diff --git a/.github/workflows/license.yml b/.github/workflows/license.yml new file mode 100644 index 00000000..c851bff3 --- /dev/null +++ b/.github/workflows/license.yml @@ -0,0 +1,39 @@ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you 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. + +name: license + +# trigger for all PRs and changes to main +on: + push: + branches: + - main + pull_request: + +jobs: + + rat: + name: Release Audit Tool (RAT) + runs-on: ubuntu-latest + steps: + - uses: actions/checkout@v4 + - name: Setup Python + uses: actions/setup-python@v5 + with: + python-version: 3.8 + - name: Audit licenses + run: ./dev/release/run-rat.sh . diff --git a/Cargo.toml b/Cargo.toml index 07e44f66..2e48f9d0 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -18,7 +18,7 @@ [package] name = "sqlparser" description = "Extensible SQL Lexer and Parser with support for ANSI SQL:2011" -version = "0.57.0" +version = "0.58.0" authors = ["Apache DataFusion "] homepage = "https://github.com/apache/datafusion-sqlparser-rs" documentation = "https://docs.rs/sqlparser/" diff --git a/README.md b/README.md index 666be17c..9dfe5081 100644 --- a/README.md +++ b/README.md @@ -21,7 +21,7 @@ [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0) [![Version](https://img.shields.io/crates/v/sqlparser.svg)](https://crates.io/crates/sqlparser) -[![Build Status](https://github.com/sqlparser-rs/sqlparser-rs/workflows/Rust/badge.svg?branch=main)](https://github.com/sqlparser-rs/sqlparser-rs/actions?query=workflow%3ARust+branch%3Amain) +[![Build Status](https://github.com/apache/datafusion-sqlparser-rs/actions/workflows/rust.yml/badge.svg)](https://github.com/sqlparser-rs/sqlparser-rs/actions?query=workflow%3ARust+branch%3Amain) [![Coverage Status](https://coveralls.io/repos/github/sqlparser-rs/sqlparser-rs/badge.svg?branch=main)](https://coveralls.io/github/sqlparser-rs/sqlparser-rs?branch=main) [![Gitter Chat](https://badges.gitter.im/sqlparser-rs/community.svg)](https://gitter.im/sqlparser-rs/community?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge) diff --git a/changelog/0.58.0.md b/changelog/0.58.0.md new file mode 100644 index 00000000..27a985d7 --- /dev/null +++ b/changelog/0.58.0.md @@ -0,0 +1,106 @@ + + +# sqlparser-rs 0.58.0 Changelog + +This release consists of 47 commits from 18 contributors. See credits at the end of this changelog for more information. + +**Fixed bugs:** + +- fix: parse snowflake fetch clause [#1894](https://github.com/apache/datafusion-sqlparser-rs/pull/1894) (Vedin) + +**Documentation updates:** + +- docs: Update rust badge [#1943](https://github.com/apache/datafusion-sqlparser-rs/pull/1943) (Olexandr88) + +**Other:** + +- Add license header check to CI [#1888](https://github.com/apache/datafusion-sqlparser-rs/pull/1888) (alamb) +- Add support of parsing struct field's options in BigQuery [#1890](https://github.com/apache/datafusion-sqlparser-rs/pull/1890) (git-hulk) +- Fix parsing error when having fields after nested struct in BigQuery [#1897](https://github.com/apache/datafusion-sqlparser-rs/pull/1897) (git-hulk) +- Extend exception handling [#1884](https://github.com/apache/datafusion-sqlparser-rs/pull/1884) (bombsimon) +- Postgres: Add support for text search types [#1889](https://github.com/apache/datafusion-sqlparser-rs/pull/1889) (MohamedAbdeen21) +- Fix `limit` in subqueries [#1899](https://github.com/apache/datafusion-sqlparser-rs/pull/1899) (Dimchikkk) +- Use `IndexColumn` in all index definitions [#1900](https://github.com/apache/datafusion-sqlparser-rs/pull/1900) (mvzink) +- Support procedure argmode [#1901](https://github.com/apache/datafusion-sqlparser-rs/pull/1901) (ZacJW) +- Fix `impl Ord for Ident` [#1893](https://github.com/apache/datafusion-sqlparser-rs/pull/1893) (eliaperantoni) +- Snowflake: support multiple column options in `CREATE VIEW` [#1891](https://github.com/apache/datafusion-sqlparser-rs/pull/1891) (eliaperantoni) +- Add support for `LANGUAGE` clause in `CREATE PROCEDURE` [#1903](https://github.com/apache/datafusion-sqlparser-rs/pull/1903) (ZacJW) +- Fix clippy lints on 1.88.0 [#1910](https://github.com/apache/datafusion-sqlparser-rs/pull/1910) (iffyio) +- Snowflake: Add support for future grants [#1906](https://github.com/apache/datafusion-sqlparser-rs/pull/1906) (yoavcloud) +- Support for Map values in ClickHouse settings [#1896](https://github.com/apache/datafusion-sqlparser-rs/pull/1896) (solontsev) +- Fix join precedence for non-snowflake queries [#1905](https://github.com/apache/datafusion-sqlparser-rs/pull/1905) (Dimchikkk) +- Support remaining pipe operators [#1879](https://github.com/apache/datafusion-sqlparser-rs/pull/1879) (simonvandel) +- Make `GenericDialect` support from-first syntax [#1911](https://github.com/apache/datafusion-sqlparser-rs/pull/1911) (simonvandel) +- Redshift utf8 idents [#1915](https://github.com/apache/datafusion-sqlparser-rs/pull/1915) (yoavcloud) +- DuckDB: Add support for multiple `TRIM` arguments [#1916](https://github.com/apache/datafusion-sqlparser-rs/pull/1916) (ryanschneider) +- Redshift alter column type no set [#1912](https://github.com/apache/datafusion-sqlparser-rs/pull/1912) (yoavcloud) +- Postgres: support `ADD CONSTRAINT NOT VALID` and `VALIDATE CONSTRAINT` [#1908](https://github.com/apache/datafusion-sqlparser-rs/pull/1908) (achristmascarl) +- Add support for MySQL MEMBER OF [#1917](https://github.com/apache/datafusion-sqlparser-rs/pull/1917) (yoavcloud) +- Add span for `Expr::TypedString` [#1919](https://github.com/apache/datafusion-sqlparser-rs/pull/1919) (feral-dot-io) +- Support for Postgres `CREATE SERVER` [#1914](https://github.com/apache/datafusion-sqlparser-rs/pull/1914) (solontsev) +- Change tag and policy names to `ObjectName` [#1892](https://github.com/apache/datafusion-sqlparser-rs/pull/1892) (eliaperantoni) +- Add support for NULL escape char in pattern match searches [#1913](https://github.com/apache/datafusion-sqlparser-rs/pull/1913) (yoavcloud) +- Add support for dropping multiple columns in Snowflake [#1918](https://github.com/apache/datafusion-sqlparser-rs/pull/1918) (yoavcloud) +- Align Snowflake dialect to new test of reserved keywords [#1924](https://github.com/apache/datafusion-sqlparser-rs/pull/1924) (yoavcloud) +- Make `GenericDialect` support trailing commas in projections [#1921](https://github.com/apache/datafusion-sqlparser-rs/pull/1921) (simonvandel) +- Add support for several Snowflake grant statements [#1922](https://github.com/apache/datafusion-sqlparser-rs/pull/1922) (yoavcloud) +- Clickhouse: support empty parenthesized options [#1925](https://github.com/apache/datafusion-sqlparser-rs/pull/1925) (solontsev) +- Add Snowflake `COPY/REVOKE CURRENT GRANTS` option [#1926](https://github.com/apache/datafusion-sqlparser-rs/pull/1926) (yoavcloud) +- Add support for Snowflake identifier function [#1929](https://github.com/apache/datafusion-sqlparser-rs/pull/1929) (yoavcloud) +- Add support for granting privileges to procedures and functions in Snowflake [#1930](https://github.com/apache/datafusion-sqlparser-rs/pull/1930) (yoavcloud) +- Add support for `+` char in Snowflake stage names [#1935](https://github.com/apache/datafusion-sqlparser-rs/pull/1935) (yoavcloud) +- Snowflake Reserved SQL Keywords as Implicit Table Alias [#1934](https://github.com/apache/datafusion-sqlparser-rs/pull/1934) (yoavcloud) +- Add support for Redshift `SELECT * EXCLUDE` [#1936](https://github.com/apache/datafusion-sqlparser-rs/pull/1936) (yoavcloud) +- Support optional semicolon between statements [#1937](https://github.com/apache/datafusion-sqlparser-rs/pull/1937) (yoavcloud) +- Snowflake: support trailing options in `CREATE TABLE` [#1931](https://github.com/apache/datafusion-sqlparser-rs/pull/1931) (yoavcloud) +- MSSQL: Add support for EXEC output and default keywords [#1940](https://github.com/apache/datafusion-sqlparser-rs/pull/1940) (yoavcloud) +- Add identifier unicode support in Mysql, Postgres and Redshift [#1933](https://github.com/apache/datafusion-sqlparser-rs/pull/1933) (etgarperets) +- Add identifier start unicode support for Postegres, MySql and Redshift [#1944](https://github.com/apache/datafusion-sqlparser-rs/pull/1944) (etgarperets) +- Fix for Postgres regex and like binary operators [#1928](https://github.com/apache/datafusion-sqlparser-rs/pull/1928) (solontsev) +- Snowflake: Improve accuracy of lookahead in implicit LIMIT alias [#1941](https://github.com/apache/datafusion-sqlparser-rs/pull/1941) (yoavcloud) +- Add support for `DROP USER` statement [#1951](https://github.com/apache/datafusion-sqlparser-rs/pull/1951) (yoavcloud) + +## Credits + +Thank you to everyone who contributed to this release. Here is a breakdown of commits (PRs merged) per contributor. + +``` + 19 Yoav Cohen + 4 Sergey Olontsev + 3 Elia Perantoni + 3 Simon Vandel Sillesen + 2 Dima + 2 ZacJW + 2 etgarperets + 2 hulk + 1 Andrew Lamb + 1 Denys Tsomenko + 1 Ifeanyi Ubah + 1 Michael Victor Zink + 1 Mohamed Abdeen + 1 Olexandr88 + 1 Ryan Schneider + 1 Simon Sawert + 1 carl + 1 feral-dot-io +``` + +Thank you also to everyone who contributed in other ways such as filing issues, reviewing PRs, and providing feedback on this release. + diff --git a/dev/release/rat_exclude_files.txt b/dev/release/rat_exclude_files.txt index 562eec2f..280b1bce 100644 --- a/dev/release/rat_exclude_files.txt +++ b/dev/release/rat_exclude_files.txt @@ -1,7 +1,8 @@ -# Files to exclude from the Apache Rat (license) check -.gitignore .tool-versions +target/* +**.gitignore +rat.txt dev/release/rat_exclude_files.txt -fuzz/.gitignore sqlparser_bench/img/flamegraph.svg - +**Cargo.lock +filtered_rat.txt diff --git a/examples/cli.rs b/examples/cli.rs index 0252fca7..08a40a6d 100644 --- a/examples/cli.rs +++ b/examples/cli.rs @@ -63,7 +63,7 @@ $ cargo run --example cli - [--dialectname] }; let contents = if filename == "-" { - println!("Parsing from stdin using {:?}", dialect); + println!("Parsing from stdin using {dialect:?}"); let mut buf = Vec::new(); stdin() .read_to_end(&mut buf) diff --git a/sqlparser_bench/benches/sqlparser_bench.rs b/sqlparser_bench/benches/sqlparser_bench.rs index 24c59c07..6132ee43 100644 --- a/sqlparser_bench/benches/sqlparser_bench.rs +++ b/sqlparser_bench/benches/sqlparser_bench.rs @@ -45,25 +45,24 @@ fn basic_queries(c: &mut Criterion) { let large_statement = { let expressions = (0..1000) - .map(|n| format!("FN_{}(COL_{})", n, n)) + .map(|n| format!("FN_{n}(COL_{n})")) .collect::>() .join(", "); let tables = (0..1000) - .map(|n| format!("TABLE_{}", n)) + .map(|n| format!("TABLE_{n}")) .collect::>() .join(" JOIN "); let where_condition = (0..1000) - .map(|n| format!("COL_{} = {}", n, n)) + .map(|n| format!("COL_{n} = {n}")) .collect::>() .join(" OR "); let order_condition = (0..1000) - .map(|n| format!("COL_{} DESC", n)) + .map(|n| format!("COL_{n} DESC")) .collect::>() .join(", "); format!( - "SELECT {} FROM {} WHERE {} ORDER BY {}", - expressions, tables, where_condition, order_condition + "SELECT {expressions} FROM {tables} WHERE {where_condition} ORDER BY {order_condition}" ) }; diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs index 3a4958c9..0897f2db 100644 --- a/src/ast/data_type.rs +++ b/src/ast/data_type.rs @@ -446,6 +446,14 @@ pub enum DataType { /// /// [PostgreSQL]: https://www.postgresql.org/docs/9.5/functions-geometry.html GeometricType(GeometricTypeKind), + /// PostgreSQL text search vectors, see [PostgreSQL]. + /// + /// [PostgreSQL]: https://www.postgresql.org/docs/17/datatype-textsearch.html + TsVector, + /// PostgreSQL text search query, see [PostgreSQL]. + /// + /// [PostgreSQL]: https://www.postgresql.org/docs/17/datatype-textsearch.html + TsQuery, } impl fmt::Display for DataType { @@ -658,7 +666,7 @@ impl fmt::Display for DataType { } DataType::Enum(vals, bits) => { match bits { - Some(bits) => write!