query/

sql.rs

// Copyright 2023 Greptime Team
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

mod show_create_table;

use std::collections::HashMap;
use std::ops::ControlFlow;
use std::sync::Arc;

use catalog::CatalogManagerRef;
use catalog::information_schema::{
    CHARACTER_SETS, COLLATIONS, COLUMNS, FLOW_STATISTICS, FLOWS, REGION_PEERS, SCHEMATA,
    STATISTICS, TABLES, VIEWS, columns, flow_statistics, flows, process_list, region_peers,
    schemata, statistics, tables,
};
use common_catalog::consts::{
    INFORMATION_SCHEMA_NAME, SEMANTIC_TYPE_FIELD, SEMANTIC_TYPE_PRIMARY_KEY,
    SEMANTIC_TYPE_TIME_INDEX,
};
use common_catalog::format_full_table_name;
use common_datasource::file_format::{FileFormat, Format, infer_schemas};
use common_datasource::lister::{Lister, Source};
use common_datasource::object_store::{LocalFileAccess, build_backend_with_path};
use common_error::ext::BoxedError;
use common_meta::SchemaOptions;
use common_meta::ddl::create_flow::{FlowType, effective_eval_schedule_from_flow_info};
use common_meta::key::flow::flow_info::FlowInfoValue;
use common_query::Output;
use common_query::prelude::greptime_timestamp;
use common_recordbatch::RecordBatches;
use common_recordbatch::adapter::RecordBatchStreamAdapter;
use common_time::Timestamp;
use common_time::timezone::get_timezone;
use datafusion::dataframe::DataFrame;
use datafusion::prelude::SessionContext;
use datafusion_expr::{Expr, SortExpr, col, lit};
use datatypes::prelude::*;
use datatypes::schema::{ColumnDefaultConstraint, ColumnSchema, Schema};
use datatypes::vectors::StringVector;
use itertools::Itertools;
use object_store::ObjectStore;
use once_cell::sync::Lazy;
use regex::Regex;
use session::context::{Channel, QueryContextRef};
pub use show_create_table::create_table_stmt;
use snafu::{OptionExt, ResultExt, ensure};
use sql::ast::{Ident, visit_expressions_mut};
use sql::parser::ParserContext;
use sql::statements::OptionMap;
use sql::statements::create::{CreateDatabase, CreateFlow, CreateView, Partitions, SqlOrTql};
use sql::statements::show::{
    ShowColumns, ShowDatabases, ShowFlowStatus, ShowFlows, ShowIndex, ShowKind, ShowProcessList,
    ShowRegion, ShowTableStatus, ShowTables, ShowVariables, ShowViews,
};
use sql::statements::statement::Statement;
use sqlparser::ast::ObjectName;
use store_api::metric_engine_consts::{is_metric_engine, is_metric_engine_internal_column};
use table::TableRef;
use table::metadata::TableInfoRef;
use table::requests::{FILE_TABLE_LOCATION_KEY, FILE_TABLE_PATTERN_KEY};

use crate::QueryEngineRef;
use crate::error::{self, Result, UnsupportedVariableSnafu};
use crate::planner::DfLogicalPlanner;

const SCHEMAS_COLUMN: &str = "Database";
const OPTIONS_COLUMN: &str = "Options";
const VIEWS_COLUMN: &str = "Views";
const FLOWS_COLUMN: &str = "Flows";
const FIELD_COLUMN: &str = "Field";
const TABLE_TYPE_COLUMN: &str = "Table_type";

const COLUMN_NAME_COLUMN: &str = "Column";
const COLUMN_GREPTIME_TYPE_COLUMN: &str = "Greptime_type";
const COLUMN_TYPE_COLUMN: &str = "Type";
const COLUMN_KEY_COLUMN: &str = "Key";
const COLUMN_EXTRA_COLUMN: &str = "Extra";
const COLUMN_PRIVILEGES_COLUMN: &str = "Privileges";
const COLUMN_COLLATION_COLUMN: &str = "Collation";
const COLUMN_NULLABLE_COLUMN: &str = "Null";
const COLUMN_DEFAULT_COLUMN: &str = "Default";
const COLUMN_COMMENT_COLUMN: &str = "Comment";
const COLUMN_SEMANTIC_TYPE_COLUMN: &str = "Semantic Type";

const YES_STR: &str = "YES";
const NO_STR: &str = "NO";
const PRI_KEY: &str = "PRI";

/// SHOW index columns
const INDEX_TABLE_COLUMN: &str = "Table";
const INDEX_NONT_UNIQUE_COLUMN: &str = "Non_unique";
const INDEX_CARDINALITY_COLUMN: &str = "Cardinality";
const INDEX_SUB_PART_COLUMN: &str = "Sub_part";
const INDEX_PACKED_COLUMN: &str = "Packed";
const INDEX_INDEX_TYPE_COLUMN: &str = "Index_type";
const INDEX_COMMENT_COLUMN: &str = "Index_comment";
const INDEX_VISIBLE_COLUMN: &str = "Visible";
const INDEX_EXPRESSION_COLUMN: &str = "Expression";
const INDEX_KEY_NAME_COLUMN: &str = "Key_name";
const INDEX_SEQ_IN_INDEX_COLUMN: &str = "Seq_in_index";
const INDEX_COLUMN_NAME_COLUMN: &str = "Column_name";

pub static DESCRIBE_TABLE_OUTPUT_SCHEMA: Lazy<Arc<Schema>> = Lazy::new(|| {
    Arc::new(Schema::new(vec![
        ColumnSchema::new(
            COLUMN_NAME_COLUMN,
            ConcreteDataType::string_datatype(),
            false,
        ),
        ColumnSchema::new(
            COLUMN_TYPE_COLUMN,
            ConcreteDataType::string_datatype(),
            false,
        ),
        ColumnSchema::new(COLUMN_KEY_COLUMN, ConcreteDataType::string_datatype(), true),
        ColumnSchema::new(
            COLUMN_NULLABLE_COLUMN,
            ConcreteDataType::string_datatype(),
            false,
        ),
        ColumnSchema::new(
            COLUMN_DEFAULT_COLUMN,
            ConcreteDataType::string_datatype(),
            false,
        ),
        ColumnSchema::new(
            COLUMN_SEMANTIC_TYPE_COLUMN,
            ConcreteDataType::string_datatype(),
            false,
        ),
    ]))
});

static SHOW_CREATE_DATABASE_OUTPUT_SCHEMA: Lazy<Arc<Schema>> = Lazy::new(|| {
    Arc::new(Schema::new(vec![
        ColumnSchema::new("Database", ConcreteDataType::string_datatype(), false),
        ColumnSchema::new(
            "Create Database",
            ConcreteDataType::string_datatype(),
            false,
        ),
    ]))
});

