update rust toolchain

rust-toolchain.toml

@@ -1,3 +1,3 @@
 [toolchain]
-channel = "nightly-2024-06-06"
+channel = "nightly-2024-09-27"
 

src/auth/src/user_provider/static_user_provider.rs

@@ -33,7 +33,7 @@ impl StaticUserProvider {
             value: value.to_string(),
             msg: "StaticUserProviderOption must be in format `<option>:<value>`",
         })?;
-        return match mode {
+        match mode {
             "file" => {
                 let users = load_credential_from_file(content)?
                     .context(InvalidConfigSnafu {
@@ -58,7 +58,7 @@ impl StaticUserProvider {
                 msg: "StaticUserProviderOption must be in format `file:<path>` or `cmd:<values>`",
             }
                 .fail(),
-        };
+        }
     }
 }
 

src/catalog/src/system_schema/information_schema/table_names.rs

@@ -12,7 +12,7 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
-/// All table names in `information_schema`.
+//! All table names in `information_schema`.
 
 pub const TABLES: &str = "tables";
 pub const COLUMNS: &str = "columns";

src/catalog/src/system_schema/pg_catalog/pg_namespace.rs

@@ -12,6 +12,9 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+//! The `pg_catalog.pg_namespace` table implementation.
+//! namespace is a schema in greptime
+
 pub(super) mod oid_map;
 
 use std::sync::{Arc, Weak};
@@ -40,9 +43,6 @@ use crate::system_schema::utils::tables::{string_column, u32_column};
 use crate::system_schema::SystemTable;
 use crate::CatalogManager;
 
-/// The `pg_catalog.pg_namespace` table implementation.
-/// namespace is a schema in greptime
-
 const NSPNAME: &str = "nspname";
 const INIT_CAPACITY: usize = 42;
 

src/cmd/src/lib.rs

@@ -74,6 +74,7 @@ pub trait App: Send {
 }
 
 /// Log the versions of the application, and the arguments passed to the cli.
+///
 /// `version` should be the same as the output of cli "--version";
 /// and the `short_version` is the short version of the codes, often consist of git branch and commit.
 pub fn log_versions(version: &str, short_version: &str, app: &str) {

src/common/base/src/secrets.rs

@@ -46,8 +46,9 @@ impl From<String> for SecretString {
     }
 }
 
-/// Wrapper type for values that contains secrets, which attempts to limit
-/// accidental exposure and ensure secrets are wiped from memory when dropped.
+/// Wrapper type for values that contains secrets.
+///
+/// It attempts to limit accidental exposure and ensure secrets are wiped from memory when dropped.
 /// (e.g. passwords, cryptographic keys, access tokens or other credentials)
 ///
 /// Access to the secret inner value occurs through the [`ExposeSecret`]

src/common/catalog/src/consts.rs

@@ -103,14 +103,15 @@ pub const INFORMATION_SCHEMA_PROCEDURE_INFO_TABLE_ID: u32 = 34;
 /// id for information_schema.region_statistics
 pub const INFORMATION_SCHEMA_REGION_STATISTICS_TABLE_ID: u32 = 35;
 
-/// ----- End of information_schema tables -----
+// ----- End of information_schema tables -----
 
 /// ----- Begin of pg_catalog tables -----
 pub const PG_CATALOG_PG_CLASS_TABLE_ID: u32 = 256;
 pub const PG_CATALOG_PG_TYPE_TABLE_ID: u32 = 257;
 pub const PG_CATALOG_PG_NAMESPACE_TABLE_ID: u32 = 258;
 
-/// ----- End of pg_catalog tables -----
+// ----- End of pg_catalog tables -----
+
 pub const MITO_ENGINE: &str = "mito";
 pub const MITO2_ENGINE: &str = "mito2";
 pub const METRIC_ENGINE: &str = "metric";

src/common/greptimedb-telemetry/src/lib.rs

@@ -199,6 +199,7 @@ pub fn default_get_uuid(working_home: &Option<String>) -> Option<String> {
 }
 