(f, "ENUM{}", bits), + Some(bits) => write!(f, "ENUM{bits}"), None => write!(f, "ENUM"), }?; write!(f, "(")?; @@ -706,16 +714,16 @@ impl fmt::Display for DataType { } // ClickHouse DataType::Nullable(data_type) => { - write!(f, "Nullable({})", data_type) + write!(f, "Nullable({data_type})") } DataType::FixedString(character_length) => { - write!(f, "FixedString({})", character_length) + write!(f, "FixedString({character_length})") } DataType::LowCardinality(data_type) => { - write!(f, "LowCardinality({})", data_type) + write!(f, "LowCardinality({data_type})") } DataType::Map(key_data_type, value_data_type) => { - write!(f, "Map({}, {})", key_data_type, value_data_type) + write!(f, "Map({key_data_type}, {value_data_type})") } DataType::Tuple(fields) => { write!(f, "Tuple({})", display_comma_separated(fields)) @@ -737,7 +745,9 @@ impl fmt::Display for DataType { DataType::NamedTable { name, columns } => { write!(f, "{} TABLE ({})", name, display_comma_separated(columns)) } - DataType::GeometricType(kind) => write!(f, "{}", kind), + DataType::GeometricType(kind) => write!(f, "{kind}"), + DataType::TsVector => write!(f, "TSVECTOR"), + DataType::TsQuery => write!(f, "TSQUERY"), } } } @@ -932,7 +942,7 @@ impl fmt::Display for CharacterLength { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { CharacterLength::IntegerLength { length, unit } => { - write!(f, "{}", length)?; + write!(f, "{length}")?; if let Some(unit) = unit { write!(f, " {unit}")?; } @@ -987,7 +997,7 @@ impl fmt::Display for BinaryLength { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { BinaryLength::IntegerLength { length } => { - write!(f, "{}", length)?; + write!(f, "{length}")?; } BinaryLength::Max => { write!(f, "MAX")?; diff --git a/src/ast/dcl.rs b/src/ast/dcl.rs index 735ab0cc..07989407 100644 --- a/src/ast/dcl.rs +++ b/src/ast/dcl.rs @@ -173,7 +173,7 @@ impl fmt::Display for AlterRoleOperation { in_database, } => { if let Some(database_name) = in_database { - write!(f, "IN DATABASE {} ", database_name)?; + write!(f, "IN DATABASE {database_name} ")?; } match config_value { @@ -187,7 +187,7 @@ impl fmt::Display for AlterRoleOperation { in_database, } => { if let Some(database_name) = in_database { - write!(f, "IN DATABASE {} ", database_name)?; + write!(f, "IN DATABASE {database_name} ")?; } match config_name { @@ -218,15 +218,15 @@ impl fmt::Display for Use { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { f.write_str("USE ")?; match self { - Use::Catalog(name) => write!(f, "CATALOG {}", name), - Use::Schema(name) => write!(f, "SCHEMA {}", name), - Use::Database(name) => write!(f, "DATABASE {}", name), - Use::Warehouse(name) => write!(f, "WAREHOUSE {}", name), - Use::Role(name) => write!(f, "ROLE {}", name), + Use::Catalog(name) => write!(f, "CATALOG {name}"), + Use::Schema(name) => write!(f, "SCHEMA {name}"), + Use::Database(name) => write!(f, "DATABASE {name}"), + Use::Warehouse(name) => write!(f, "WAREHOUSE {name}"), + Use::Role(name) => write!(f, "ROLE {name}"), Use::SecondaryRoles(secondary_roles) => { - write!(f, "SECONDARY ROLES {}", secondary_roles) + write!(f, "SECONDARY ROLES {secondary_roles}") } - Use::Object(name) => write!(f, "{}", name), + Use::Object(name) => write!(f, "{name}"), Use::Default => write!(f, "DEFAULT"), } } diff --git a/src/ast/ddl.rs b/src/ast/ddl.rs index 059c6196..51e05784 100644 --- a/src/ast/ddl.rs +++ b/src/ast/ddl.rs @@ -30,11 +30,11 @@ use sqlparser_derive::{Visit, VisitMut}; use crate::ast::value::escape_single_quote_string; use crate::ast::{ - display_comma_separated, display_separated, CommentDef, CreateFunctionBody, + display_comma_separated, display_separated, ArgMode, CommentDef, CreateFunctionBody, CreateFunctionUsing, DataType, Expr, FunctionBehavior, FunctionCalledOnNull, - FunctionDeterminismSpecifier, FunctionParallel, Ident, MySQLColumnPosition, ObjectName, - OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, SqlOption, Tag, Value, - ValueWithSpan, + FunctionDeterminismSpecifier, FunctionParallel, Ident, IndexColumn, MySQLColumnPosition, + ObjectName, OperateFunctionArg, OrderByExpr, ProjectionSelect, SequenceOptions, SqlOption, Tag, + Value, ValueWithSpan, }; use crate::keywords::Keyword; use crate::tokenizer::Token; @@ -57,7 +57,7 @@ impl fmt::Display for ReplicaIdentity { ReplicaIdentity::None => f.write_str("NONE"), ReplicaIdentity::Full => f.write_str("FULL"), ReplicaIdentity::Default => f.write_str("DEFAULT"), - ReplicaIdentity::Index(idx) => write!(f, "USING INDEX {}", idx), + ReplicaIdentity::Index(idx) => write!(f, "USING INDEX {idx}"), } } } @@ -67,8 +67,11 @@ impl fmt::Display for ReplicaIdentity { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum AlterTableOperation { - /// `ADD ` - AddConstraint(TableConstraint), + /// `ADD [NOT VALID]` + AddConstraint { + constraint: TableConstraint, + not_valid: bool, + }, /// `ADD [COLUMN] [IF NOT EXISTS] ` AddColumn { /// `[COLUMN]`. @@ -137,10 +140,10 @@ pub enum AlterTableOperation { name: Ident, drop_behavior: Option, }, - /// `DROP [ COLUMN ] [ IF EXISTS ] [ CASCADE ]` + /// `DROP [ COLUMN ] [ IF EXISTS ] [ , , ... ] [ CASCADE ]` DropColumn { has_column_keyword: bool, - column_name: Ident, + column_names: Vec, if_exists: bool, drop_behavior: Option, }, @@ -344,6 +347,10 @@ pub enum AlterTableOperation { equals: bool, value: ValueWithSpan, }, + /// `VALIDATE CONSTRAINT ` + ValidateConstraint { + name: Ident, + }, } /// An `ALTER Policy` (`Statement::AlterPolicy`) operation @@ -450,7 +457,7 @@ pub enum Owner { impl fmt::Display for Owner { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - Owner::Ident(ident) => write!(f, "{}", ident), + Owner::Ident(ident) => write!(f, "{ident}"), Owner::CurrentRole => write!(f, "CURRENT_ROLE"), Owner::CurrentUser => write!(f, "CURRENT_USER"), Owner::SessionUser => write!(f, "SESSION_USER"), @@ -494,7 +501,16 @@ impl fmt::Display for AlterTableOperation { display_separated(new_partitions, " "), ine = if *if_not_exists { " IF NOT EXISTS" } else { "" } ), - AlterTableOperation::AddConstraint(c) => write!(f, "ADD {c}"), + AlterTableOperation::AddConstraint { + not_valid, + constraint, + } => { + write!(f, "ADD {constraint}")?; + if *not_valid { + write!(f, " NOT VALID")?; + } + Ok(()) + } AlterTableOperation::AddColumn { column_keyword, if_not_exists, @@ -525,7 +541,7 @@ impl fmt::Display for AlterTableOperation { if *if_not_exists { write!(f, " IF NOT EXISTS")?; } - write!(f, " {} ({})", name, query) + write!(f, " {name} ({query})") } AlterTableOperation::Algorithm { equals, algorithm } => { write!( @@ -540,7 +556,7 @@ impl fmt::Display for AlterTableOperation { if *if_exists { write!(f, " IF EXISTS")?; } - write!(f, " {}", name) + write!(f, " {name}") } AlterTableOperation::MaterializeProjection { if_exists, @@ -551,9 +567,9 @@ impl fmt::Display for AlterTableOperation { if *if_exists { write!(f, " IF EXISTS")?; } - write!(f, " {}", name)?; + write!(f, " {name}")?; if let Some(partition) = partition { - write!(f, " IN PARTITION {}", partition)?; + write!(f, " IN PARTITION {partition}")?; } Ok(()) } @@ -566,9 +582,9 @@ impl fmt::Display for AlterTableOperation { if *if_exists { write!(f, " IF EXISTS")?; } - write!(f, " {}", name)?; + write!(f, " {name}")?; if let Some(partition) = partition { - write!(f, " IN PARTITION {}", partition)?; + write!(f, " IN PARTITION {partition}")?; } Ok(()) } @@ -615,7 +631,7 @@ impl fmt::Display for AlterTableOperation { AlterTableOperation::DropIndex { name } => write!(f, "DROP INDEX {name}"), AlterTableOperation::DropColumn { has_column_keyword, - column_name, + column_names: column_name, if_exists, drop_behavior, } => write!( @@ -623,7 +639,7 @@ impl fmt::Display for AlterTableOperation { "DROP {}{}{}{}", if *has_column_keyword { "COLUMN " } else { "" }, if *if_exists { "IF EXISTS " } else { "" }, - column_name, + display_comma_separated(column_name), match drop_behavior { None => "", Some(DropBehavior::Restrict) => " RESTRICT", @@ -772,6 +788,9 @@ impl fmt::Display for AlterTableOperation { AlterTableOperation::ReplicaIdentity { identity } => { write!(f, "REPLICA IDENTITY {identity}") } + AlterTableOperation::ValidateConstraint { name } => { + write!(f, "VALIDATE CONSTRAINT {name}") + } } } } @@ -893,7 +912,10 @@ pub enum AlterColumnOperation { data_type: DataType, /// PostgreSQL specific using: Option, + /// Set to true if the statement includes the `SET DATA TYPE` keywords + had_set: bool, }, + /// `ADD GENERATED { ALWAYS | BY DEFAULT } AS IDENTITY [ ( sequence_options ) ]` /// /// Note: this is a PostgreSQL-specific operation. @@ -914,12 +936,19 @@ impl fmt::Display for AlterColumnOperation { AlterColumnOperation::DropDefault => { write!(f, "DROP DEFAULT") } - AlterColumnOperation::SetDataType { data_type, using } => { - if let Some(expr) = using { - write!(f, "SET DATA TYPE {data_type} USING {expr}") - } else { - write!(f, "SET DATA TYPE {data_type}") + AlterColumnOperation::SetDataType { + data_type, + using, + had_set, + } => { + if *had_set { + write!(f, "SET DATA ")?; } + write!(f, "TYPE {data_type}")?; + if let Some(expr) = using { + write!(f, " USING {expr}")?; + } + Ok(()) } AlterColumnOperation::AddGenerated { generated_as, @@ -979,7 +1008,7 @@ pub enum TableConstraint { /// [1]: IndexType index_type: Option, /// Identifiers of the columns that are unique. - columns: Vec, + columns: Vec, index_options: Vec, characteristics: Option, /// Optional Postgres nulls handling: `[ NULLS [ NOT ] DISTINCT ]` @@ -1015,7 +1044,7 @@ pub enum TableConstraint { /// [1]: IndexType index_type: Option, /// Identifiers of the columns that form the primary key. - columns: Vec, + columns: Vec, index_options: Vec, characteristics: Option, }, @@ -1060,7 +1089,7 @@ pub enum TableConstraint { /// [1]: IndexType index_type: Option, /// Referred column identifier list. - columns: Vec, + columns: Vec, }, /// MySQLs [fulltext][1] definition. Since the [`SPATIAL`][2] definition is exactly the same, /// and MySQL displays both the same way, it is part of this definition as well. @@ -1083,7 +1112,7 @@ pub enum TableConstraint { /// Optional index name. opt_index_name: Option, /// Referred column identifier list. - columns: Vec, + columns: Vec, }, } @@ -1168,7 +1197,7 @@ impl fmt::Display for TableConstraint { write!(f, " ON UPDATE {action}")?; } if let Some(characteristics) = characteristics { - write!(f, " {}", characteristics)?; + write!(f, " {characteristics}")?; } Ok(()) } @@ -1308,7 +1337,7 @@ impl fmt::Display for IndexType { Self::SPGiST => write!(f, "SPGIST"), Self::BRIN => write!(f, "BRIN"), Self::Bloom => write!(f, "BLOOM"), - Self::Custom(name) => write!(f, "{}", name), + Self::Custom(name) => write!(f, "{name}"), } } } @@ -1367,11 +1396,16 @@ impl fmt::Display for NullsDistinctOption { pub struct ProcedureParam { pub name: Ident, pub data_type: DataType, + pub mode: Option, } impl fmt::Display for ProcedureParam { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { - write!(f, "{} {}", self.name, self.data_type) + if let Some(mode) = &self.mode { + write!(f, "{mode} {} {}", self.name, self.data_type) + } else { + write!