static SHOW_CREATE_TABLE_OUTPUT_SCHEMA: Lazy<Arc<Schema>> = Lazy::new(|| {
    Arc::new(Schema::new(vec![
        ColumnSchema::new("Table", ConcreteDataType::string_datatype(), false),
        ColumnSchema::new("Create Table", ConcreteDataType::string_datatype(), false),
    ]))
});

static SHOW_CREATE_FLOW_OUTPUT_SCHEMA: Lazy<Arc<Schema>> = Lazy::new(|| {
    Arc::new(Schema::new(vec![
        ColumnSchema::new("Flow", ConcreteDataType::string_datatype(), false),
        ColumnSchema::new("Create Flow", ConcreteDataType::string_datatype(), false),
    ]))
});

static SHOW_CREATE_VIEW_OUTPUT_SCHEMA: Lazy<Arc<Schema>> = Lazy::new(|| {
    Arc::new(Schema::new(vec![
        ColumnSchema::new("View", ConcreteDataType::string_datatype(), false),
        ColumnSchema::new("Create View", ConcreteDataType::string_datatype(), false),
    ]))
});

pub async fn show_databases(
    stmt: ShowDatabases,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe =
        show_databases_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for `SHOW DATABASES` without executing it.
pub async fn show_databases_dataframe(
    stmt: &ShowDatabases,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let projects = if stmt.full {
        vec![
            (schemata::SCHEMA_NAME, SCHEMAS_COLUMN),
            (schemata::SCHEMA_OPTS, OPTIONS_COLUMN),
        ]
    } else {
        vec![(schemata::SCHEMA_NAME, SCHEMAS_COLUMN)]
    };

    let filters = vec![col(schemata::CATALOG_NAME).eq(lit(query_ctx.current_catalog()))];
    let like_field = Some(schemata::SCHEMA_NAME);
    let sort = vec![col(schemata::SCHEMA_NAME).sort(true, true)];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        SCHEMATA,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        &stmt.kind,
    )
    .await
}

/// Replaces column identifier references in a SQL expression.
/// Used for backward compatibility where old column names should work with new ones.
fn replace_column_in_expr(expr: &mut sqlparser::ast::Expr, from_column: &str, to_column: &str) {
    let _ = visit_expressions_mut(expr, |e| {
        match e {
            sqlparser::ast::Expr::Identifier(ident)
                if ident.value.eq_ignore_ascii_case(from_column) =>
            {
                ident.value = to_column.to_string();
            }
            sqlparser::ast::Expr::CompoundIdentifier(idents) => {
                if let Some(last) = idents.last_mut()
                    && last.value.eq_ignore_ascii_case(from_column)
                {
                    last.value = to_column.to_string();
                }
            }
            _ => {}
        }
        ControlFlow::<()>::Continue(())
    });
}

/// Cast a `show` statement execution into a query from tables in  `information_schema`.
/// - `table_name`: the table name in `information_schema`,
/// - `projects`: query projection, a list of `(column, renamed_column)`,
/// - `filters`: filter expressions for query,
/// - `like_field`: the field to filter by the predicate `ShowKind::Like`,
/// - `sort`: sort the results by the specified sorting expressions,
/// - `kind`: the show kind
#[allow(clippy::too_many_arguments)]
async fn query_from_information_schema_table(
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
    table_name: &str,
    select: Vec<Expr>,
    projects: Vec<(&str, &str)>,
    filters: Vec<Expr>,
    like_field: Option<&str>,
    sort: Vec<SortExpr>,
    kind: ShowKind,
) -> Result<Output> {
    let dataframe = query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        table_name,
        select,
        projects,
        filters,
        like_field,
        sort,
        &kind,
    )
    .await?;
    dataframe_to_output(dataframe).await
}

async fn dataframe_to_output(dataframe: DataFrame) -> Result<Output> {
    let stream = dataframe.execute_stream().await?;
    Ok(Output::new_with_stream(Box::pin(
        RecordBatchStreamAdapter::try_new(stream).context(error::CreateRecordBatchSnafu)?,
    )))
}

/// Builds the [`DataFrame`] for a `SHOW` statement without executing it,
/// so `Describe` handlers can derive the output schema from the same
/// projection the executor uses.
#[allow(clippy::too_many_arguments)]
async fn query_from_information_schema_dataframe(
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
    table_name: &str,
    select: Vec<Expr>,
    projects: Vec<(&str, &str)>,
    filters: Vec<Expr>,
    like_field: Option<&str>,
    sort: Vec<SortExpr>,
    kind: &ShowKind,
) -> Result<DataFrame> {
    let table = catalog_manager
        .table(
            query_ctx.current_catalog(),
            INFORMATION_SCHEMA_NAME,
            table_name,
            Some(&query_ctx),
        )
        .await
        .context(error::CatalogSnafu)?
        .with_context(|| error::TableNotFoundSnafu {
            table: format_full_table_name(
                query_ctx.current_catalog(),
                INFORMATION_SCHEMA_NAME,
                table_name,
            ),
        })?;

    let dataframe = query_engine.read_table(table)?;

    // Apply filters
    let dataframe = filters.into_iter().try_fold(dataframe, |df, expr| {
        df.filter(expr).context(error::PlanSqlSnafu)
    })?;

    // Apply `like` predicate if exists
    let dataframe = if let (ShowKind::Like(ident), Some(field)) = (&kind, like_field) {
        dataframe
            .filter(col(field).like(lit(ident.value.clone())))
            .context(error::PlanSqlSnafu)?
    } else {
        dataframe
    };

    // Apply sorting
    let dataframe = if sort.is_empty() {
        dataframe
    } else {
        dataframe.sort(sort).context(error::PlanSqlSnafu)?
    };

    // Apply select
    let dataframe = if select.is_empty() {
        if projects.is_empty() {
            dataframe
        } else {
            let projection = projects
                .iter()
                .map(|x| col(x.0).alias(x.1))
                .collect::<Vec<_>>();
            dataframe.select(projection).context(error::PlanSqlSnafu)?
        }
    } else {
        dataframe.select(select).context(error::PlanSqlSnafu)?
    };

    // Apply projection
    let dataframe = projects
        .into_iter()
        .try_fold(dataframe, |df, (column, renamed_column)| {
            df.with_column_renamed(column, renamed_column)
                .context(error::PlanSqlSnafu)
        })?;

    let dataframe = match kind {
        ShowKind::All | ShowKind::Like(_) => {
            // Like kind is processed above
            dataframe
        }
        ShowKind::Where(filter) => {
            // Cast the results into VIEW for `where` clause,
            // which is evaluated against the column names displayed by the SHOW statement.
            let view = dataframe.into_view();
            let dataframe = SessionContext::new_with_state(
                query_engine
                    .engine_context(query_ctx.clone())
                    .state()
                    .clone(),
            )
            .read_table(view)?;

            let planner = query_engine.planner();
            let planner = planner
                .as_any()
                .downcast_ref::<DfLogicalPlanner>()
                .expect("Must be the datafusion planner");

            let filter = planner
                .sql_to_expr(filter.clone(), dataframe.schema(), false, query_ctx)
                .await?;