 /// Report version info to GreptimeDB.
+///
 /// We do not collect any identity-sensitive information.
 /// This task is scheduled to run every 30 minutes.
 /// The task will be disabled default. It can be enabled by setting the build feature `greptimedb-telemetry`

src/common/macro/src/lib.rs

@@ -35,7 +35,9 @@ pub fn aggr_func_type_store_derive(input: TokenStream) -> TokenStream {
 }
 
 /// A struct can be used as a creator for aggregate function if it has been annotated with this
-/// attribute first. This attribute add a necessary field which is intended to store the input
+/// attribute first.
+///
+/// This attribute add a necessary field which is intended to store the input
 /// data's types to the struct.
 /// This attribute is expected to be used along with derive macro [AggrFuncTypeStore].
 #[proc_macro_attribute]
@@ -44,9 +46,10 @@ pub fn as_aggr_func_creator(args: TokenStream, input: TokenStream) -> TokenStrea
 }
 
 /// Attribute macro to convert an arithimetic function to a range function. The annotated function
-/// should accept servaral arrays as input and return a single value as output. This procedure
-/// macro can works on any number of input parameters. Return type can be either primitive type
-/// or wrapped in `Option`.
+/// should accept servaral arrays as input and return a single value as output.
+///
+/// This procedure macro can works on any number of input parameters. Return type can be either
+/// primitive type or wrapped in `Option`.
 ///
 /// # Example
 /// Take `count_over_time()` in PromQL as an example:

src/common/meta/src/cluster.rs

@@ -55,6 +55,7 @@ pub trait ClusterInfo {
 }
 
 /// The key of [NodeInfo] in the storage. The format is `__meta_cluster_node_info-{cluster_id}-{role}-{node_id}`.
+///
 /// This key cannot be used to describe the `Metasrv` because the `Metasrv` does not have
 /// a `cluster_id`, it serves multiple clusters.
 #[derive(Debug, Clone, Eq, Hash, PartialEq, Serialize, Deserialize)]

src/common/meta/src/key/flow/flow_name.rs

@@ -104,7 +104,7 @@ impl<'a> MetadataKey<'a, FlowNameKeyInner<'a>> for FlowNameKeyInner<'_> {
         .into_bytes()
     }
 
-    fn from_bytes(bytes: &'a [u8]) -> Result<FlowNameKeyInner> {
+    fn from_bytes(bytes: &'a [u8]) -> Result<FlowNameKeyInner<'a>> {
         let key = std::str::from_utf8(bytes).map_err(|e| {
             error::InvalidMetadataSnafu {
                 err_msg: format!(

src/common/recordbatch/src/recordbatch.rs

@@ -220,7 +220,7 @@ pub struct RecordBatchRowIterator<'a> {
 }
 
 impl<'a> RecordBatchRowIterator<'a> {
-    fn new(record_batch: &'a RecordBatch) -> RecordBatchRowIterator {
+    fn new(record_batch: &'a RecordBatch) -> RecordBatchRowIterator<'a> {
         RecordBatchRowIterator {
             record_batch,
             rows: record_batch.df_record_batch.num_rows(),

src/common/time/src/interval.rs

@@ -62,6 +62,7 @@ impl From<ArrowIntervalUnit> for IntervalUnit {
 }
 
 /// Interval Type represents a period of time.
+///
 /// It is composed of months, days and nanoseconds.
 /// 3 kinds of interval are supported: year-month, day-time and
 /// month-day-nano, which will be stored in the following format.

src/datatypes/src/types/cast.rs

@@ -82,8 +82,8 @@ pub fn cast_with_opt(
     }
 }
 
-/// Return true if the src_value can be casted to dest_type,
-/// Otherwise, return false.
+/// Return true if the src_value can be casted to dest_type, Otherwise, return false.
+///
 /// Notice: this function does not promise that the `cast_with_opt` will succeed,
 /// it only checks whether the src_value can be casted to dest_type.
 pub fn can_cast_type(src_value: &Value, dest_type: &ConcreteDataType) -> bool {

src/datatypes/src/types/primitive_type.rs

@@ -83,9 +83,10 @@ pub trait LogicalPrimitiveType: 'static + Sized {
     fn cast_value_ref(value: ValueRef) -> Result<Option<Self::Wrapper>>;
 }
 