(f, "{} {}", self.name, self.data_type) + } } } @@ -1421,17 +1455,41 @@ impl fmt::Display for ColumnDef { pub struct ViewColumnDef { pub name: Ident, pub data_type: Option, - pub options: Option>, + pub options: Option, +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum ColumnOptions { + CommaSeparated(Vec), + SpaceSeparated(Vec), +} + +impl ColumnOptions { + pub fn as_slice(&self) -> &[ColumnOption] { + match self { + ColumnOptions::CommaSeparated(options) => options.as_slice(), + ColumnOptions::SpaceSeparated(options) => options.as_slice(), + } + } } impl fmt::Display for ViewColumnDef { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.name)?; if let Some(data_type) = self.data_type.as_ref() { - write!(f, " {}", data_type)?; + write!(f, " {data_type}")?; } if let Some(options) = self.options.as_ref() { - write!(f, " {}", display_comma_separated(options.as_slice()))?; + match options { + ColumnOptions::CommaSeparated(column_options) => { + write!(f, " {}", display_comma_separated(column_options.as_slice()))?; + } + ColumnOptions::SpaceSeparated(column_options) => { + write!(f, " {}", display_separated(column_options.as_slice(), " "))? + } + } } Ok(()) } @@ -1651,7 +1709,7 @@ pub struct ColumnPolicyProperty { /// ``` /// [Snowflake]: https://docs.snowflake.com/en/sql-reference/sql/create-table pub with: bool, - pub policy_name: Ident, + pub policy_name: ObjectName, pub using_columns: Option>, } @@ -1816,7 +1874,7 @@ impl fmt::Display for ColumnOption { } => { write!(f, "{}", if *is_primary { "PRIMARY KEY" } else { "UNIQUE" })?; if let Some(characteristics) = characteristics { - write!(f, " {}", characteristics)?; + write!(f, " {characteristics}")?; } Ok(()) } @@ -1838,7 +1896,7 @@ impl fmt::Display for ColumnOption { write!(f, " ON UPDATE {action}")?; } if let Some(characteristics) = characteristics { - write!(f, " {}", characteristics)?; + write!(f, " {characteristics}")?; } Ok(()) } @@ -1898,7 +1956,7 @@ impl fmt::Display for ColumnOption { write!(f, "{parameters}") } OnConflict(keyword) => { - write!(f, "ON CONFLICT {:?}", keyword)?; + write!(f, "ON CONFLICT {keyword:?}")?; Ok(()) } Policy(parameters) => { diff --git a/src/ast/dml.rs b/src/ast/dml.rs index 292650c8..e179f5d7 100644 --- a/src/ast/dml.rs +++ b/src/ast/dml.rs @@ -55,7 +55,7 @@ impl Display for IndexColumn { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.column)?; if let Some(operator_class) = &self.operator_class { - write!(f, " {}", operator_class)?; + write!(f, " {operator_class}")?; } Ok(()) } @@ -266,7 +266,7 @@ impl Display for CreateTable { name = self.name, )?; if let Some(on_cluster) = &self.on_cluster { - write!(f, " ON CLUSTER {}", on_cluster)?; + write!(f, " ON CLUSTER {on_cluster}")?; } if !self.columns.is_empty() || !self.constraints.is_empty() { f.write_str(" (")?; @@ -383,15 +383,15 @@ impl Display for CreateTable { match &self.table_options { options @ CreateTableOptions::With(_) | options @ CreateTableOptions::Plain(_) - | options @ CreateTableOptions::TableProperties(_) => write!(f, " {}", options)?, + | options @ CreateTableOptions::TableProperties(_) => write!(f, " {options}")?, _ => (), } if let Some(primary_key) = &self.primary_key { - write!(f, " PRIMARY KEY {}", primary_key)?; + write!(f, " PRIMARY KEY {primary_key}")?; } if let Some(order_by) = &self.order_by { - write!(f, " ORDER BY {}", order_by)?; + write!(f, " ORDER BY {order_by}")?; } if let Some(inherits) = &self.inherits { write!(f, " INHERITS ({})", display_comma_separated(inherits))?; @@ -403,7 +403,7 @@ impl Display for CreateTable { write!(f, " CLUSTER BY {cluster_by}")?; } if let options @ CreateTableOptions::Options(_) = &self.table_options { - write!(f, " {}", options)?; + write!(f, " {options}")?; } if let Some(external_volume) = self.external_volume.as_ref() { write!(f, " EXTERNAL_VOLUME = '{external_volume}'")?; diff --git a/src/ast/helpers/key_value_options.rs b/src/ast/helpers/key_value_options.rs index 06f028dd..796bfd5e 100644 --- a/src/ast/helpers/key_value_options.rs +++ b/src/ast/helpers/key_value_options.rs @@ -67,7 +67,7 @@ impl fmt::Display for KeyValueOptions { } else { f.write_str(" ")?; } - write!(f, "{}", option)?; + write!(f, "{option}")?; } } Ok(()) diff --git a/src/ast/helpers/stmt_create_table.rs b/src/ast/helpers/stmt_create_table.rs index d66a869b..60b8fb2a 100644 --- a/src/ast/helpers/stmt_create_table.rs +++ b/src/ast/helpers/stmt_create_table.rs @@ -383,6 +383,26 @@ impl CreateTableBuilder { self } + /// Returns true if the statement has exactly one source of info on the schema of the new table. + /// This is Snowflake-specific, some dialects allow more than one source. + pub(crate) fn validate_schema_info(&self) -> bool { + let mut sources = 0; + if !self.columns.is_empty() { + sources += 1; + } + if self.query.is_some() { + sources += 1; + } + if self.like.is_some() { + sources += 1; + } + if self.clone.is_some() { + sources += 1; + } + + sources == 1 + } + pub fn build(self) -> Statement { Statement::CreateTable(CreateTable { or_replace: self.or_replace, diff --git a/src/ast/mod.rs b/src/ast/mod.rs index a1d8ff6f..1f5311af 100644 --- a/src/ast/mod.rs +++ b/src/ast/mod.rs @@ -28,6 +28,7 @@ use helpers::{ stmt_data_loading::{FileStagingCommand, StageLoadSelectItemKind}, }; +use core::cmp::Ordering; use core::ops::Deref; use core::{ fmt::{self, Display}, @@ -60,13 +61,14 @@ pub use self::ddl::{ AlterColumnOperation, AlterConnectorOwner, AlterIndexOperation, AlterPolicyOperation, AlterTableAlgorithm, AlterTableLock, AlterTableOperation, AlterType, AlterTypeAddValue, AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue, - ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnPolicy, ColumnPolicyProperty, - ConstraintCharacteristics, CreateConnector, CreateDomain, CreateFunction, Deduplicate, - DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, IdentityParameters, - IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, IdentityPropertyOrder, - IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, - ProcedureParam, ReferentialAction, ReplicaIdentity, TableConstraint, TagsColumnOption, - UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef, + ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, ColumnPolicy, + ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, CreateFunction, + Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode, + IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, + IdentityPropertyOrder, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner, + Partition, ProcedureParam, ReferentialAction, ReplicaIdentity, TableConstraint, + TagsColumnOption, UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, + ViewColumnDef, }; pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert}; pub use self::operator::{BinaryOperator, UnaryOperator}; @@ -172,7 +174,7 @@ fn format_statement_list(f: &mut fmt::Formatter, statements: &[Statement]) -> fm } /// An identifier, decomposed into its value or character data and the quote style. -#[derive(Debug, Clone, PartialOrd, Ord)] +#[derive(Debug, Clone)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct Ident { @@ -214,6 +216,35 @@ impl core::hash::Hash for Ident { impl Eq for Ident {} +impl PartialOrd for Ident { + fn partial_cmp(&self, other: &Self) -> Option { + Some(self.cmp(other)) + } +} + +impl Ord for Ident { + fn cmp(&self, other: &Self) -> Ordering { + let Ident { + value, + quote_style, + // exhaustiveness check; we ignore spans in ordering + span: _, + } = self; + + let Ident { + value: other_value, + quote_style: other_quote_style, + // exhaustiveness check; we ignore spans in ordering + span: _, + } = other; + + // First compare by value, then by quote_style + value + .cmp(other_value) + .then_with(|| quote_style.cmp(other_quote_style)) + } +} + impl Ident { /// Create a new identifier with the given value and no quotes and an empty span. pub fn new(value: S) -> Self @@ -313,12 +344,14 @@ impl fmt::Display for ObjectName { #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub enum ObjectNamePart { Identifier(Ident), + Function(ObjectNamePartFunction), } impl ObjectNamePart { pub fn as_ident(&self) -> Option<&Ident> { match self { ObjectNamePart::Identifier(ident) => Some(ident), + ObjectNamePart::Function(_) => None, } } } @@ -326,11 +359,31 @@ impl ObjectNamePart { impl fmt::Display for ObjectNamePart { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - ObjectNamePart::Identifier(ident) => write!(f, "{}", ident), + ObjectNamePart::Identifier(ident) => write!(f, "{ident}"), + ObjectNamePart::Function(func) => write!(f, "{func}"), } } } +/// An object name part that consists of a function that dynamically +/// constructs identifiers. +/// +/// - [Snowflake](https://docs.snowflake.com/en/sql-reference/identifier-literal) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ObjectNamePartFunction { + pub name: Ident, + pub args: Vec, +} + +impl fmt::Display for ObjectNamePartFunction { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{}(", self.name)?; + write!(f, "{})", display_comma_separated(&self.args)) + } +} + /// Represents an Array Expression, either /// `ARRAY[..]`, or `[..]` #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -428,14 +481,22 @@ impl fmt::Display for Interval { pub struct StructField { pub field_name: Option, pub field_type: DataType, + /// Struct field options. + /// See [BigQuery](https://cloud.google.com/bigquery/docs/reference/standard-sql/data-definition-language#column_name_and_column_schema) + pub options: Option>, } impl fmt::Display for StructField { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { if let Some(name) = &self.field_name { - write!(f, "{name} {}", self.field_type) + write!(f, "{name} {}", self.field_type)?; } else { - write!(f, "{}", self.field_type) + write!(f, "{}", self.field_type)?; + } + if let Some(options) = &self.options { + write!(f, " OPTIONS({})", display_separated(options, ", ")) + } else { + Ok(()) } } } @@ -740,7 +801,7 @@ pub enum Expr { /// `[ NOT ] IN (SELECT ...)` InSubquery { expr: Box, - subquery: Box, + subquery: Box, negated: bool, }, /// `[ NOT ] IN UNNEST(array_expression)` @@ -770,7 +831,7 @@ pub enum Expr { any: bool, expr: Box, pattern: Box, - escape_char: Option, + escape_char: Option, }, /// `ILIKE` (case-insensitive `LIKE`) ILike { @@ -780,14 +841,14 @@ pub enum Expr { any: bool, expr: Box, pattern: Box, - escape_char: Option, + escape_char: Option, }, /// SIMILAR TO regex SimilarTo { negated: bool, expr: Box, pattern: Box, - escape_char: Option, + escape_char: Option, }, /// MySQL: RLIKE regex or REGEXP regex RLike { @@ -957,7 +1018,7 @@ pub enum Expr { data_type: DataType, /// The value of the constant. /// Hint: you can unwrap the string value using `value.into_string()`. - value: Value, + value: ValueWithSpan, }, /// Scalar function call e.g. `LEFT(foo, 5)` Function(Function), @@ -1085,6 +1146,8 @@ pub enum Expr { /// [Databricks](https://docs.databricks.com/en/sql/language-manual/sql-ref-lambda-functions.html) /// [DuckDb](https://duckdb.org/docs/sql/functions/lambda.html) Lambda(LambdaFunction), + /// Checks membership of a value in a JSON array + MemberOf(MemberOf), } impl Expr { @@ -1171,8 +1234,8 @@ pub enum AccessExpr { impl fmt::Display for AccessExpr { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { - AccessExpr::Dot(expr) => write!(f, ".{}", expr), - AccessExpr::Subscript(subscript) => write!(f, "[{}]", subscript), + AccessExpr::Dot(expr) => write!(f, ".