            // Apply the `where` clause filters
            dataframe.filter(filter).context(error::PlanSqlSnafu)?
        }
    };

    Ok(dataframe)
}

/// Execute `SHOW COLUMNS` statement.
pub async fn show_columns(
    stmt: ShowColumns,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe = show_columns_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for `SHOW COLUMNS` without executing it.
pub async fn show_columns_dataframe(
    stmt: &ShowColumns,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let schema_name = if let Some(database) = &stmt.database {
        database.clone()
    } else {
        query_ctx.current_schema()
    };

    let projects = if stmt.full {
        vec![
            (columns::COLUMN_NAME, FIELD_COLUMN),
            (columns::DATA_TYPE, COLUMN_TYPE_COLUMN),
            (columns::COLLATION_NAME, COLUMN_COLLATION_COLUMN),
            (columns::IS_NULLABLE, COLUMN_NULLABLE_COLUMN),
            (columns::COLUMN_KEY, COLUMN_KEY_COLUMN),
            (columns::COLUMN_DEFAULT, COLUMN_DEFAULT_COLUMN),
            (columns::COLUMN_COMMENT, COLUMN_COMMENT_COLUMN),
            (columns::PRIVILEGES, COLUMN_PRIVILEGES_COLUMN),
            (columns::EXTRA, COLUMN_EXTRA_COLUMN),
            (columns::GREPTIME_DATA_TYPE, COLUMN_GREPTIME_TYPE_COLUMN),
        ]
    } else {
        vec![
            (columns::COLUMN_NAME, FIELD_COLUMN),
            (columns::DATA_TYPE, COLUMN_TYPE_COLUMN),
            (columns::IS_NULLABLE, COLUMN_NULLABLE_COLUMN),
            (columns::COLUMN_KEY, COLUMN_KEY_COLUMN),
            (columns::COLUMN_DEFAULT, COLUMN_DEFAULT_COLUMN),
            (columns::EXTRA, COLUMN_EXTRA_COLUMN),
            (columns::GREPTIME_DATA_TYPE, COLUMN_GREPTIME_TYPE_COLUMN),
        ]
    };

    let filters = vec![
        col(columns::TABLE_NAME).eq(lit(&stmt.table)),
        col(columns::TABLE_SCHEMA).eq(lit(schema_name.clone())),
        col(columns::TABLE_CATALOG).eq(lit(query_ctx.current_catalog())),
    ];
    let like_field = Some(columns::COLUMN_NAME);
    let sort = vec![col(columns::COLUMN_NAME).sort(true, true)];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        COLUMNS,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        &stmt.kind,
    )
    .await
}

/// Execute `SHOW INDEX` statement.
pub async fn show_index(
    stmt: ShowIndex,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe = show_index_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for `SHOW INDEX` without executing it.
pub async fn show_index_dataframe(
    stmt: &ShowIndex,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let schema_name = if let Some(database) = &stmt.database {
        database.clone()
    } else {
        query_ctx.current_schema()
    };

    let select = vec![
        col(statistics::TABLE_NAME).alias(INDEX_TABLE_COLUMN),
        col(statistics::NON_UNIQUE).alias(INDEX_NONT_UNIQUE_COLUMN),
        col(statistics::INDEX_NAME).alias(INDEX_KEY_NAME_COLUMN),
        col(statistics::SEQ_IN_INDEX).alias(INDEX_SEQ_IN_INDEX_COLUMN),
        col(statistics::COLUMN_NAME).alias(INDEX_COLUMN_NAME_COLUMN),
        col(statistics::COLLATION).alias(COLUMN_COLLATION_COLUMN),
        col(statistics::CARDINALITY).alias(INDEX_CARDINALITY_COLUMN),
        col(statistics::SUB_PART).alias(INDEX_SUB_PART_COLUMN),
        col(statistics::PACKED).alias(INDEX_PACKED_COLUMN),
        col(statistics::NULLABLE).alias(COLUMN_NULLABLE_COLUMN),
        col(statistics::INDEX_TYPE).alias(INDEX_INDEX_TYPE_COLUMN),
        col(statistics::COMMENT).alias(COLUMN_COMMENT_COLUMN),
        col(statistics::INDEX_COMMENT).alias(INDEX_COMMENT_COLUMN),
        col(statistics::IS_VISIBLE).alias(INDEX_VISIBLE_COLUMN),
        col(statistics::EXPRESSION).alias(INDEX_EXPRESSION_COLUMN),
    ];

    let projects = vec![
        (statistics::TABLE_NAME, INDEX_TABLE_COLUMN),
        (INDEX_NONT_UNIQUE_COLUMN, INDEX_NONT_UNIQUE_COLUMN),
        (statistics::INDEX_NAME, INDEX_KEY_NAME_COLUMN),
        (statistics::SEQ_IN_INDEX, INDEX_SEQ_IN_INDEX_COLUMN),
        (statistics::COLUMN_NAME, INDEX_COLUMN_NAME_COLUMN),
        (COLUMN_COLLATION_COLUMN, COLUMN_COLLATION_COLUMN),
        (INDEX_CARDINALITY_COLUMN, INDEX_CARDINALITY_COLUMN),
        (INDEX_SUB_PART_COLUMN, INDEX_SUB_PART_COLUMN),
        (INDEX_PACKED_COLUMN, INDEX_PACKED_COLUMN),
        (COLUMN_NULLABLE_COLUMN, COLUMN_NULLABLE_COLUMN),
        (statistics::INDEX_TYPE, INDEX_INDEX_TYPE_COLUMN),
        (COLUMN_COMMENT_COLUMN, COLUMN_COMMENT_COLUMN),
        (INDEX_COMMENT_COLUMN, INDEX_COMMENT_COLUMN),
        (INDEX_VISIBLE_COLUMN, INDEX_VISIBLE_COLUMN),
        (INDEX_EXPRESSION_COLUMN, INDEX_EXPRESSION_COLUMN),
    ];

    let filters = vec![
        col(statistics::TABLE_NAME).eq(lit(&stmt.table)),
        col(statistics::TABLE_SCHEMA).eq(lit(schema_name.clone())),
    ];
    let like_field = None;
    let sort = vec![
        col(statistics::INDEX_NAME).sort(true, true),
        col(statistics::SEQ_IN_INDEX).sort(true, true),
    ];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        STATISTICS,
        select,
        projects,
        filters,
        like_field,
        sort,
        &stmt.kind,
    )
    .await
}