-/// A new type for [WrapperType], complement the `Ord` feature for it. Wrapping non ordered
-/// primitive types like `f32` and `f64` in `OrdPrimitive` can make them be used in places that
-/// require `Ord`. For example, in `Median` UDAFs.
+/// A new type for [WrapperType], complement the `Ord` feature for it.
+///
+/// Wrapping non ordered primitive types like `f32` and `f64` in `OrdPrimitive`
+/// can make them be used in places that require `Ord`. For example, in `Median` UDAFs.
 #[derive(Debug, Clone, Copy, PartialEq)]
 pub struct OrdPrimitive<T: WrapperType>(pub T);
 

src/datatypes/src/vectors/list.rs

@@ -157,7 +157,7 @@ pub struct ListIter<'a> {
 }
 
 impl<'a> ListIter<'a> {
-    fn new(vector: &'a ListVector) -> ListIter {
+    fn new(vector: &'a ListVector) -> ListIter<'a> {
         ListIter { vector, idx: 0 }
     }
 }

src/index/src/inverted_index/search/fst_values_mapper.rs

@@ -19,7 +19,9 @@ use crate::inverted_index::error::Result;
 use crate::inverted_index::format::reader::InvertedIndexReader;
 
 /// `FstValuesMapper` maps FST-encoded u64 values to their corresponding bitmaps
-/// within an inverted index. The higher 32 bits of each u64 value represent the
+/// within an inverted index.
+///
+/// The higher 32 bits of each u64 value represent the
 /// bitmap offset and the lower 32 bits represent its size. This mapper uses these
 /// combined offset-size pairs to fetch and union multiple bitmaps into a single `BitVec`.
 pub struct FstValuesMapper<'a> {

src/index/src/inverted_index/search/index_apply/predicates_apply.rs

@@ -134,7 +134,7 @@ impl PredicatesIndexApplier {
     fn bitmap_full_range(metadata: &InvertedIndexMetas) -> BitVec {
         let total_count = metadata.total_row_count;
         let segment_count = metadata.segment_row_count;
-        let len = (total_count + segment_count - 1) / segment_count;
+        let len = total_count.div_ceil(segment_count);
         BitVec::repeat(true, len as _)
     }
 }

src/meta-srv/src/lib.rs

@@ -16,7 +16,6 @@
 #![feature(result_flattening)]
 #![feature(assert_matches)]
 #![feature(extract_if)]
-#![feature(option_take_if)]
 
 pub mod bootstrap;
 mod cache_invalidator;

src/mito2/src/config.rs

@@ -499,7 +499,7 @@ fn divide_num_cpus(divisor: usize) -> usize {
     let cores = common_config::utils::get_cpus();
     debug_assert!(cores > 0);
 
-    (cores + divisor - 1) / divisor
+    cores.div_ceil(divisor)
 }
 
 #[cfg(test)]

src/mito2/src/read/seq_scan.rs

@@ -724,7 +724,7 @@ fn maybe_split_parts(mut parts: Vec<ScanPart>, parallelism: usize) -> Vec<ScanPa
         // Safety: `can_split_preserve_order()` ensures file_ranges.len() == 1.
         // Splits part into `num_parts_to_split + 1` new parts if possible.
         let target_part_num = num_parts_to_split + 1;
-        let ranges_per_part = (part.file_ranges[0].len() + target_part_num - 1) / target_part_num;
+        let ranges_per_part = part.file_ranges[0].len().div_ceil(target_part_num);
         // `can_split_preserve_order()` ensures part.file_ranges[0].len() > 1.
         assert!(ranges_per_part > 0);
         for ranges in part.file_ranges[0].chunks(ranges_per_part) {

src/mito2/src/sst/parquet/metadata.rs

@@ -35,7 +35,11 @@ pub(crate) struct MetadataLoader<'a> {
 
 impl<'a> MetadataLoader<'a> {
     /// Create a new parquet metadata loader.
-    pub fn new(object_store: ObjectStore, file_path: &'a str, file_size: u64) -> MetadataLoader {
+    pub fn new(
+        object_store: ObjectStore,
+        file_path: &'a str,
+        file_size: u64,
+    ) -> MetadataLoader<'a> {
         Self {
             object_store,
             file_path,

src/pipeline/src/etl/processor.rs

@@ -62,7 +62,8 @@ const TARGET_FIELDS_NAME: &str = "target_fields";
 // const ON_FAILURE_NAME: &str = "on_failure";
 // const TAG_NAME: &str = "tag";
 