{expr}"), + AccessExpr::Subscript(subscript) => write!(f, "[{subscript}]"), } } } @@ -1374,12 +1437,12 @@ impl fmt::Display for Expr { match self { Expr::Identifier(s) => write!(f, "{s}"), Expr::Wildcard(_) => f.write_str("*"), - Expr::QualifiedWildcard(prefix, _) => write!(f, "{}.*", prefix), + Expr::QualifiedWildcard(prefix, _) => write!(f, "{prefix}.*"), Expr::CompoundIdentifier(s) => write!(f, "{}", display_separated(s, ".")), Expr::CompoundFieldAccess { root, access_chain } => { - write!(f, "{}", root)?; + write!(f, "{root}")?; for field in access_chain { - write!(f, "{}", field)?; + write!(f, "{field}")?; } Ok(()) } @@ -1447,7 +1510,7 @@ impl fmt::Display for Expr { } => match escape_char { Some(ch) => write!( f, - "{} {}LIKE {}{} ESCAPE '{}'", + "{} {}LIKE {}{} ESCAPE {}", expr, if *negated { "NOT " } else { "" }, if *any { "ANY " } else { "" }, @@ -1472,7 +1535,7 @@ impl fmt::Display for Expr { } => match escape_char { Some(ch) => write!( f, - "{} {}ILIKE {}{} ESCAPE '{}'", + "{} {}ILIKE {}{} ESCAPE {}", expr, if *negated { "NOT " } else { "" }, if *any { "ANY" } else { "" }, @@ -1508,7 +1571,7 @@ impl fmt::Display for Expr { } => { let not_ = if *negated { "NOT " } else { "" }; if form.is_none() { - write!(f, "{} IS {}NORMALIZED", expr, not_) + write!(f, "{expr} IS {not_}NORMALIZED") } else { write!( f, @@ -1527,7 +1590,7 @@ impl fmt::Display for Expr { } => match escape_char { Some(ch) => write!( f, - "{} {}SIMILAR TO {} ESCAPE '{}'", + "{} {}SIMILAR TO {} ESCAPE {}", expr, if *negated { "NOT " } else { "" }, pattern, @@ -1830,7 +1893,7 @@ impl fmt::Display for Expr { } } Expr::Named { expr, name } => { - write!(f, "{} AS {}", expr, name) + write!(f, "{expr} AS {name}") } Expr::Dictionary(fields) => { write!(f, "{{{}}}", display_comma_separated(fields)) @@ -1873,6 +1936,7 @@ impl fmt::Display for Expr { } Expr::Prior(expr) => write!(f, "PRIOR {expr}"), Expr::Lambda(lambda) => write!(f, "{lambda}"), + Expr::MemberOf(member_of) => write!(f, "{member_of}"), } } } @@ -2386,7 +2450,7 @@ impl fmt::Display for ConditionalStatements { } Ok(()) } - ConditionalStatements::BeginEnd(bes) => write!(f, "{}", bes), + ConditionalStatements::BeginEnd(bes) => write!(f, "{bes}"), } } } @@ -2906,9 +2970,7 @@ impl Display for Set { write!( f, "SET {modifier}ROLE {role_name}", - modifier = context_modifier - .map(|m| format!("{}", m)) - .unwrap_or_default() + modifier = context_modifier.map(|m| format!("{m}")).unwrap_or_default() ) } Self::SetSessionParam(kind) => write!(f, "SET {kind}"), @@ -2941,7 +3003,7 @@ impl Display for Set { charset_name, collation_name, } => { - write!(f, "SET NAMES {}", charset_name)?; + write!(f, "SET NAMES {charset_name}")?; if let Some(collation) = collation_name { f.write_str(" COLLATE ")?; @@ -2964,7 +3026,7 @@ impl Display for Set { write!( f, "SET {}{}{} = {}", - scope.map(|s| format!("{}", s)).unwrap_or_default(), + scope.map(|s| format!("{s}")).unwrap_or_default(), if *hivevar { "HIVEVAR:" } else { "" }, variable, display_comma_separated(values) @@ -2982,6 +3044,36 @@ impl From for Statement { } } +/// A representation of a `WHEN` arm with all the identifiers catched and the statements to execute +/// for the arm. +/// +/// Snowflake: +/// BigQuery: +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct ExceptionWhen { + pub idents: Vec, + pub statements: Vec, +} + +impl Display for ExceptionWhen { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!( + f, + "WHEN {idents} THEN", + idents = display_separated(&self.idents, " OR ") + )?; + + if !self.statements.is_empty() { + write!(f, " ")?; + format_statement_list(f, &self.statements)?; + } + + Ok(()) + } +} + /// A top-level statement (SELECT, INSERT, CREATE, etc.) #[allow(clippy::large_enum_variant)] #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] @@ -3248,6 +3340,8 @@ pub enum Statement { secret_type: Ident, options: Vec, }, + /// A `CREATE SERVER` statement. + CreateServer(CreateServerStatement), /// ```sql /// CREATE POLICY /// ``` @@ -3670,17 +3764,20 @@ pub enum Statement { /// END; /// ``` statements: Vec, - /// Statements of an exception clause. + /// Exception handling with exception clauses. /// Example: /// ```sql - /// BEGIN - /// SELECT 1; - /// EXCEPTION WHEN ERROR THEN - /// SELECT 2; - /// SELECT 3; - /// END; + /// EXCEPTION + /// WHEN EXCEPTION_1 THEN + /// SELECT 2; + /// WHEN EXCEPTION_2 OR EXCEPTION_3 THEN + /// SELECT 3; + /// WHEN OTHER THEN + /// SELECT 4; + /// ``` /// - exception_statements: Option>, + /// + exception: Option>, /// TRUE if the statement has an `END` keyword. has_end_keyword: bool, }, @@ -3873,6 +3970,7 @@ pub enum Statement { or_alter: bool, name: ObjectName, params: Option>, + language: Option, body: ConditionalStatements, }, /// ```sql @@ -3920,6 +4018,7 @@ pub enum Statement { with_grant_option: bool, as_grantor: Option, granted_by: Option, + current_grants: Option, }, /// ```sql /// DENY privileges ON object TO grantees @@ -3960,6 +4059,12 @@ pub enum Statement { immediate: bool, into: Vec, using: Vec, + /// Whether the last parameter is the return value of the procedure + /// MSSQL: + output: bool, + /// Whether to invoke the procedure with the default parameter values + /// MSSQL: + default: bool, }, /// ```sql /// PREPARE name [ ( data_type [, ...] ) ] AS statement @@ -4021,7 +4126,7 @@ pub enum Statement { /// A SQL query that specifies what to explain statement: Box, /// Optional output format of explain - format: Option, + format: Option, /// Postgres style utility options, `(analyze, verbose true)` options: Option>, }, @@ -4173,7 +4278,7 @@ pub enum Statement { /// ```sql /// NOTIFY channel [ , payload ] /// ``` - /// send a notification event together with an optional “payload” string to channel + /// send a notification event together with an optional "payload" string to channel /// /// See Postgres NOTIFY { @@ -4236,6 +4341,28 @@ pub enum Statement { Return(ReturnStatement), } +/// ```sql +/// {COPY | REVOKE} CURRENT GRANTS +/// ``` +/// +/// - [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/grant-ownership#optional-parameters) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum CurrentGrantsKind { + CopyCurrentGrants, + RevokeCurrentGrants, +} + +impl fmt::Display for CurrentGrantsKind { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + match self { + CurrentGrantsKind::CopyCurrentGrants => write!(f, "COPY CURRENT GRANTS"), + CurrentGrantsKind::RevokeCurrentGrants => write!(f, "REVOKE CURRENT GRANTS"), + } + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -4332,7 +4459,7 @@ impl fmt::Display for Statement { write!(f, "{describe_alias} ")?; if let Some(format) = hive_format { - write!(f, "{} ", format)?; + write!(f, "{format} ")?; } if *has_table_keyword { write!(f, "TABLE ")?; @@ -4367,7 +4494,7 @@ impl fmt::Display for Statement { } if let Some(format) = format { - write!(f, "FORMAT {format} ")?; + write!(f, "{format} ")?; } if let Some(options) = options { @@ -4776,6 +4903,7 @@ impl fmt::Display for Statement { name, or_alter, params, + language, body, } => { write!( @@ -4791,6 +4919,10 @@ impl fmt::Display for Statement { } } + if let Some(language) = language { + write!(f, " LANGUAGE {language}")?; + } + write!(f, " AS {body}") } Statement::CreateMacro { @@ -5099,6 +5231,9 @@ impl fmt::Display for Statement { write!(f, " )")?; Ok(()) } + Statement::CreateServer(stmt) => { + write!(f, "{stmt}") + } Statement::CreatePolicy { name, table_name, @@ -5163,7 +5298,7 @@ impl fmt::Display for Statement { if *only { write!(f, "ONLY ")?; } - write!(f, "{name} ", name = name)?; + write!(f, "{name} ")?; if let Some(cluster) = on_cluster { write!(f, "ON CLUSTER {cluster} ")?; } @@ -5241,7 +5376,7 @@ impl fmt::Display for Statement { )?; if !session_params.options.is_empty() { if *set { - write!(f, " {}", session_params)?; + write!(f, " {session_params}")?; } else { let options = session_params .options @@ -5275,7 +5410,7 @@ impl fmt::Display for Statement { if *purge { " PURGE" } else { "" }, )?; if let Some(table_name) = table.as_ref() { - write!(f, " ON {}", table_name)?; + write!(f, " ON {table_name}")?; }; Ok(()) } @@ -5525,12 +5660,12 @@ impl fmt::Display for Statement { transaction, modifier, statements, - exception_statements, + exception, has_end_keyword, } => { if *syntax_begin { if let Some(modifier) = *modifier { - write!(f, "BEGIN {}", modifier)?; + write!(f, "BEGIN {modifier}")?; } else { write!(f, "BEGIN")?; } @@ -5547,11 +5682,10 @@ impl fmt::Display for Statement { write!(f, " ")?; format_statement_list(f, statements)?; } - if let Some(exception_statements) = exception_statements { - write!(f, " EXCEPTION WHEN ERROR THEN")?; - if !exception_statements.is_empty() { - write!(f, " ")?; - format_statement_list(f, exception_statements)?; + if let Some(exception_when) = exception { + write!(f, " EXCEPTION")?; + for when in exception_when { + write!(f, " {when}")?; } } if *has_end_keyword { @@ -5567,7 +5701,7 @@ impl fmt::Display for Statement { if *end_syntax { write!(f, "END")?; if let Some(modifier) = *modifier { - write!(f, " {}", modifier)?; + write!(f, " {modifier}")?; } if *chain { write!(f, " AND CHAIN")?; @@ -5632,6 +5766,7 @@ impl fmt::Display for Statement { with_grant_option, as_grantor, granted_by, + current_grants, } => { write!(f, "GRANT {privileges} ")?; if let Some(objects) = objects { @@ -5641,6 +5776,9 @@ impl fmt::Display for Statement { if *with_grant_option { write!(f, " WITH GRANT OPTION")?; } + if let Some(current_grants) = current_grants { + write!(f, " {current_grants}")?; + } if let Some(grantor) = as_grantor { write!(f, " AS {grantor}")?; } @@ -5666,7 +5804,7 @@ impl fmt::Display for Statement { write!(f, " GRANTED BY {grantor}")?; } if let Some(cascade) = cascade { - write!(f, " {}", cascade)?; + write!(f, " {cascade}")?; } Ok(()) } @@ -5683,6 +5821,8 @@ impl fmt::Display for Statement { immediate, into, using, + output, + default, } => { let (open, close) = if *has_parentheses { ("(", ")") @@ -5703,6 +5843,12 @@ impl fmt::Display for Statement { if !using.is_empty() { write!(f, " USING {}", display_comma_separated(using))?; }; + if *output { + write!(f, " OUTPUT")?; + } + if *default { + write!(f, " DEFAULT")?; + } Ok(()) } Statement::Prepare { @@ -5845,13 +5991,13 @@ impl fmt::Display for Statement { if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, )?; if !directory_table_params.options.is_empty() { - write!(f, " DIRECTORY=({})", directory_table_params)?; + write!(f, " DIRECTORY=({directory_table_params})")?; } if !file_format.options.is_empty() { - write!(f, " FILE_FORMAT=({})", file_format)?; + write!(f, " FILE_FORMAT=({file_format})")?; } if !copy_options.options.is_empty() { - write!(f, " COPY_OPTIONS=({})", copy_options)?; + write!(f, " COPY_OPTIONS=({copy_options})")?; } if comment.is_some() { write!(f, " COMMENT='{}'", comment.as_ref().unwrap())?; @@ -5874,7 +6020,7 @@ impl fmt::Display for Statement { validation_mode, partition, } => { - write!(f, "COPY INTO {}", into)?; + write!(f, "COPY INTO {into}")?; if let Some(into_columns) = into_columns { write!(f, " ({})", display_comma_separated(into_columns))?; } @@ -5890,12 +6036,12 @@ impl fmt::Display for Statement { )?; } if let Some(from_obj_alias) = from_obj_alias { - write!(f, " AS {}", from_obj_alias)?; + write!(f, " AS {from_obj_alias}")?; } write!(f, ")")?; } else if let Some(from_obj) = from_obj { // Standard data load - write!(f, " FROM {}{}", from_obj, stage_params)?; + write!