/// Execute `SHOW REGION` statement.
pub async fn show_region(
    stmt: ShowRegion,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe = show_region_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for `SHOW REGION` without executing it.
pub async fn show_region_dataframe(
    stmt: &ShowRegion,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let schema_name = if let Some(database) = &stmt.database {
        database.clone()
    } else {
        query_ctx.current_schema()
    };

    let filters = vec![
        col(region_peers::TABLE_NAME).eq(lit(&stmt.table)),
        col(region_peers::TABLE_SCHEMA).eq(lit(schema_name.clone())),
        col(region_peers::TABLE_CATALOG).eq(lit(query_ctx.current_catalog())),
    ];
    let projects = vec![
        (region_peers::TABLE_NAME, "Table"),
        (region_peers::REGION_ID, "Region"),
        (region_peers::PEER_ID, "Peer"),
        (region_peers::IS_LEADER, "Leader"),
    ];

    let like_field = None;
    let sort = vec![
        col(columns::REGION_ID).sort(true, true),
        col(columns::PEER_ID).sort(true, true),
    ];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        REGION_PEERS,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        &stmt.kind,
    )
    .await
}

/// Execute [`ShowTables`] statement and return the [`Output`] if success.
pub async fn show_tables(
    stmt: ShowTables,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe = show_tables_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for [`ShowTables`] without executing it.
pub async fn show_tables_dataframe(
    stmt: &ShowTables,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let schema_name = if let Some(database) = &stmt.database {
        database.clone()
    } else {
        query_ctx.current_schema()
    };

    // MySQL renames `table_name` to `Tables_in_{schema}` for protocol compatibility
    let tables_column = format!("Tables_in_{}", schema_name);
    let projects = if stmt.full {
        vec![
            (tables::TABLE_NAME, tables_column.as_str()),
            (tables::TABLE_TYPE, TABLE_TYPE_COLUMN),
        ]
    } else {
        vec![(tables::TABLE_NAME, tables_column.as_str())]
    };
    let filters = vec![
        col(tables::TABLE_SCHEMA).eq(lit(schema_name.clone())),
        col(tables::TABLE_CATALOG).eq(lit(query_ctx.current_catalog())),
    ];
    let like_field = Some(tables::TABLE_NAME);
    let sort = vec![col(tables::TABLE_NAME).sort(true, true)];

    // Transform the WHERE clause for backward compatibility:
    // Replace "Tables" with "Tables_in_{schema}" to support old queries
    let rewritten_kind = match &stmt.kind {
        ShowKind::Where(filter) => {
            let mut filter = filter.clone();
            replace_column_in_expr(&mut filter, "Tables", &tables_column);
            ShowKind::Where(filter)
        }
        kind => kind.clone(),
    };

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        TABLES,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        &rewritten_kind,
    )
    .await
}

/// Execute [`ShowTableStatus`] statement and return the [`Output`] if success.
pub async fn show_table_status(
    stmt: ShowTableStatus,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe =
        show_table_status_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for [`ShowTableStatus`] without executing it.
pub async fn show_table_status_dataframe(
    stmt: &ShowTableStatus,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let schema_name = if let Some(database) = &stmt.database {
        database.clone()
    } else {
        query_ctx.current_schema()
    };

    // Refer to https://dev.mysql.com/doc/refman/8.4/en/show-table-status.html
    let projects = vec![
        (tables::TABLE_NAME, "Name"),
        (tables::ENGINE, "Engine"),
        (tables::VERSION, "Version"),
        (tables::ROW_FORMAT, "Row_format"),
        (tables::TABLE_ROWS, "Rows"),
        (tables::AVG_ROW_LENGTH, "Avg_row_length"),
        (tables::DATA_LENGTH, "Data_length"),
        (tables::MAX_DATA_LENGTH, "Max_data_length"),
        (tables::INDEX_LENGTH, "Index_length"),
        (tables::DATA_FREE, "Data_free"),
        (tables::AUTO_INCREMENT, "Auto_increment"),
        (tables::CREATE_TIME, "Create_time"),
        (tables::UPDATE_TIME, "Update_time"),
        (tables::CHECK_TIME, "Check_time"),
        (tables::TABLE_COLLATION, "Collation"),
        (tables::CHECKSUM, "Checksum"),
        (tables::CREATE_OPTIONS, "Create_options"),
        (tables::TABLE_COMMENT, "Comment"),
    ];

    let filters = vec![
        col(tables::TABLE_SCHEMA).eq(lit(schema_name.clone())),
        col(tables::TABLE_CATALOG).eq(lit(query_ctx.current_catalog())),
    ];
    let like_field = Some(tables::TABLE_NAME);
    let sort = vec![col(tables::TABLE_NAME).sort(true, true)];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        TABLES,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        &stmt.kind,
    )
    .await
}

/// Execute `SHOW COLLATION` statement and returns the `Output` if success.
pub async fn show_collations(
    kind: ShowKind,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe =
        show_collations_dataframe(&kind, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for `SHOW COLLATION` without executing it.
pub async fn show_collations_dataframe(
    kind: &ShowKind,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    // Refer to https://dev.mysql.com/doc/refman/8.0/en/show-collation.html
    let projects = vec![
        ("collation_name", "Collation"),
        ("character_set_name", "Charset"),
        ("id", "Id"),
        ("is_default", "Default"),
        ("is_compiled", "Compiled"),
        ("sortlen", "Sortlen"),
    ];

    let filters = vec![];
    let like_field = Some("collation_name");
    let sort = vec![];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        COLLATIONS,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        kind,
    )
    .await
}

/// Execute `SHOW CHARSET` statement and returns the `Output` if success.
pub async fn show_charsets(
    kind: ShowKind,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe =
        show_charsets_dataframe(&kind, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for `SHOW CHARSET` without executing it.
pub async fn show_charsets_dataframe(
    kind: &ShowKind,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    // Refer to https://dev.mysql.com/doc/refman/8.0/en/show-character-set.html
    let projects = vec![
        ("character_set_name", "Charset"),
        ("description", "Description"),
        ("default_collate_name", "Default collation"),
        ("maxlen", "Maxlen"),
    ];

    let filters = vec![];
    let like_field = Some("character_set_name");
    let sort = vec![];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        CHARACTER_SETS,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        kind,
    )
    .await
}

pub fn show_variable(stmt: ShowVariables, query_ctx: QueryContextRef) -> Result<Output> {
    let variable = stmt.variable.to_string().to_uppercase();
    let value = match variable.as_str() {
        "SYSTEM_TIME_ZONE" | "SYSTEM_TIMEZONE" => get_timezone(None).to_string(),
        "TIME_ZONE" | "TIMEZONE" => query_ctx.timezone().to_string(),
        "READ_PREFERENCE" => query_ctx.read_preference().to_string(),
        "DATESTYLE" => {
            let (style, order) = *query_ctx.configuration_parameter().pg_datetime_style();
            format!("{}, {}", style, order)
        }
        "INTERVALSTYLE" => {
            let style = *query_ctx
                .configuration_parameter()
                .pg_intervalstyle_format();
            style.to_string()
        }
        "MAX_EXECUTION_TIME"
            if query_ctx.channel() == Channel::Mysql => {
                query_ctx.query_timeout_as_millis().to_string()
            }
        "STATEMENT_TIMEOUT"
            // Add time units to postgres query timeout display.
            if query_ctx.channel() == Channel::Postgres => {
                let mut timeout = query_ctx.query_timeout_as_millis().to_string();
                timeout.push_str("ms");
                timeout
            }
        _ => return UnsupportedVariableSnafu { name: variable }.fail(),
    };
    let schema = Arc::new(Schema::new(vec![ColumnSchema::new(
        variable,
        ConcreteDataType::string_datatype(),
        false,
    )]));
    let records = RecordBatches::try_from_columns(
        schema,
        vec![Arc::new(StringVector::from(vec![value])) as _],
    )
    .context(error::CreateRecordBatchSnafu)?;
    Ok(Output::new_with_record_batches(records))
}