-/// Processor trait defines the interface for all processors
+/// Processor trait defines the interface for all processors.
+///
 /// A processor is a transformation that can be applied to a field in a document
 /// It can be used to extract, transform, or enrich data
 /// Now Processor only have one input field. In the future, we may support multiple input fields.

src/pipeline/src/manager.rs

@@ -26,6 +26,7 @@ pub mod table;
 pub mod util;
 
 /// Pipeline version. An optional timestamp with nanosecond precision.
+///
 /// If the version is None, it means the latest version of the pipeline.
 /// User can specify the version by providing a timestamp string formatted as iso8601.
 /// When it used in cache key, it will be converted to i64 meaning the number of nanoseconds since the epoch.

src/promql/src/extension_plan/histogram_fold.rs

@@ -44,7 +44,9 @@ use datatypes::vectors::MutableVector;
 use futures::{ready, Stream, StreamExt};
 
 /// `HistogramFold` will fold the conventional (non-native) histogram ([1]) for later
-/// computing. Specifically, it will transform the `le` and `field` column into a complex
+/// computing.
+///
+/// Specifically, it will transform the `le` and `field` column into a complex
 /// type, and samples on other tag columns:
 /// - `le` will become a [ListArray] of [f64]. With each bucket bound parsed
 /// - `field` will become a [ListArray] of [f64]

src/promql/src/extension_plan/scalar_calculate.rs

@@ -46,7 +46,8 @@ use crate::error::{ColumnNotFoundSnafu, DataFusionPlanningSnafu, DeserializeSnaf
 /// `ScalarCalculate` is the custom logical plan to calculate
 /// [`scalar`](https://prometheus.io/docs/prometheus/latest/querying/functions/#scalar)
 /// in PromQL, return NaN when have multiple time series.
-/// return the time series as scalar value when only have one time series.
+///
+/// Return the time series as scalar value when only have one time series.
 #[derive(Debug, Clone, PartialEq, Eq, Hash)]
 pub struct ScalarCalculate {
     start: Millisecond,

src/script/src/python/rspython/vector_impl.rs

@@ -12,6 +12,8 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+//! PyVectors' rustpython specify methods
+
 use arrow::compute::kernels::numeric;
 use common_time::date::Date;
 use common_time::datetime::DateTime;
@@ -40,7 +42,6 @@ use crate::python::ffi_types::vector::{
     arrow_rfloordiv, arrow_rsub, arrow_rtruediv, rspy_is_pyobj_scalar, wrap_result, PyVector,
 };
 use crate::python::rspython::utils::{is_instance, obj_cast_to};
-/// PyVectors' rustpython specify methods
 
 fn to_type_error(vm: &'_ VirtualMachine) -> impl FnOnce(String) -> PyBaseExceptionRef + '_ {
     |msg: String| vm.new_type_error(msg)

src/servers/src/export_metrics.rs

@@ -31,7 +31,9 @@ use crate::error::{InvalidExportMetricsConfigSnafu, Result, SendPromRemoteReques
 use crate::prom_store::{snappy_compress, to_grpc_row_insert_requests};
 use crate::query_handler::PromStoreProtocolHandlerRef;
 
-/// Use to export the metrics generated by greptimedb, encoded to Prometheus [RemoteWrite format](https://prometheus.io/docs/concepts/remote_write_spec/),
+/// Use to export the metrics generated by greptimedb.
+///
+/// Encoded to Prometheus [RemoteWrite format](https://prometheus.io/docs/concepts/remote_write_spec/),
 /// and send to Prometheus remote-write compatible receiver (e.g. send to `greptimedb` itself)
 #[derive(Clone, Debug, Serialize, Deserialize, PartialEq, Eq)]
 #[serde(default)]

src/servers/src/query_handler.rs

@@ -13,6 +13,7 @@
 // limitations under the License.
 