(f, " FROM {from_obj}{stage_params}")?; if let Some(from_obj_alias) = from_obj_alias { write!(f, " AS {from_obj_alias}")?; } @@ -5908,24 +6054,24 @@ impl fmt::Display for Statement { write!(f, " FILES = ('{}')", display_separated(files, "', '"))?; } if let Some(pattern) = pattern { - write!(f, " PATTERN = '{}'", pattern)?; + write!(f, " PATTERN = '{pattern}'")?; } if let Some(partition) = partition { write!(f, " PARTITION BY {partition}")?; } if !file_format.options.is_empty() { - write!(f, " FILE_FORMAT=({})", file_format)?; + write!(f, " FILE_FORMAT=({file_format})")?; } if !copy_options.options.is_empty() { match kind { CopyIntoSnowflakeKind::Table => { - write!(f, " COPY_OPTIONS=({})", copy_options)? + write!(f, " COPY_OPTIONS=({copy_options})")? } CopyIntoSnowflakeKind::Location => write!(f, " {copy_options}")?, } } if let Some(validation_mode) = validation_mode { - write!(f, " VALIDATION_MODE = {}", validation_mode)?; + write!(f, " VALIDATION_MODE = {validation_mode}")?; } Ok(()) } @@ -5971,10 +6117,10 @@ impl fmt::Display for Statement { } => { write!(f, "OPTIMIZE TABLE {name}")?; if let Some(on_cluster) = on_cluster { - write!(f, " ON CLUSTER {on_cluster}", on_cluster = on_cluster)?; + write!(f, " ON CLUSTER {on_cluster}")?; } if let Some(partition) = partition { - write!(f, " {partition}", partition = partition)?; + write!(f, " {partition}")?; } if *include_final { write!(f, " FINAL")?; @@ -6101,7 +6247,7 @@ impl fmt::Display for SetAssignment { write!( f, "{}{} = {}", - self.scope.map(|s| format!("{}", s)).unwrap_or_default(), + self.scope.map(|s| format!("{s}")).unwrap_or_default(), self.name, self.value ) @@ -6830,7 +6976,7 @@ impl fmt::Display for GranteeName { match self { GranteeName::ObjectName(name) => name.fmt(f), GranteeName::UserHost { user, host } => { - write!(f, "{}@{}", user, host) + write!(f, "{user}@{host}") } } } @@ -6845,6 +6991,24 @@ pub enum GrantObjects { AllSequencesInSchema { schemas: Vec }, /// Grant privileges on `ALL TABLES IN SCHEMA [, ...]` AllTablesInSchema { schemas: Vec }, + /// Grant privileges on `ALL VIEWS IN SCHEMA [, ...]` + AllViewsInSchema { schemas: Vec }, + /// Grant privileges on `ALL MATERIALIZED VIEWS IN SCHEMA [, ...]` + AllMaterializedViewsInSchema { schemas: Vec }, + /// Grant privileges on `ALL EXTERNAL TABLES IN SCHEMA [, ...]` + AllExternalTablesInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE SCHEMAS IN DATABASE [, ...]` + FutureSchemasInDatabase { databases: Vec }, + /// Grant privileges on `FUTURE TABLES IN SCHEMA [, ...]` + FutureTablesInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE VIEWS IN SCHEMA [, ...]` + FutureViewsInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE EXTERNAL TABLES IN SCHEMA [, ...]` + FutureExternalTablesInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE MATERIALIZED VIEWS IN SCHEMA [, ...]` + FutureMaterializedViewsInSchema { schemas: Vec }, + /// Grant privileges on `FUTURE SEQUENCES IN SCHEMA [, ...]` + FutureSequencesInSchema { schemas: Vec }, /// Grant privileges on specific databases Databases(Vec), /// Grant privileges on specific schemas @@ -6873,6 +7037,25 @@ pub enum GrantObjects { ReplicationGroup(Vec), /// Grant privileges on external volumes ExternalVolumes(Vec), + /// Grant privileges on a procedure. In dialects that + /// support overloading, the argument types must be specified. + /// + /// For example: + /// `GRANT USAGE ON PROCEDURE foo(varchar) TO ROLE role1` + Procedure { + name: ObjectName, + arg_types: Vec, + }, + + /// Grant privileges on a function. In dialects that + /// support overloading, the argument types must be specified. + /// + /// For example: + /// `GRANT USAGE ON FUNCTION foo(varchar) TO ROLE role1` + Function { + name: ObjectName, + arg_types: Vec, + }, } impl fmt::Display for GrantObjects { @@ -6913,6 +7096,69 @@ impl fmt::Display for GrantObjects { display_comma_separated(schemas) ) } + GrantObjects::AllExternalTablesInSchema { schemas } => { + write!( + f, + "ALL EXTERNAL TABLES IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::AllViewsInSchema { schemas } => { + write!( + f, + "ALL VIEWS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::AllMaterializedViewsInSchema { schemas } => { + write!( + f, + "ALL MATERIALIZED VIEWS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::FutureSchemasInDatabase { databases } => { + write!( + f, + "FUTURE SCHEMAS IN DATABASE {}", + display_comma_separated(databases) + ) + } + GrantObjects::FutureTablesInSchema { schemas } => { + write!( + f, + "FUTURE TABLES IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::FutureExternalTablesInSchema { schemas } => { + write!( + f, + "FUTURE EXTERNAL TABLES IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::FutureViewsInSchema { schemas } => { + write!( + f, + "FUTURE VIEWS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::FutureMaterializedViewsInSchema { schemas } => { + write!( + f, + "FUTURE MATERIALIZED VIEWS IN SCHEMA {}", + display_comma_separated(schemas) + ) + } + GrantObjects::FutureSequencesInSchema { schemas } => { + write!( + f, + "FUTURE SEQUENCES IN SCHEMA {}", + display_comma_separated(schemas) + ) + } GrantObjects::ResourceMonitors(objects) => { write!(f, "RESOURCE MONITOR {}", display_comma_separated(objects)) } @@ -6934,6 +7180,20 @@ impl fmt::Display for GrantObjects { GrantObjects::ExternalVolumes(objects) => { write!(f, "EXTERNAL VOLUME {}", display_comma_separated(objects)) } + GrantObjects::Procedure { name, arg_types } => { + write!(f, "PROCEDURE {name}")?; + if !arg_types.is_empty() { + write!(f, "({})", display_comma_separated(arg_types))?; + } + Ok(()) + } + GrantObjects::Function { name, arg_types } => { + write!(f, "FUNCTION {name}")?; + if !arg_types.is_empty() { + write!(f, "({})", display_comma_separated(arg_types))?; + } + Ok(()) + } } } } @@ -7000,7 +7260,7 @@ pub enum AssignmentTarget { impl fmt::Display for AssignmentTarget { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - AssignmentTarget::ColumnName(column) => write!(f, "{}", column), + AssignmentTarget::ColumnName(column) => write!(f, "{column}"), AssignmentTarget::Tuple(columns) => write!(f, "({})", display_comma_separated(columns)), } } @@ -7245,8 +7505,8 @@ impl fmt::Display for FunctionArguments { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { FunctionArguments::None => Ok(()), - FunctionArguments::Subquery(query) => write!(f, "({})", query), - FunctionArguments::List(args) => write!(f, "({})", args), + FunctionArguments::Subquery(query) => write!(f, "({query})"), + FunctionArguments::List(args) => write!(f, "({args})"), } } } @@ -7267,7 +7527,7 @@ pub struct FunctionArgumentList { impl fmt::Display for FunctionArgumentList { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { if let Some(duplicate_treatment) = self.duplicate_treatment { - write!(f, "{} ", duplicate_treatment)?; + write!(f, "{duplicate_treatment} ")?; } write!(f, "{}", display_comma_separated(&self.args))?; if !self.clauses.is_empty() { @@ -7327,7 +7587,7 @@ impl fmt::Display for FunctionArgumentClause { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { match self { FunctionArgumentClause::IgnoreOrRespectNulls(null_treatment) => { - write!(f, "{}", null_treatment) + write!(f, "{null_treatment}") } FunctionArgumentClause::OrderBy(order_by) => { write!(f, "ORDER BY {}", display_comma_separated(order_by)) @@ -7381,6 +7641,25 @@ impl fmt::Display for DuplicateTreatment { } } +#[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub enum AnalyzeFormatKind { + /// e.g. `EXPLAIN ANALYZE FORMAT JSON SELECT * FROM tbl` + Keyword(AnalyzeFormat), + /// e.g. `EXPLAIN ANALYZE FORMAT=JSON SELECT * FROM tbl` + Assignment(AnalyzeFormat), +} + +impl fmt::Display for AnalyzeFormatKind { + fn fmt(&self, f: &mut core::fmt::Formatter<'_>) -> core::fmt::Result { + match self { + AnalyzeFormatKind::Keyword(format) => write!(f, "FORMAT {format}"), + AnalyzeFormatKind::Assignment(format) => write!(f, "FORMAT={format}"), + } + } +} + #[derive(Debug, Copy, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -7388,6 +7667,8 @@ pub enum AnalyzeFormat { TEXT, GRAPHVIZ, JSON, + TRADITIONAL, + TREE, } impl fmt::Display for AnalyzeFormat { @@ -7396,6 +7677,8 @@ impl fmt::Display for AnalyzeFormat { AnalyzeFormat::TEXT => "TEXT", AnalyzeFormat::GRAPHVIZ => "GRAPHVIZ", AnalyzeFormat::JSON => "JSON", + AnalyzeFormat::TRADITIONAL => "TRADITIONAL", + AnalyzeFormat::TREE => "TREE", }) } } @@ -7508,6 +7791,7 @@ pub enum ObjectType { Sequence, Stage, Type, + User, } impl fmt::Display for ObjectType { @@ -7523,6 +7807,7 @@ impl fmt::Display for ObjectType { ObjectType::Sequence => "SEQUENCE", ObjectType::Stage => "STAGE", ObjectType::Type => "TYPE", + ObjectType::User => "USER", }) } } @@ -7783,12 +8068,12 @@ pub enum SqlOption { impl fmt::Display for SqlOption { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - SqlOption::Clustered(c) => write!(f, "{}", c), + SqlOption::Clustered(c) => write!(f, "{c}"), SqlOption::Ident(ident) => { - write!(f, "{}", ident) + write!(f, "{ident}") } SqlOption::KeyValue { key: name, value } => { - write!(f, "{} = {}", name, value) + write!(f, "{name} = {value}") } SqlOption::Partition { column_name, @@ -7828,7 +8113,7 @@ impl fmt::Display for SqlOption { SqlOption::NamedParenthesizedList(value) => { write!(f, "{} = ", value.key)?; if let Some(key) = &value.name { - write!(f, "{}", key)?; + write!(f, "{key}")?; } if !value.values.is_empty() { write!(f, "({})", display_comma_separated(&value.values))? @@ -7871,6 +8156,70 @@ impl fmt::Display for SecretOption { } } +/// A `CREATE SERVER` statement. +/// +/// [PostgreSQL Documentation](https://www.postgresql.org/docs/current/sql-createserver.html) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateServerStatement { + pub name: ObjectName, + pub if_not_exists: bool, + pub server_type: Option, + pub version: Option, + pub foreign_data_wrapper: ObjectName, + pub options: Option>, +} + +impl fmt::Display for CreateServerStatement { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + let CreateServerStatement { + name, + if_not_exists, + server_type, + version, + foreign_data_wrapper, + options, + } = self; + + write!( + f, + "CREATE SERVER {if_not_exists}{name} ", + if_not_exists = if *if_not_exists { "IF NOT EXISTS " } else { "" }, + )?; + + if let Some(st) = server_type { + write!(f, "TYPE {st} ")?; + } + + if let Some(v) = version { + write!(f, "VERSION {v} ")?; + } + + write!(f, "FOREIGN DATA WRAPPER {foreign_data_wrapper}")?; + + if let Some(o) = options { + write!(f, " OPTIONS ({o})", o = display_comma_separated(o))?; + } + + Ok(()) + } +} + +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct CreateServerOption { + pub key: Ident, + pub value: Ident, +} + +impl fmt::Display for CreateServerOption { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{} {}", self.key, self.value) + } +} + #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] @@ -7885,7 +8234,7 @@ impl fmt::Display for AttachDuckDBDatabaseOption { AttachDuckDBDatabaseOption::ReadOnly(Some(true)) => write!(f, "READ_ONLY true"), AttachDuckDBDatabaseOption::ReadOnly(Some(false)) => write!(f, "READ_ONLY false"), AttachDuckDBDatabaseOption::ReadOnly(None) => write!(f, "READ_ONLY"), - AttachDuckDBDatabaseOption::Type(t) => write!(f, "TYPE {}", t), + AttachDuckDBDatabaseOption::Type(t) => write!