pub async fn show_status(_query_ctx: QueryContextRef) -> Result<Output> {
    let schema = Arc::new(Schema::new(vec![
        ColumnSchema::new("Variable_name", ConcreteDataType::string_datatype(), false),
        ColumnSchema::new("Value", ConcreteDataType::string_datatype(), true),
    ]));
    let records = RecordBatches::try_from_columns(
        schema,
        vec![
            Arc::new(StringVector::from(Vec::<&str>::new())) as _,
            Arc::new(StringVector::from(Vec::<&str>::new())) as _,
        ],
    )
    .context(error::CreateRecordBatchSnafu)?;
    Ok(Output::new_with_record_batches(records))
}

pub async fn show_search_path(_query_ctx: QueryContextRef) -> Result<Output> {
    let schema = Arc::new(Schema::new(vec![ColumnSchema::new(
        "search_path",
        ConcreteDataType::string_datatype(),
        false,
    )]));
    let records = RecordBatches::try_from_columns(
        schema,
        vec![Arc::new(StringVector::from(vec![_query_ctx.current_schema()])) as _],
    )
    .context(error::CreateRecordBatchSnafu)?;
    Ok(Output::new_with_record_batches(records))
}

pub fn show_create_database(database_name: &str, options: OptionMap) -> Result<Output> {
    let stmt = CreateDatabase {
        name: ObjectName::from(vec![Ident::new(database_name)]),
        if_not_exists: true,
        options,
    };
    let sql = format!("{stmt}");
    let columns = vec![
        Arc::new(StringVector::from(vec![database_name.to_string()])) as _,
        Arc::new(StringVector::from(vec![sql])) as _,
    ];
    let records =
        RecordBatches::try_from_columns(SHOW_CREATE_DATABASE_OUTPUT_SCHEMA.clone(), columns)
            .context(error::CreateRecordBatchSnafu)?;
    Ok(Output::new_with_record_batches(records))
}

pub fn show_create_table(
    table_info: TableInfoRef,
    schema_options: Option<SchemaOptions>,
    partitions: Option<Partitions>,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let table_name = table_info.name.clone();

    let quote_style = query_ctx.quote_style();

    let mut stmt = create_table_stmt(&table_info, schema_options, quote_style)?;
    stmt.partitions = partitions.map(|mut p| {
        p.set_quote(quote_style);
        p
    });
    let sql = format!("{}", stmt);
    let columns = vec![
        Arc::new(StringVector::from(vec![table_name])) as _,
        Arc::new(StringVector::from(vec![sql])) as _,
    ];
    let records = RecordBatches::try_from_columns(SHOW_CREATE_TABLE_OUTPUT_SCHEMA.clone(), columns)
        .context(error::CreateRecordBatchSnafu)?;

    Ok(Output::new_with_record_batches(records))
}

pub fn show_create_foreign_table_for_pg(
    table: TableRef,
    _query_ctx: QueryContextRef,
) -> Result<Output> {
    let table_info = table.table_info();

    let table_meta = &table_info.meta;
    let table_name = &table_info.name;
    let schema = &table_info.meta.schema;
    let is_metric_engine = is_metric_engine(&table_meta.engine);

    let columns = schema
        .column_schemas()
        .iter()
        .filter_map(|c| {
            if is_metric_engine && is_metric_engine_internal_column(&c.name) {
                None
            } else {
                Some(format!(
                    "\"{}\" {}",
                    c.name,
                    c.data_type.postgres_datatype_name()
                ))
            }
        })
        .join(",\n  ");

    let sql = format!(
        r#"CREATE FOREIGN TABLE ft_{} (
  {}
)
SERVER greptimedb
OPTIONS (table_name '{}')"#,
        table_name, columns, table_name
    );

    let columns = vec![
        Arc::new(StringVector::from(vec![table_name.clone()])) as _,
        Arc::new(StringVector::from(vec![sql])) as _,
    ];
    let records = RecordBatches::try_from_columns(SHOW_CREATE_TABLE_OUTPUT_SCHEMA.clone(), columns)
        .context(error::CreateRecordBatchSnafu)?;

    Ok(Output::new_with_record_batches(records))
}

pub fn show_create_view(
    view_name: ObjectName,
    definition: &str,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let mut parser_ctx =
        ParserContext::new(query_ctx.sql_dialect(), definition).context(error::SqlSnafu)?;

    let Statement::CreateView(create_view) =
        parser_ctx.parse_statement().context(error::SqlSnafu)?
    else {
        // MUST be `CreateView` statement.
        unreachable!();
    };

    let stmt = CreateView {
        name: view_name.clone(),
        columns: create_view.columns,
        query: create_view.query,
        or_replace: create_view.or_replace,
        if_not_exists: create_view.if_not_exists,
    };

    let sql = format!("{}", stmt);
    let columns = vec![
        Arc::new(StringVector::from(vec![view_name.to_string()])) as _,
        Arc::new(StringVector::from(vec![sql])) as _,
    ];
    let records = RecordBatches::try_from_columns(SHOW_CREATE_VIEW_OUTPUT_SCHEMA.clone(), columns)
        .context(error::CreateRecordBatchSnafu)?;

    Ok(Output::new_with_record_batches(records))
}

/// Execute [`ShowViews`] statement and return the [`Output`] if success.
pub async fn show_views(
    stmt: ShowViews,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe = show_views_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for [`ShowViews`] without executing it.
pub async fn show_views_dataframe(
    stmt: &ShowViews,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let schema_name = if let Some(database) = &stmt.database {
        database.clone()
    } else {
        query_ctx.current_schema()
    };

    let projects = vec![(tables::TABLE_NAME, VIEWS_COLUMN)];
    let filters = vec![
        col(tables::TABLE_SCHEMA).eq(lit(schema_name.clone())),
        col(tables::TABLE_CATALOG).eq(lit(query_ctx.current_catalog())),
    ];
    let like_field = Some(tables::TABLE_NAME);
    let sort = vec![col(tables::TABLE_NAME).sort(true, true)];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        VIEWS,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        &stmt.kind,
    )
    .await
}

