is reported as cancelled via `CompactStatus::report_compact_task`. It's
-///   the final state.
+/// 1. Success: an assigned task is reported as success via `CompactStatus::report_compact_task`.
+///    It's the final state.
+/// 2. Failed: an Failed task is reported as success via `CompactStatus::report_compact_task`.
+///    It's the final state.
+/// 3. Cancelled: a task is reported as cancelled via `CompactStatus::report_compact_task`. It's
+///    the final state.
 pub struct CompactorManagerInner {
     pub task_expired_seconds: u64,
     pub heartbeat_expired_seconds: u64,

src/meta/src/hummock/manager/gc.rs

@@ -173,7 +173,7 @@ impl HummockManager {
     /// Starts a full GC.
     /// 1. Meta node sends a `FullScanTask` to a compactor in this method.
     /// 2. The compactor returns scan result of object store to meta node. See
-    /// `HummockManager::full_scan_inner` in storage crate.
+    ///    `HummockManager::full_scan_inner` in storage crate.
     /// 3. Meta node decides which SSTs to delete. See `HummockManager::complete_full_gc`.
     ///
     /// Returns Ok(false) if there is no worker available.

src/meta/src/manager/catalog/mod.rs

@@ -187,6 +187,7 @@ impl CatalogManager {
     /// We identify a 'legacy' source based on two conditions:
     /// 1. The `format_encode_options` in `source_info` is empty.
     /// 2. Keys with certain prefixes belonging to `format_encode_options` exist in `with_properties` instead.
+    ///
     /// And if the source is identified as 'legacy', we copy the misplaced keys from `with_properties` to `format_encode_options`.
     async fn source_backward_compat_check(&self) -> MetaResult<()> {
         let core = &mut *self.core.lock().await;
@@ -897,6 +898,7 @@ impl CatalogManager {
     ///    with:
     ///    1. `stream_job_status` = CREATING
     ///    2. Not belonging to a background stream job.
+    ///
     ///    Clean up these hanging tables by the id.
     pub async fn clean_dirty_tables(&self, fragment_manager: FragmentManagerRef) -> MetaResult<()> {
         let core = &mut *self.core.lock().await;

src/meta/src/manager/env.rs

@@ -192,7 +192,7 @@ pub struct MetaOpts {
     /// The Dashboard service uses this in the following ways:
     /// 1. Query Prometheus for relevant metrics to find Stream Graph Bottleneck, and display it.
     /// 2. Provide cluster diagnostics, at `/api/monitor/diagnose` to troubleshoot cluster.
-    /// These are just examples which show how the Meta Dashboard Service queries Prometheus.
+    ///    These are just examples which show how the Meta Dashboard Service queries Prometheus.
     pub prometheus_endpoint: Option<String>,
 
     /// The additional selector used when querying Prometheus.
@@ -251,7 +251,7 @@ pub struct MetaOpts {
     /// When `hybrid_partition_vnode_count` > 0, in hybrid compaction group
     /// - Tables with high write throughput will be split at vnode granularity
     /// - Tables with high size tables will be split by table granularity
-    /// When `hybrid_partition_vnode_count` = 0,no longer be special alignment operations for the hybird compaction group
+    ///   When `hybrid_partition_vnode_count` = 0,no longer be special alignment operations for the hybird compaction group
     pub hybrid_partition_node_count: u32,
 
     pub event_log_enabled: bool,

src/meta/src/storage/transaction.rs

@@ -17,9 +17,9 @@ use crate::storage::{ColumnFamily, Key, Value};
 /// A `Transaction` executes several writes(aka. operations) to meta store atomically with optional
 /// preconditions checked. It executes as follow:
 /// 1. If all `preconditions` are valid, all `operations` are executed; Otherwise no operation
-/// is executed.
+///    is executed.
 /// 2. Upon `commit` the transaction, the `TransactionAbort` error will be returned if
-/// any precondition was not met in previous step.
+///    any precondition was not met in previous step.
 #[derive(Default)]
 pub struct Transaction {
     preconditions: Vec<Precondition>,