(f, "TYPE {t}"), } } } @@ -9198,12 +9547,12 @@ impl Display for RowAccessPolicy { #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct Tag { - pub key: Ident, + pub key: ObjectName, pub value: String, } impl Tag { - pub fn new(key: Ident, value: String) -> Self { + pub fn new(key: ObjectName, value: String) -> Self { Self { key, value } } } @@ -9408,10 +9757,10 @@ impl fmt::Display for ShowStatementIn { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.clause)?; if let Some(parent_type) = &self.parent_type { - write!(f, " {}", parent_type)?; + write!(f, " {parent_type}")?; } if let Some(parent_name) = &self.parent_name { - write!(f, " {}", parent_name)?; + write!(f, " {parent_name}")?; } Ok(()) } @@ -9492,7 +9841,7 @@ impl fmt::Display for TableObject { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { Self::TableName(table_name) => write!(f, "{table_name}"), - Self::TableFunction(func) => write!(f, "FUNCTION {}", func), + Self::TableFunction(func) => write!(f, "FUNCTION {func}"), } } } @@ -9680,7 +10029,7 @@ pub struct ReturnStatement { impl fmt::Display for ReturnStatement { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match &self.value { - Some(ReturnStatementValue::Expr(expr)) => write!(f, "RETURN {}", expr), + Some(ReturnStatementValue::Expr(expr)) => write!(f, "RETURN {expr}"), None => write!(f, "RETURN"), } } @@ -9729,8 +10078,31 @@ impl fmt::Display for NullInclusion { } } +/// Checks membership of a value in a JSON array +/// +/// Syntax: +/// ```sql +/// MEMBER OF() +/// ``` +/// [MySQL](https://dev.mysql.com/doc/refman/8.4/en/json-search-functions.html#operator_member-of) +#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)] +#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))] +#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] +pub struct MemberOf { + pub value: Box, + pub array: Box, +} + +impl fmt::Display for MemberOf { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + write!(f, "{} MEMBER OF({})", self.value, self.array) + } +} + #[cfg(test)] mod tests { + use crate::tokenizer::Location; + use super::*; #[test] @@ -10026,4 +10398,16 @@ mod tests { test_steps(OneOrManyWithParens::Many(vec![2]), vec![2], 3); test_steps(OneOrManyWithParens::Many(vec![3, 4]), vec![3, 4], 4); } + + // Tests that the position in the code of an `Ident` does not affect its + // ordering. + #[test] + fn test_ident_ord() { + let mut a = Ident::with_span(Span::new(Location::new(1, 1), Location::new(1, 1)), "a"); + let mut b = Ident::with_span(Span::new(Location::new(2, 2), Location::new(2, 2)), "b"); + + assert!(a < b); + std::mem::swap(&mut a.span, &mut b.span); + assert!(a < b); + } } diff --git a/src/ast/query.rs b/src/ast/query.rs index 1fb93b6c..7ffb64d9 100644 --- a/src/ast/query.rs +++ b/src/ast/query.rs @@ -321,6 +321,11 @@ pub struct Select { pub top_before_distinct: bool, /// projection expressions pub projection: Vec, + /// Excluded columns from the projection expression which are not specified + /// directly after a wildcard. + /// + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_EXCLUDE_list.html) + pub exclude: Option, /// INTO pub into: Option, /// FROM @@ -401,6 +406,10 @@ impl fmt::Display for Select { indented_list(f, &self.projection)?; } + if let Some(exclude) = &self.exclude { + write!(f, " {exclude}")?; + } + if let Some(ref into) = self.into { f.write_str(" ")?; into.fmt(f)?; @@ -1047,7 +1056,7 @@ impl fmt::Display for ConnectBy { #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))] pub struct Setting { pub key: Ident, - pub value: Value, + pub value: Expr, } impl fmt::Display for Setting { @@ -1183,7 +1192,7 @@ impl fmt::Display for TableIndexHints { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{} {} ", self.hint_type, self.index_type)?; if let Some(for_clause) = &self.for_clause { - write!(f, "FOR {} ", for_clause)?; + write!(f, "FOR {for_clause} ")?; } write!(f, "({})", display_comma_separated(&self.index_names)) } @@ -1459,7 +1468,7 @@ impl fmt::Display for TableSampleQuantity { } write!(f, "{}", self.value)?; if let Some(unit) = &self.unit { - write!(f, " {}", unit)?; + write!(f, " {unit}")?; } if self.parenthesized { write!(f, ")")?; @@ -1552,7 +1561,7 @@ impl fmt::Display for TableSampleBucket { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { write!(f, "BUCKET {} OUT OF {}", self.bucket, self.total)?; if let Some(on) = &self.on { - write!(f, " ON {}", on)?; + write!(f, " ON {on}")?; } Ok(()) } @@ -1561,19 +1570,19 @@ impl fmt::Display for TableSample { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { write!(f, "{}", self.modifier)?; if let Some(name) = &self.name { - write!(f, " {}", name)?; + write!(f, " {name}")?; } if let Some(quantity) = &self.quantity { - write!(f, " {}", quantity)?; + write!(f, " {quantity}")?; } if let Some(seed) = &self.seed { - write!(f, " {}", seed)?; + write!(f, " {seed}")?; } if let Some(bucket) = &self.bucket { - write!(f, " ({})", bucket)?; + write!(f, " ({bucket})")?; } if let Some(offset) = &self.offset { - write!(f, " OFFSET {}", offset)?; + write!(f, " OFFSET {offset}")?; } Ok(()) } @@ -1651,7 +1660,7 @@ impl fmt::Display for RowsPerMatch { RowsPerMatch::AllRows(mode) => { write!(f, "ALL ROWS PER MATCH")?; if let Some(mode) = mode { - write!(f, " {}", mode)?; + write!(f, " {mode}")?; } Ok(()) } @@ -1777,7 +1786,7 @@ impl fmt::Display for MatchRecognizePattern { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { use MatchRecognizePattern::*; match self { - Symbol(symbol) => write!(f, "{}", symbol), + Symbol(symbol) => write!(f, "{symbol}"), Exclude(symbol) => write!(f, "{{- {symbol} -}}"), Permute(symbols) => write!(f, "PERMUTE({})", display_comma_separated(symbols)), Concat(patterns) => write!(f, "{}", display_separated(patterns, " ")), @@ -2148,7 +2157,7 @@ impl fmt::Display for TableAliasColumnDef { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.name)?; if let Some(ref data_type) = self.data_type { - write!(f, " {}", data_type)?; + write!(f, " {data_type}")?; } Ok(()) } @@ -2398,7 +2407,7 @@ impl fmt::Display for OrderBy { write!(f, " {}", display_comma_separated(exprs))?; } OrderByKind::All(all) => { - write!(f, " ALL{}", all)?; + write!(f, " ALL{all}")?; } } @@ -2429,7 +2438,7 @@ impl fmt::Display for OrderByExpr { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}{}", self.expr, self.options)?; if let Some(ref with_fill) = self.with_fill { - write!(f, " {}", with_fill)? + write!(f, " {with_fill}")? } Ok(()) } @@ -2452,13 +2461,13 @@ impl fmt::Display for WithFill { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "WITH FILL")?; if let Some(ref from) = self.from { - write!(f, " FROM {}", from)?; + write!(f, " FROM {from}")?; } if let Some(ref to) = self.to { - write!(f, " TO {}", to)?; + write!(f, " TO {to}")?; } if let Some(ref step) = self.step { - write!(f, " STEP {}", step)?; + write!(f, " STEP {step}")?; } Ok(()) } @@ -2487,7 +2496,7 @@ impl fmt::Display for InterpolateExpr { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.column)?; if let Some(ref expr) = self.expr { - write!(f, " AS {}", expr)?; + write!(f, " AS {expr}")?; } Ok(()) } @@ -2565,7 +2574,7 @@ impl fmt::Display for LimitClause { Ok(()) } LimitClause::OffsetCommaLimit { offset, limit } => { - write!(f, " LIMIT {}, {}", offset, limit) + write!(f, " LIMIT {offset}, {limit}") } } } @@ -2684,6 +2693,79 @@ pub enum PipeOperator { /// Syntax: `|> TABLESAMPLE SYSTEM (10 PERCENT) /// See more at TableSample { sample: Box }, + /// Renames columns in the input table. + /// + /// Syntax: `|> RENAME old_name AS new_name, ...` + /// + /// See more at + Rename { mappings: Vec }, + /// Combines the input table with one or more tables using UNION. + /// + /// Syntax: `|> UNION [ALL|DISTINCT] (), (), ...` + /// + /// See more at + Union { + set_quantifier: SetQuantifier, + queries: Vec, + }, + /// Returns only the rows that are present in both the input table and the specified tables. + /// + /// Syntax: `|> INTERSECT [DISTINCT] (), (), ...` + /// + /// See more at + Intersect { + set_quantifier: SetQuantifier, + queries: Vec, + }, + /// Returns only the rows that are present in the input table but not in the specified tables. + /// + /// Syntax: `|> EXCEPT DISTINCT (), (), ...` + /// + /// See more at + Except { + set_quantifier: SetQuantifier, + queries: Vec, + }, + /// Calls a table function or procedure that returns a table. + /// + /// Syntax: `|> CALL function_name(args) [AS alias]` + /// + /// See more at + Call { + function: Function, + alias: Option, + }, + /// Pivots data from rows to columns. + /// + /// Syntax: `|> PIVOT(aggregate_function(column) FOR pivot_column IN (value1, value2, ...)) [AS alias]` + /// + /// See more at + Pivot { + aggregate_functions: Vec, + value_column: Vec, + value_source: PivotValueSource, + alias: Option, + }, + /// The `UNPIVOT` pipe operator transforms columns into rows. + /// + /// Syntax: + /// ```sql + /// |> UNPIVOT(value_column FOR name_column IN (column1, column2, ...)) [alias] + /// ``` + /// + /// See more at + Unpivot { + value_column: Ident, + name_column: Ident, + unpivot_columns: Vec, + alias: Option, + }, + /// Joins the input table with another table. + /// + /// Syntax: `|> [JOIN_TYPE] JOIN [alias] ON ` or `|> [JOIN_TYPE] JOIN
[alias] USING ()` + /// + /// See more at + Join(Join), } impl fmt::Display for PipeOperator { @@ -2702,12 +2784,12 @@ impl fmt::Display for PipeOperator { write!(f, "DROP {}", display_comma_separated(columns.as_slice())) } PipeOperator::As { alias } => { - write!(f, "AS {}", alias) + write!(f, "AS {alias}") } PipeOperator::Limit { expr, offset } => { - write!(f, "LIMIT {}", expr)?; + write!(f, "LIMIT {expr}")?; if let Some(offset) = offset { - write!(f, " OFFSET {}", offset)?; + write!(f, " OFFSET {offset}")?; } Ok(()) } @@ -2730,16 +2812,96 @@ impl fmt::Display for PipeOperator { } PipeOperator::Where { expr } => { - write!(f, "WHERE {}", expr) + write!(f, "WHERE {expr}") } PipeOperator::OrderBy { exprs } => { write!(f, "ORDER BY {}", display_comma_separated(exprs.as_slice())) } PipeOperator::TableSample { sample } => { - write!(f, "{}", sample) + write!(f, "{sample}") + } + PipeOperator::Rename { mappings } => { + write!(f, "RENAME {}", display_comma_separated(mappings)) + } + PipeOperator::Union { + set_quantifier, + queries, + } => Self::fmt_set_operation(f, "UNION", set_quantifier, queries), + PipeOperator::Intersect { + set_quantifier, + queries, + } => Self::fmt_set_operation(f, "INTERSECT", set_quantifier, queries), + PipeOperator::Except { + set_quantifier, + queries, + } => Self::fmt_set_operation(f, "EXCEPT", set_quantifier, queries), + PipeOperator::Call { function, alias } => { + write!(f, "CALL {function}")?; + Self::fmt_optional_alias(f, alias) + } + PipeOperator::Pivot { + aggregate_functions, + value_column, + value_source, + alias, + } => { + write!( + f, + "PIVOT({} FOR {} IN ({}))", + display_comma_separated(aggregate_functions), + Expr::CompoundIdentifier(value_column.to_vec()), + value_source + )?; + Self::fmt_optional_alias(f, alias) + } + PipeOperator::Unpivot { + value_column, + name_column, + unpivot_columns, + alias, + } => { + write!( + f, + "UNPIVOT({} FOR {} IN ({}))", + value_column, + name_column, + display_comma_separated(unpivot_columns) + )?; + Self::fmt_optional_alias(f, alias) + } + PipeOperator::Join(join) => write!(f, "{join}"), + } + } +} + +impl PipeOperator { + /// Helper function to format optional alias for pipe operators + fn fmt_optional_alias(f: &mut fmt::Formatter<'_>, alias: &Option) -> fmt::Result { + if let Some(alias) = alias { + write!