/// Execute [`ShowFlows`] statement and return the [`Output`] if success.
pub async fn show_flows(
    stmt: ShowFlows,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe = show_flows_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for [`ShowFlows`] without executing it.
pub async fn show_flows_dataframe(
    stmt: &ShowFlows,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let projects = vec![(flows::FLOW_NAME, FLOWS_COLUMN)];
    let filters = vec![col(flows::TABLE_CATALOG).eq(lit(query_ctx.current_catalog()))];
    let like_field = Some(flows::FLOW_NAME);
    let sort = vec![col(flows::FLOW_NAME).sort(true, true)];

    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx,
        FLOWS,
        vec![],
        projects,
        filters,
        like_field,
        sort,
        &stmt.kind,
    )
    .await
}

/// Execute [`ShowFlowStatus`] statement and return the [`Output`] if success.
pub async fn show_flow_status(
    stmt: ShowFlowStatus,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let projects = vec![
        (flow_statistics::FLOW_ID, flow_statistics::FLOW_ID),
        (flow_statistics::FLOW_NAME, flow_statistics::FLOW_NAME),
        (flow_statistics::START_TIME, flow_statistics::START_TIME),
        (
            flow_statistics::LAST_EXECUTION_TIME,
            flow_statistics::LAST_EXECUTION_TIME,
        ),
        (
            flow_statistics::UPTIME_SECONDS,
            flow_statistics::UPTIME_SECONDS,
        ),
        (flow_statistics::STATE_SIZE, flow_statistics::STATE_SIZE),
    ];
    let like_field = Some(flow_statistics::FLOW_NAME);
    let sort = vec![col(flow_statistics::FLOW_NAME).sort(true, true)];

    query_from_information_schema_table(
        query_engine,
        catalog_manager,
        query_ctx,
        FLOW_STATISTICS,
        vec![],
        projects,
        vec![],
        like_field,
        sort,
        stmt.kind,
    )
    .await
}

#[cfg(feature = "enterprise")]
pub async fn show_triggers(
    stmt: sql::statements::show::trigger::ShowTriggers,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    const TRIGGER_NAME: &str = "trigger_name";
    const TRIGGERS_COLUMN: &str = "Triggers";

    let projects = vec![(TRIGGER_NAME, TRIGGERS_COLUMN)];
    let like_field = Some(TRIGGER_NAME);
    let sort = vec![col(TRIGGER_NAME).sort(true, true)];

    query_from_information_schema_table(
        query_engine,
        catalog_manager,
        query_ctx,
        catalog::information_schema::TRIGGERS,
        vec![],
        projects,
        vec![],
        like_field,
        sort,
        stmt.kind,
    )
    .await
}

pub fn show_create_flow(
    flow_name: ObjectName,
    flow_val: FlowInfoValue,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let mut parser_ctx =
        ParserContext::new(query_ctx.sql_dialect(), flow_val.raw_sql()).context(error::SqlSnafu)?;

    let query = parser_ctx.parse_statement().context(error::SqlSnafu)?;

    // since prom ql will parse `now()` to a fixed time, we need to not use it for generating raw query
    let raw_query = match &query {
        Statement::Tql(_) => flow_val.raw_sql().clone(),
        _ => query.to_string(),
    };

    let query = Box::new(SqlOrTql::try_from_statement(query, &raw_query).context(error::SqlSnafu)?);

    let comment = if flow_val.comment().is_empty() {
        None
    } else {
        Some(flow_val.comment().clone())
    };

    let stmt = CreateFlow {
        flow_name,
        sink_table_name: ObjectName::from(vec![
            Ident::new(&flow_val.sink_table_name().schema_name),
            Ident::new(&flow_val.sink_table_name().table_name),
        ]),
        // notice we don't want `OR REPLACE` and `IF NOT EXISTS` in same sql since it's unclear what to do
        // so we set `or_replace` to false.
        or_replace: false,
        if_not_exists: true,
        expire_after: flow_val.expire_after(),
        eval_interval: flow_val.eval_interval(),
        eval_offset: effective_eval_schedule_from_flow_info(&flow_val)
            .map_err(BoxedError::new)
            .context(error::QueryExecutionSnafu)?
            .map(|schedule| schedule.anchor_secs)
            .filter(|anchor_secs| *anchor_secs != 0),
        comment,
        flow_options: OptionMap::from_filtered_string_map(
            flow_val.options(),
            &[FlowType::FLOW_TYPE_KEY],
        ),
        query,
    };

    let sql = format!("{}", stmt);
    let columns = vec![
        Arc::new(StringVector::from(vec![flow_val.flow_name().clone()])) as _,
        Arc::new(StringVector::from(vec![sql])) as _,
    ];
    let records = RecordBatches::try_from_columns(SHOW_CREATE_FLOW_OUTPUT_SCHEMA.clone(), columns)
        .context(error::CreateRecordBatchSnafu)?;

    Ok(Output::new_with_record_batches(records))
}

pub fn describe_table(table: TableRef) -> Result<Output> {
    let table_info = table.table_info();
    let columns_schemas = table_info.meta.schema.column_schemas();
    let columns = vec![
        describe_column_names(columns_schemas),
        describe_column_types(columns_schemas),
        describe_column_keys(columns_schemas, &table_info.meta.primary_key_indices),
        describe_column_nullables(columns_schemas),
        describe_column_defaults(columns_schemas),
        describe_column_semantic_types(columns_schemas, &table_info.meta.primary_key_indices),
    ];
    let records = RecordBatches::try_from_columns(DESCRIBE_TABLE_OUTPUT_SCHEMA.clone(), columns)
        .context(error::CreateRecordBatchSnafu)?;
    Ok(Output::new_with_record_batches(records))
}

fn describe_column_names(columns_schemas: &[ColumnSchema]) -> VectorRef {
    Arc::new(StringVector::from_iterator(
        columns_schemas.iter().map(|cs| cs.name.as_str()),
    ))
}

fn describe_column_types(columns_schemas: &[ColumnSchema]) -> VectorRef {
    Arc::new(StringVector::from(
        columns_schemas
            .iter()
            .map(|cs| describe_column_type_name(&cs.data_type))
            .collect::<Vec<_>>(),
    ))
}

fn describe_column_type_name(data_type: &ConcreteDataType) -> String {
    match data_type {
        ConcreteDataType::Json(json_type) if json_type.is_json2() => "Json2".to_string(),
        data_type => data_type.name(),
    }
}

fn describe_column_keys(
    columns_schemas: &[ColumnSchema],
    primary_key_indices: &[usize],
) -> VectorRef {
    Arc::new(StringVector::from_iterator(
        columns_schemas.iter().enumerate().map(|(i, cs)| {
            if cs.is_time_index() || primary_key_indices.contains(&i) {
                PRI_KEY
            } else {
                ""
            }
        }),
    ))
}