 //! All query handler traits for various request protocols, like SQL or GRPC.
+//!
 //! Instance that wishes to support certain request protocol, just implement the corresponding
 //! trait, the Server will handle codec for you.
 //!
@@ -122,6 +123,7 @@ pub trait OpenTelemetryProtocolHandler {
 }
 
 /// LogHandler is responsible for handling log related requests.
+///
 /// It should be able to insert logs and manage pipelines.
 /// The pipeline is a series of transformations that can be applied to logs.
 /// The pipeline is stored in the database and can be retrieved by name.

src/sql/src/statements/tql.rs

@@ -131,6 +131,7 @@ impl Display for TqlAnalyze {
 }
 
 /// Intermediate structure used to unify parameter mappings for various TQL operations.
+///
 /// This struct serves as a common parameter container for parsing TQL queries
 /// and constructing corresponding TQL operations: `TqlEval`, `TqlAnalyze` or `TqlExplain`.
 #[derive(Debug)]

src/store-api/src/manifest/action.rs

@@ -31,8 +31,9 @@ pub fn supported_protocol_version() -> (ProtocolVersion, ProtocolVersion) {
 }
 
 /// Protocol action that used to block older clients from reading or writing the log when backwards
-/// incompatible changes are made to the protocol. clients should be tolerant of messages and
-/// fields that they do not understand.
+/// incompatible changes are made to the protocol.
+///
+/// clients should be tolerant of messages and fields that they do not understand.
 #[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq)]
 pub struct ProtocolAction {
     pub min_reader_version: ProtocolVersion,

src/store-api/src/path_utils.rs

@@ -13,9 +13,9 @@
 // limitations under the License.
 
 //! Path constants for table engines, cluster states and WAL
+//! All paths relative to data_home(file storage) or root path(S3, OSS etc).
 
 use crate::storage::{RegionId, RegionNumber, TableId};
-/// All paths relative to data_home(file storage) or root path(S3, OSS etc).
 
 /// WAL dir for local file storage
 pub const WAL_DIR: &str = "wal/";
@@ -45,7 +45,9 @@ pub fn region_dir(path: &str, region_id: RegionId) -> String {
     )
 }
 
-/// get_storage_path returns the storage path from the region_dir. It will always return the storage path if the region_dir is valid, otherwise None.
+/// get_storage_path returns the storage path from the region_dir.
+///
+/// It will always return the storage path if the region_dir is valid, otherwise None.
 /// The storage path is constructed from the catalog and schema, which are generated by `common_meta::ddl::utils::region_storage_path`.
 /// We can extract the catalog and schema from the region_dir by following example:
 /// ```

src/store-api/src/region_engine.rs

@@ -247,6 +247,7 @@ impl ScannerProperties {
 }
 
 /// A scanner that provides a way to scan the region concurrently.
+///
 /// The scanner splits the region into partitions so that each partition can be scanned concurrently.
 /// You can use this trait to implement an [`ExecutionPlan`](datafusion_physical_plan::ExecutionPlan).
 pub trait RegionScanner: Debug + DisplayAs + Send {

src/table/src/metadata.rs

@@ -100,7 +100,8 @@ pub struct TableIdent {
     pub version: TableVersion,
 }
 
-/// The table metadata
+/// The table metadata.
+///
 /// Note: if you add new fields to this struct, please ensure 'new_meta_builder' function works.
 /// TODO(dennis): find a better way to ensure 'new_meta_builder' works when adding new fields.
 #[derive(Clone, Debug, Builder, PartialEq, Eq)]

src/table/src/stats.rs

@@ -14,7 +14,8 @@
 
 use datatypes::value::Value;
 
-/// Statistics for a relation
+/// Statistics for a relation.
+///
 /// Fields are optional and can be inexact because the sources
 /// sometimes provide approximate estimates for performance reasons
 /// and the transformations output are not always predictable.