(f, " AS {alias}")?; + } + Ok(()) + } + + /// Helper function to format set operations (UNION, INTERSECT, EXCEPT) with queries + fn fmt_set_operation( + f: &mut fmt::Formatter<'_>, + operation: &str, + set_quantifier: &SetQuantifier, + queries: &[Query], + ) -> fmt::Result { + write!(f, "{operation}")?; + match set_quantifier { + SetQuantifier::None => {} + _ => { + write!(f, " {set_quantifier}")?; } } + write!(f, " ")?; + let parenthesized_queries: Vec = + queries.iter().map(|query| format!("({query})")).collect(); + write!(f, "{}", display_comma_separated(&parenthesized_queries)) } } @@ -3016,7 +3178,7 @@ pub enum FormatClause { impl fmt::Display for FormatClause { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { match self { - FormatClause::Identifier(ident) => write!(f, "FORMAT {}", ident), + FormatClause::Identifier(ident) => write!(f, "FORMAT {ident}"), FormatClause::Null => write!(f, "FORMAT NULL"), } } @@ -3078,9 +3240,9 @@ impl fmt::Display for ForClause { without_array_wrapper, } => { write!(f, "FOR JSON ")?; - write!(f, "{}", for_json)?; + write!(f, "{for_json}")?; if let Some(root) = root { - write!(f, ", ROOT('{}')", root)?; + write!(f, ", ROOT('{root}')")?; } if *include_null_values { write!(f, ", INCLUDE_NULL_VALUES")?; @@ -3098,7 +3260,7 @@ impl fmt::Display for ForClause { r#type, } => { write!(f, "FOR XML ")?; - write!(f, "{}", for_xml)?; + write!(f, "{for_xml}")?; if *binary_base64 { write!(f, ", BINARY BASE64")?; } @@ -3106,7 +3268,7 @@ impl fmt::Display for ForClause { write!(f, ", TYPE")?; } if let Some(root) = root { - write!(f, ", ROOT('{}')", root)?; + write!(f, ", ROOT('{root}')")?; } if *elements { write!(f, ", ELEMENTS")?; @@ -3133,7 +3295,7 @@ impl fmt::Display for ForXml { ForXml::Raw(root) => { write!(f, "RAW")?; if let Some(root) = root { - write!(f, "('{}')", root)?; + write!(f, "('{root}')")?; } Ok(()) } @@ -3142,7 +3304,7 @@ impl fmt::Display for ForXml { ForXml::Path(root) => { write!(f, "PATH")?; if let Some(root) = root { - write!(f, "('{}')", root)?; + write!(f, "('{root}')")?; } Ok(()) } @@ -3205,7 +3367,7 @@ impl fmt::Display for JsonTableColumn { JsonTableColumn::Named(json_table_named_column) => { write!(f, "{json_table_named_column}") } - JsonTableColumn::ForOrdinality(ident) => write!(f, "{} FOR ORDINALITY", ident), + JsonTableColumn::ForOrdinality(ident) => write!(f, "{ident} FOR ORDINALITY"), JsonTableColumn::Nested(json_table_nested_column) => { write!(f, "{json_table_nested_column}") } @@ -3271,10 +3433,10 @@ impl fmt::Display for JsonTableNamedColumn { self.path )?; if let Some(on_empty) = &self.on_empty { - write!(f, " {} ON EMPTY", on_empty)?; + write!(f, " {on_empty} ON EMPTY")?; } if let Some(on_error) = &self.on_error { - write!(f, " {} ON ERROR", on_error)?; + write!(f, " {on_error} ON ERROR")?; } Ok(()) } @@ -3296,7 +3458,7 @@ impl fmt::Display for JsonTableColumnErrorHandling { match self { JsonTableColumnErrorHandling::Null => write!(f, "NULL"), JsonTableColumnErrorHandling::Default(json_string) => { - write!(f, "DEFAULT {}", json_string) + write!(f, "DEFAULT {json_string}") } JsonTableColumnErrorHandling::Error => write!(f, "ERROR"), } @@ -3429,12 +3591,12 @@ impl fmt::Display for XmlTableColumn { default, nullable, } => { - write!(f, " {}", r#type)?; + write!(f, " {type}")?; if let Some(p) = path { - write!(f, " PATH {}", p)?; + write!(f, " PATH {p}")?; } if let Some(d) = default { - write!(f, " DEFAULT {}", d)?; + write!(f, " DEFAULT {d}")?; } if !*nullable { write!(f, " NOT NULL")?; @@ -3465,7 +3627,7 @@ impl fmt::Display for XmlPassingArgument { } write!(f, "{}", self.expr)?; if let Some(alias) = &self.alias { - write!(f, " AS {}", alias)?; + write!(f, " AS {alias}")?; } Ok(()) } diff --git a/src/ast/spans.rs b/src/ast/spans.rs index 14664b4c..3e82905e 100644 --- a/src/ast/spans.rs +++ b/src/ast/spans.rs @@ -15,7 +15,7 @@ // specific language governing permissions and limitations // under the License. -use crate::ast::query::SelectItemQualifiedWildcardKind; +use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions}; use core::iter; use crate::tokenizer::Span; @@ -28,16 +28,17 @@ use super::{ ConstraintCharacteristics, CopySource, CreateIndex, CreateTable, CreateTableOptions, Cte, Delete, DoUpdate, ExceptSelectItem, ExcludeSelectItem, Expr, ExprWithAlias, Fetch, FromTable, Function, FunctionArg, FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, - FunctionArguments, GroupByExpr, HavingBound, IfStatement, IlikeSelectItem, Insert, Interpolate, - InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, LateralView, - LimitClause, MatchRecognizePattern, Measure, NamedParenthesizedList, NamedWindowDefinition, - ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, OnInsert, OpenStatement, - OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, ProjectionSelect, Query, - RaiseStatement, RaiseStatementValue, ReferentialAction, RenameSelectItem, ReplaceSelectElement, - ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SqlOption, Statement, Subscript, - SymbolDefinition, TableAlias, TableAliasColumnDef, TableConstraint, TableFactor, TableObject, - TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use, Value, Values, ViewColumnDef, - WhileStatement, WildcardAdditionalOptions, With, WithFill, + FunctionArguments, GroupByExpr, HavingBound, IfStatement, IlikeSelectItem, IndexColumn, Insert, + Interpolate, InterpolateExpr, Join, JoinConstraint, JoinOperator, JsonPath, JsonPathElem, + LateralView, LimitClause, MatchRecognizePattern, Measure, NamedParenthesizedList, + NamedWindowDefinition, ObjectName, ObjectNamePart, Offset, OnConflict, OnConflictAction, + OnInsert, OpenStatement, OrderBy, OrderByExpr, OrderByKind, Partition, PivotValueSource, + ProjectionSelect, Query, RaiseStatement, RaiseStatementValue, ReferentialAction, + RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem, + SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef, + TableConstraint, TableFactor, TableObject, TableOptionsClustered, TableWithJoins, + UpdateTableFromKind, Use, Value, Values, ViewColumnDef, WhileStatement, + WildcardAdditionalOptions, With, WithFill, }; /// Given an iterator of spans, return the [Span::union] of all spans. @@ -422,6 +423,7 @@ impl Spanned for Statement { Statement::CreateIndex(create_index) => create_index.span(), Statement::CreateRole { .. } => Span::empty(), Statement::CreateSecret { .. } => Span::empty(), + Statement::CreateServer { .. } => Span::empty(), Statement::CreateConnector { .. } => Span::empty(), Statement::AlterTable { name, @@ -650,7 +652,7 @@ impl Spanned for TableConstraint { name.iter() .map(|i| i.span) .chain(index_name.iter().map(|i| i.span)) - .chain(columns.iter().map(|i| i.span)) + .chain(columns.iter().map(|i| i.span())) .chain(characteristics.iter().map(|i| i.span())), ), TableConstraint::PrimaryKey { @@ -664,7 +666,7 @@ impl Spanned for TableConstraint { name.iter() .map(|i| i.span) .chain(index_name.iter().map(|i| i.span)) - .chain(columns.iter().map(|i| i.span)) + .chain(columns.iter().map(|i| i.span())) .chain(characteristics.iter().map(|i| i.span())), ), TableConstraint::ForeignKey { @@ -700,7 +702,7 @@ impl Spanned for TableConstraint { } => union_spans( name.iter() .map(|i| i.span) - .chain(columns.iter().map(|i| i.span)), + .chain(columns.iter().map(|i| i.span())), ), TableConstraint::FulltextOrSpatial { fulltext: _, @@ -711,7 +713,7 @@ impl Spanned for TableConstraint { opt_index_name .iter() .map(|i| i.span) - .chain(columns.iter().map(|i| i.span)), + .chain(columns.iter().map(|i| i.span())), ), } } @@ -745,6 +747,12 @@ impl Spanned for CreateIndex { } } +impl Spanned for IndexColumn { + fn span(&self) -> Span { + self.column.span() + } +} + impl Spanned for CaseStatement { fn span(&self) -> Span { let CaseStatement { @@ -917,6 +925,7 @@ impl Spanned for AlterColumnOperation { AlterColumnOperation::SetDataType { data_type: _, using, + had_set: _, } => using.as_ref().map_or(Span::empty(), |u| u.span()), AlterColumnOperation::AddGenerated { .. } => Span::empty(), } @@ -984,10 +993,13 @@ impl Spanned for ViewColumnDef { options, } = self; - union_spans( - core::iter::once(name.span) - .chain(options.iter().flat_map(|i| i.iter().map(|k| k.span()))), - ) + name.span.union_opt(&options.as_ref().map(|o| o.span())) + } +} + +impl Spanned for ColumnOptions { + fn span(&self) -> Span { + union_spans(self.as_slice().iter().map(|i| i.span())) } } @@ -1048,7 +1060,9 @@ impl Spanned for CreateTableOptions { match self { CreateTableOptions::None => Span::empty(), CreateTableOptions::With(vec) => union_spans(vec.iter().map(|i| i.span())), - CreateTableOptions::Options(vec) => union_spans(vec.iter().map(|i| i.span())), + CreateTableOptions::Options(vec) => { + union_spans(vec.as_slice().iter().map(|i| i.span())) + } CreateTableOptions::Plain(vec) => union_spans(vec.iter().map(|i| i.span())), CreateTableOptions::TableProperties(vec) => union_spans(vec.iter().map(|i| i.span())), } @@ -1062,7 +1076,10 @@ impl Spanned for CreateTableOptions { impl Spanned for AlterTableOperation { fn span(&self) -> Span { match self { - AlterTableOperation::AddConstraint(table_constraint) => table_constraint.span(), + AlterTableOperation::AddConstraint { + constraint, + not_valid: _, + } => constraint.span(), AlterTableOperation::AddColumn { column_keyword: _, if_not_exists: _, @@ -1095,10 +1112,10 @@ impl Spanned for AlterTableOperation { } => name.span, AlterTableOperation::DropColumn { has_column_keyword: _, - column_name, + column_names, if_exists: _, drop_behavior: _, - } => column_name.span, + } => union_spans(column_names.iter().map(|i| i.span)), AlterTableOperation::AttachPartition { partition } => partition.span(), AlterTableOperation::DetachPartition { partition } => partition.span(), AlterTableOperation::FreezePartition { @@ -1183,6 +1200,7 @@ impl Spanned for AlterTableOperation { AlterTableOperation::AutoIncrement { value, .. } => value.span(), AlterTableOperation::Lock { .. } => Span::empty(), AlterTableOperation::ReplicaIdentity { .. } => Span::empty(), + AlterTableOperation::ValidateConstraint { name } => name.span, } } } @@ -1398,7 +1416,6 @@ impl Spanned for AssignmentTarget { /// f.e. `IS NULL ` reports as `::span`. /// /// Missing spans: -/// - [Expr::TypedString] # missing span for data_type /// - [Expr::MatchAgainst] # MySQL specific /// - [Expr::RLike] # MySQL specific /// - [Expr::Struct] # BigQuery specific @@ -1607,6 +1624,7 @@ impl Spanned for Expr { Expr::OuterJoin(expr) => expr.span(), Expr::Prior(expr) => expr.span(), Expr::Lambda(_) => Span::empty(), + Expr::MemberOf(member_of) => member_of.value.span().union(&member_of.array.span()), } } } @@ -1653,6 +1671,10 @@ impl Spanned for ObjectNamePart { fn span(&self) -> Span { match self { ObjectNamePart::Identifier(ident) => ident.span, + ObjectNamePart::Function(func) => func + .name + .span + .union(&union_spans(func.args.iter().map(|i| i.span()))), } } } @@ -2198,6 +2220,7 @@ impl Spanned for Select { distinct: _, // todo top: _, // todo, mysql specific projection, + exclude: _, into, from, lateral_views, diff --git a/src/ast/value.rs b/src/ast/value.rs index 98616407..fdfa6a67 100644 --- a/src/ast/value.rs +++ b/src/ast/value.rs @@ -116,7 +116,6 @@ impl From for Value { derive(Visit, VisitMut), visit(with = "visit_value") )] - pub enum Value { /// Numeric literal #[cfg(not(feature = "bigdecimal"))] @@ -551,16 +550,16 @@ impl fmt::Display for EscapeUnicodeStringLiteral<'_> { write!(f, r#"\\"#)?; } x if x.is_ascii() => { - write!(f, "{}", c)?; + write!