fn describe_column_nullables(columns_schemas: &[ColumnSchema]) -> VectorRef {
    Arc::new(StringVector::from_iterator(columns_schemas.iter().map(
        |cs| {
            if cs.is_nullable() { YES_STR } else { NO_STR }
        },
    )))
}

fn describe_column_defaults(columns_schemas: &[ColumnSchema]) -> VectorRef {
    Arc::new(StringVector::from(
        columns_schemas
            .iter()
            .map(|cs| {
                cs.default_constraint()
                    .map_or(String::from(""), |dc| dc.to_string())
            })
            .collect::<Vec<String>>(),
    ))
}

fn describe_column_semantic_types(
    columns_schemas: &[ColumnSchema],
    primary_key_indices: &[usize],
) -> VectorRef {
    Arc::new(StringVector::from_iterator(
        columns_schemas.iter().enumerate().map(|(i, cs)| {
            if primary_key_indices.contains(&i) {
                SEMANTIC_TYPE_PRIMARY_KEY
            } else if cs.is_time_index() {
                SEMANTIC_TYPE_TIME_INDEX
            } else {
                SEMANTIC_TYPE_FIELD
            }
        }),
    ))
}

// lists files in the frontend to reduce unnecessary scan requests repeated in each datanode.
pub async fn prepare_file_table_files(
    options: &HashMap<String, String>,
    local_file_access: &LocalFileAccess,
) -> Result<(ObjectStore, Vec<String>)> {
    let url = options
        .get(FILE_TABLE_LOCATION_KEY)
        .context(error::MissingRequiredFieldSnafu {
            name: FILE_TABLE_LOCATION_KEY,
        })?;

    let regex = options
        .get(FILE_TABLE_PATTERN_KEY)
        .map(|x| Regex::new(x))
        .transpose()
        .context(error::BuildRegexSnafu)?;
    let backend = build_backend_with_path(url, options, local_file_access)
        .await
        .context(error::BuildBackendSnafu)?;
    let source = if let Some(filename) = backend.object_path {
        Source::Filename(filename)
    } else {
        Source::Dir
    };
    let lister = Lister::new(backend.object_store.clone(), source, url.clone(), regex);
    // If we scan files in a directory every time the database restarts,
    // then it might lead to a potential undefined behavior:
    // If a user adds a file with an incompatible schema to that directory,
    // it will make the external table unavailable.
    let files = lister
        .list()
        .await
        .context(error::ListObjectsSnafu)?
        .into_iter()
        .filter_map(|entry| {
            if entry.path().ends_with('/') {
                None
            } else {
                Some(entry.path().to_string())
            }
        })
        .collect::<Vec<_>>();
    Ok((backend.object_store, files))
}

pub async fn infer_file_table_schema(
    object_store: &ObjectStore,
    files: &[String],
    options: &HashMap<String, String>,
) -> Result<Schema> {
    let format = parse_file_table_format(options)?;
    let merged = infer_schemas(object_store, files, format.as_ref())
        .await
        .context(error::InferSchemaSnafu)?;
    Schema::try_from(merged).context(error::ConvertSchemaSnafu)
}

// Converts the file column schemas to table column schemas.
// Returns the column schemas and the time index column name.
//
// More specifically, this function will do the following:
// 1. Add a default time index column if there is no time index column
//    in the file column schemas, or
// 2. If the file column schemas contain a column with name conflicts with
//    the default time index column, it will replace the column schema
//    with the default one.
pub fn file_column_schemas_to_table(
    file_column_schemas: &[ColumnSchema],
) -> (Vec<ColumnSchema>, String) {
    let mut column_schemas = file_column_schemas.to_owned();
    if let Some(time_index_column) = column_schemas.iter().find(|c| c.is_time_index()) {
        let time_index = time_index_column.name.clone();
        return (column_schemas, time_index);
    }

    let timestamp_type = ConcreteDataType::timestamp_millisecond_datatype();
    let default_zero = Value::Timestamp(Timestamp::new_millisecond(0));
    let timestamp_column_schema = ColumnSchema::new(greptime_timestamp(), timestamp_type, false)
        .with_time_index(true)
        .with_default_constraint(Some(ColumnDefaultConstraint::Value(default_zero)))
        .unwrap();

    if let Some(column_schema) = column_schemas
        .iter_mut()
        .find(|column_schema| column_schema.name == greptime_timestamp())
    {
        // Replace the column schema with the default one
        *column_schema = timestamp_column_schema;
    } else {
        column_schemas.push(timestamp_column_schema);
    }

    (column_schemas, greptime_timestamp().to_string())
}

/// This function checks if the column schemas from a file can be matched with
/// the column schemas of a table.
///
/// More specifically, for each column seen in the table schema,
/// - If the same column does exist in the file schema, it checks if the data
///   type of the file column can be casted into the form of the table column.
/// - If the same column does not exist in the file schema, it checks if the
///   table column is nullable or has a default constraint.
pub fn check_file_to_table_schema_compatibility(
    file_column_schemas: &[ColumnSchema],
    table_column_schemas: &[ColumnSchema],
) -> Result<()> {
    let file_schemas_map = file_column_schemas
        .iter()
        .map(|s| (s.name.clone(), s))
        .collect::<HashMap<_, _>>();

    for table_column in table_column_schemas {
        if let Some(file_column) = file_schemas_map.get(&table_column.name) {
            // TODO(zhongzc): a temporary solution, we should use `can_cast_to` once it's ready.
            ensure!(
                file_column
                    .data_type
                    .can_arrow_type_cast_to(&table_column.data_type),
                error::ColumnSchemaIncompatibleSnafu {
                    column: table_column.name.clone(),
                    file_type: file_column.data_type.clone(),
                    table_type: table_column.data_type.clone(),
                }
            );
        } else {
            ensure!(
                table_column.is_nullable() || table_column.default_constraint().is_some(),
                error::ColumnSchemaNoDefaultSnafu {
                    column: table_column.name.clone(),
                }
            );
        }
    }

    Ok(())
}

fn parse_file_table_format(options: &HashMap<String, String>) -> Result<Box<dyn FileFormat>> {
    Ok(
        match Format::try_from(options).context(error::ParseFileFormatSnafu)? {
            Format::Csv(format) => Box::new(format),
            Format::Json(format) => Box::new(format),
            Format::Parquet(format) => Box::new(format),
            Format::Orc(format) => Box::new(format),
        },
    )
}

pub async fn show_processlist(
    stmt: ShowProcessList,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<Output> {
    let dataframe =
        show_processlist_dataframe(&stmt, query_engine, catalog_manager, query_ctx).await?;
    dataframe_to_output(dataframe).await
}

/// Builds the [`DataFrame`] for `SHOW PROCESSLIST` without executing it.
pub async fn show_processlist_dataframe(
    stmt: &ShowProcessList,
    query_engine: &QueryEngineRef,
    catalog_manager: &CatalogManagerRef,
    query_ctx: QueryContextRef,
) -> Result<DataFrame> {
    let projects = if stmt.full {
        vec![
            (process_list::ID, "Id"),
            (process_list::CATALOG, "Catalog"),
            (process_list::SCHEMAS, "Schema"),
            (process_list::CLIENT, "Client"),
            (process_list::FRONTEND, "Frontend"),
            (process_list::START_TIMESTAMP, "Start Time"),
            (process_list::ELAPSED_TIME, "Elapsed Time"),
            (process_list::QUERY, "Query"),
        ]
    } else {
        vec![
            (process_list::ID, "Id"),
            (process_list::CATALOG, "Catalog"),
            (process_list::QUERY, "Query"),
            (process_list::ELAPSED_TIME, "Elapsed Time"),
        ]
    };

    let filters = if query_ctx.current_user().is_admin() {
        vec![]
    } else {
        vec![col(process_list::CATALOG).eq(lit(query_ctx.current_catalog()))]
    };
    let like_field = None;
    let sort = vec![col("id").sort(true, true)];
    query_from_information_schema_dataframe(
        query_engine,
        catalog_manager,
        query_ctx.clone(),
        "process_list",
        vec![],
        projects,
        filters,
        like_field,
        sort,
        &ShowKind::All,
    )
    .await
}

#[cfg(test)]
mod test {
    use std::sync::Arc;

    use common_query::{Output, OutputData};
    use common_recordbatch::{RecordBatch, RecordBatches};
    use common_time::Timezone;
    use common_time::timestamp::TimeUnit;
    use datatypes::prelude::ConcreteDataType;
    use datatypes::schema::{ColumnDefaultConstraint, ColumnSchema, Schema, SchemaRef};
    use datatypes::types::json_type::JsonNativeType;
    use datatypes::vectors::{StringVector, TimestampMillisecondVector, UInt32Vector, VectorRef};
    use session::context::QueryContextBuilder;
    use snafu::ResultExt;
    use sql::ast::{Ident, ObjectName};
    use sql::statements::show::ShowVariables;
    use table::TableRef;
    use table::test_util::MemTable;

    use super::{describe_column_type_name, show_variable};
    use crate::error;
    use crate::error::Result;
    use crate::sql::{
        DESCRIBE_TABLE_OUTPUT_SCHEMA, NO_STR, SEMANTIC_TYPE_FIELD, SEMANTIC_TYPE_TIME_INDEX,
        YES_STR, describe_table,
    };

    #[test]
    fn test_describe_table_multiple_columns() -> Result<()> {
        let table_name = "test_table";
        let schema = vec![
            ColumnSchema::new("t1", ConcreteDataType::uint32_datatype(), true),
            ColumnSchema::new(
                "t2",
                ConcreteDataType::timestamp_datatype(TimeUnit::Millisecond),
                false,
            )
            .with_default_constraint(Some(ColumnDefaultConstraint::Function(String::from(
                "current_timestamp()",
            ))))
            .unwrap()
            .with_time_index(true),
        ];
        let data = vec![
            Arc::new(UInt32Vector::from_slice([0])) as _,
            Arc::new(TimestampMillisecondVector::from_slice([0])) as _,
        ];
        let expected_columns = vec![
            Arc::new(StringVector::from(vec!["t1", "t2"])) as _,
            Arc::new(StringVector::from(vec!["UInt32", "TimestampMillisecond"])) as _,
            Arc::new(StringVector::from(vec!["", "PRI"])) as _,
            Arc::new(StringVector::from(vec![YES_STR, NO_STR])) as _,
            Arc::new(StringVector::from(vec!["", "current_timestamp()"])) as _,
            Arc::new(StringVector::from(vec![
                SEMANTIC_TYPE_FIELD,
                SEMANTIC_TYPE_TIME_INDEX,
            ])) as _,
        ];

        describe_table_test_by_schema(table_name, schema, data, expected_columns)
    }

    #[test]
    fn test_describe_column_type_name_json2() {
        assert_eq!(
            describe_column_type_name(&ConcreteDataType::json2(JsonNativeType::Null)),
            "Json2"
        );
        assert_eq!(
            describe_column_type_name(&ConcreteDataType::uint32_datatype()),
            "UInt32"
        );
    }

    fn describe_table_test_by_schema(
        table_name: &str,
        schema: Vec<ColumnSchema>,
        data: Vec<VectorRef>,
        expected_columns: Vec<VectorRef>,
    ) -> Result<()> {
        let table_schema = SchemaRef::new(Schema::new(schema));
        let table = prepare_describe_table(table_name, table_schema, data);

        let expected =
            RecordBatches::try_from_columns(DESCRIBE_TABLE_OUTPUT_SCHEMA.clone(), expected_columns)
                .context(error::CreateRecordBatchSnafu)?;

        if let OutputData::RecordBatches(res) = describe_table(table)?.data {
            assert_eq!(res.take(), expected.take());
        } else {
            panic!("describe table must return record batch");
        }

        Ok(())
    }

    fn prepare_describe_table(
        table_name: &str,
        table_schema: SchemaRef,
        data: Vec<VectorRef>,
    ) -> TableRef {
        let record_batch = RecordBatch::new(table_schema, data).unwrap();
        MemTable::table(table_name, record_batch)
    }

    #[test]
    fn test_show_variable() {
        assert_eq!(
            exec_show_variable("SYSTEM_TIME_ZONE", "Asia/Shanghai").unwrap(),
            "UTC"
        );
        assert_eq!(
            exec_show_variable("SYSTEM_TIMEZONE", "Asia/Shanghai").unwrap(),
            "UTC"
        );
        assert_eq!(
            exec_show_variable("TIME_ZONE", "Asia/Shanghai").unwrap(),
            "Asia/Shanghai"
        );
        assert_eq!(
            exec_show_variable("TIMEZONE", "Asia/Shanghai").unwrap(),
            "Asia/Shanghai"
        );
        assert!(exec_show_variable("TIME ZONE", "Asia/Shanghai").is_err());
        assert!(exec_show_variable("SYSTEM TIME ZONE", "Asia/Shanghai").is_err());
    }

    fn exec_show_variable(variable: &str, tz: &str) -> Result<String> {
        let stmt = ShowVariables {
            variable: ObjectName::from(vec![Ident::new(variable)]),
        };
        let ctx = Arc::new(
            QueryContextBuilder::default()
                .timezone(Timezone::from_tz_string(tz).unwrap())
                .build(),
        );
        match show_variable(stmt, ctx) {
            Ok(Output {
                data: OutputData::RecordBatches(record),
                ..
            }) => {
                let record = record.take().first().cloned().unwrap();
                Ok(record.iter_column_as_string(0).next().unwrap().unwrap())
            }
            Ok(_) => unreachable!(),
            Err(e) => Err(e),
        }
    }
}