(f, "{c}")?; } _ => { let codepoint = c as u32; // if the character fits in 32 bits, we can use the \XXXX format // otherwise, we need to use the \+XXXXXX format if codepoint <= 0xFFFF { - write!(f, "\\{:04X}", codepoint)?; + write!(f, "\\{codepoint:04X}")?; } else { - write!(f, "\\+{:06X}", codepoint)?; + write!(f, "\\+{codepoint:06X}")?; } } } diff --git a/src/ast/visitor.rs b/src/ast/visitor.rs index ab4f73aa..8e0a3139 100644 --- a/src/ast/visitor.rs +++ b/src/ast/visitor.rs @@ -926,10 +926,10 @@ mod tests { #[test] fn overflow() { let cond = (0..1000) - .map(|n| format!("X = {}", n)) + .map(|n| format!("X = {n}")) .collect::>() .join(" OR "); - let sql = format!("SELECT x where {0}", cond); + let sql = format!("SELECT x where {cond}"); let dialect = GenericDialect {}; let tokens = Tokenizer::new(&dialect, sql.as_str()).tokenize().unwrap(); diff --git a/src/dialect/bigquery.rs b/src/dialect/bigquery.rs index 68ca1390..d53c9db0 100644 --- a/src/dialect/bigquery.rs +++ b/src/dialect/bigquery.rs @@ -46,7 +46,11 @@ pub struct BigQueryDialect; impl Dialect for BigQueryDialect { fn parse_statement(&self, parser: &mut Parser) -> Option> { - self.maybe_parse_statement(parser) + if parser.parse_keyword(Keyword::BEGIN) { + return Some(parser.parse_begin_exception_end()); + } + + None } /// See @@ -140,49 +144,8 @@ impl Dialect for BigQueryDialect { fn supports_pipe_operator(&self) -> bool { true } -} -impl BigQueryDialect { - fn maybe_parse_statement(&self, parser: &mut Parser) -> Option> { - if parser.peek_keyword(Keyword::BEGIN) { - return Some(self.parse_begin(parser)); - } - None - } - - /// Parse a `BEGIN` statement. - /// - fn parse_begin(&self, parser: &mut Parser) -> Result { - parser.expect_keyword(Keyword::BEGIN)?; - - let statements = parser.parse_statement_list(&[Keyword::EXCEPTION, Keyword::END])?; - - let has_exception_when_clause = parser.parse_keywords(&[ - Keyword::EXCEPTION, - Keyword::WHEN, - Keyword::ERROR, - Keyword::THEN, - ]); - let exception_statements = if has_exception_when_clause { - if !parser.peek_keyword(Keyword::END) { - Some(parser.parse_statement_list(&[Keyword::END])?) - } else { - Some(Default::default()) - } - } else { - None - }; - - parser.expect_keyword(Keyword::END)?; - - Ok(Statement::StartTransaction { - begin: true, - statements, - exception_statements, - has_end_keyword: true, - transaction: None, - modifier: None, - modes: Default::default(), - }) + fn supports_create_table_multi_schema_info_sources(&self) -> bool { + true } } diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs index 3366c670..fa18463a 100644 --- a/src/dialect/duckdb.rs +++ b/src/dialect/duckdb.rs @@ -94,4 +94,8 @@ impl Dialect for DuckDbDialect { fn supports_order_by_all(&self) -> bool { true } + + fn supports_select_wildcard_exclude(&self) -> bool { + true + } } diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs index 8f57e487..be2cc007 100644 --- a/src/dialect/generic.rs +++ b/src/dialect/generic.rs @@ -52,6 +52,10 @@ impl Dialect for GenericDialect { true } + fn supports_left_associative_joins_without_parens(&self) -> bool { + true + } + fn supports_connect_by(&self) -> bool { true } @@ -108,6 +112,14 @@ impl Dialect for GenericDialect { true } + fn supports_from_first_select(&self) -> bool { + true + } + + fn supports_projection_trailing_commas(&self) -> bool { + true + } + fn supports_asc_desc_in_column_definition(&self) -> bool { true } @@ -167,4 +179,8 @@ impl Dialect for GenericDialect { fn supports_filter_during_aggregation(&self) -> bool { true } + + fn supports_select_wildcard_exclude(&self) -> bool { + true + } } diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs index a4c899e6..c79b279d 100644 --- a/src/dialect/mod.rs +++ b/src/dialect/mod.rs @@ -49,7 +49,7 @@ pub use self::postgresql::PostgreSqlDialect; pub use self::redshift::RedshiftSqlDialect; pub use self::snowflake::SnowflakeDialect; pub use self::sqlite::SQLiteDialect; -use crate::ast::{ColumnOption, Expr, GranteesType, Statement}; +use crate::ast::{ColumnOption, Expr, GranteesType, Ident, ObjectNamePart, Statement}; pub use crate::keywords; use crate::keywords::Keyword; use crate::parser::{Parser, ParserError}; @@ -278,6 +278,34 @@ pub trait Dialect: Debug + Any { false } + /// Indicates whether the dialect supports left-associative join parsing + /// by default when parentheses are omitted in nested joins. + /// + /// Most dialects (like MySQL or Postgres) assume **left-associative** precedence, + /// so a query like: + /// + /// ```sql + /// SELECT * FROM t1 NATURAL JOIN t5 INNER JOIN t0 ON ... + /// ``` + /// is interpreted as: + /// ```sql + /// ((t1 NATURAL JOIN t5) INNER JOIN t0 ON ...) + /// ``` + /// and internally represented as a **flat list** of joins. + /// + /// In contrast, some dialects (e.g. **Snowflake**) assume **right-associative** + /// precedence and interpret the same query as: + /// ```sql + /// (t1 NATURAL JOIN (t5 INNER JOIN t0 ON ...)) + /// ``` + /// which results in a **nested join** structure in the AST. + /// + /// If this method returns `false`, the parser must build nested join trees + /// even in the absence of parentheses to reflect the correct associativity + fn supports_left_associative_joins_without_parens(&self) -> bool { + true + } + /// Returns true if the dialect supports the `(+)` syntax for OUTER JOIN. fn supports_outer_join_operator(&self) -> bool { false @@ -542,6 +570,33 @@ pub trait Dialect: Debug + Any { false } + /// Returns true if the dialect supports an exclude option + /// following a wildcard in the projection section. For example: + /// `SELECT * EXCLUDE col1 FROM tbl`. + /// + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_EXCLUDE_list.html) + /// [Snowflake](https://docs.snowflake.com/en/sql-reference/sql/select) + fn supports_select_wildcard_exclude(&self) -> bool { + false + } + + /// Returns true if the dialect supports an exclude option + /// as the last item in the projection section, not necessarily + /// after a wildcard. For example: + /// `SELECT *, c1, c2 EXCLUDE c3 FROM tbl` + /// + /// [Redshift](https://docs.aws.amazon.com/redshift/latest/dg/r_EXCLUDE_list.html) + fn supports_select_exclude(&self) -> bool { + false + } + + /// Returne true if the dialect supports specifying multiple options + /// in a `CREATE TABLE` statement for the structure of the new table. For example: + /// `CREATE TABLE t (a INT, b INT) AS SELECT 1 AS b, 2 AS a` + fn supports_create_table_multi_schema_info_sources(&self) -> bool { + false + } + /// Dialect-specific infix parser override /// /// This method is called to parse the next infix expression. @@ -587,7 +642,7 @@ pub trait Dialect: Debug + Any { } let token = parser.peek_token(); - debug!("get_next_precedence_full() {:?}", token); + debug!("get_next_precedence_full() {token:?}"); match token.token { Token::Word(w) if w.keyword == Keyword::OR => Ok(p!(Or)), Token::Word(w) if w.keyword == Keyword::AND => Ok(p!(And)), @@ -621,6 +676,7 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::REGEXP => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::MATCH => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::SIMILAR => Ok(p!(Like)), + Token::Word(w) if w.keyword == Keyword::MEMBER => Ok(p!(Like)), _ => Ok(self.prec_unknown()), }, Token::Word(w) if w.keyword == Keyword::IS => Ok(p!(Is)), @@ -633,6 +689,7 @@ pub trait Dialect: Debug + Any { Token::Word(w) if w.keyword == Keyword::REGEXP => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::MATCH => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::SIMILAR => Ok(p!(Like)), + Token::Word(w) if w.keyword == Keyword::MEMBER => Ok(p!(Like)), Token::Word(w) if w.keyword == Keyword::OPERATOR => Ok(p!(Between)), Token::Word(w) if w.keyword == Keyword::DIV => Ok(p!(MulDivModOp)), Token::Period => Ok(p!(Period)), @@ -962,11 +1019,17 @@ pub trait Dialect: Debug + Any { explicit || self.is_column_alias(kw, parser) } + /// Returns true if the specified keyword should be parsed as a table identifier. + /// See [keywords::RESERVED_FOR_TABLE_ALIAS] + fn is_table_alias(&self, kw: &Keyword, _parser: &mut Parser) -> bool { + !keywords::RESERVED_FOR_TABLE_ALIAS.contains(kw) + } + /// Returns true if the specified keyword should be parsed as a table factor alias. /// When explicit is true, the keyword is preceded by an `AS` word. Parser is provided /// to enable looking ahead if needed. - fn is_table_factor_alias(&self, explicit: bool, kw: &Keyword, _parser: &mut Parser) -> bool { - explicit || !keywords::RESERVED_FOR_TABLE_ALIAS.contains(kw) + fn is_table_factor_alias(&self, explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { + explicit || self.is_table_alias(kw, parser) } /// Returns true if this dialect supports querying historical table data @@ -1028,6 +1091,37 @@ pub trait Dialect: Debug + Any { fn supports_set_names(&self) -> bool { false } + + fn supports_space_separated_column_options(&self) -> bool { + false + } + + /// Returns true if the dialect supports the `USING` clause in an `ALTER COLUMN` statement. + /// Example: + /// ```sql + /// ALTER TABLE tbl ALTER COLUMN col SET DATA TYPE USING ` + /// ``` + fn supports_alter_column_type_using(&self) -> bool { + false + } + + /// Returns true if the dialect supports `ALTER TABLE tbl DROP COLUMN c1, ..., cn` + fn supports_comma_separated_drop_column_list(&self) -> bool { + false + } + + /// Returns true if the dialect considers the specified ident as a function + /// that returns an identifier. Typically used to generate identifiers + /// programmatically. + /// + /// - [Snowflake](https://docs.snowflake.com/en/sql-reference/identifier-literal) + fn is_identifier_generating_function_name( + &self, + _ident: &Ident, + _name_parts: &[ObjectNamePart], + ) -> bool { + false + } } /// This represents the operators for which precedence must be defined diff --git a/src/dialect/mysql.rs b/src/dialect/mysql.rs index f69e4243..f7b5f574 100644 --- a/src/dialect/mysql.rs +++ b/src/dialect/mysql.rs @@ -43,15 +43,19 @@ impl Dialect for MySqlDialect { // See https://dev.mysql.com/doc/refman/8.0/en/identifiers.html. // Identifiers which begin with a digit are recognized while tokenizing numbers, // so they can be distinguished from exponent numeric literals. + // MySQL also implements non ascii utf-8 charecters ch.is_alphabetic() || ch == '_' || ch == '$' || ch == '@' || ('\u{0080}'..='\u{ffff}').contains(&ch) + || !ch.is_ascii() } fn is_identifier_part(&self, ch: char) -> bool { - self.is_identifier_start(ch) || ch.is_ascii_digit() + self.is_identifier_start(ch) || ch.is_ascii_digit() || + // MySQL implements Unicode characters in identifiers. + !ch.is_ascii() } fn is_delimited_identifier_start(&self, ch: char) -> bool { diff --git a/src/dialect/postgresql.rs b/src/dialect/postgresql.rs index 9b08b8f3..9cea252c 100644 --- a/src/dialect/postgresql.rs +++ b/src/dialect/postgresql.rs @@ -65,14 +65,15 @@ impl Dialect for PostgreSqlDialect { } fn is_identifier_start(&self, ch: char) -> bool { - // See https://www.postgresql.org/docs/11/sql-syntax-lexical.html#SQL-SYNTAX-IDENTIFIERS - // We don't yet support identifiers beginning with "letters with - // diacritical marks" - ch.is_alphabetic() || ch == '_' + ch.is_alphabetic() || ch == '_' || + // PostgreSQL implements Unicode characters in identifiers. + !ch.is_ascii() } fn is_identifier_part(&self, ch: char) -> bool { - ch.is_alphabetic() || ch.is_ascii_digit() || ch == '$' || ch == '_' + ch.is_alphabetic() || ch.is_ascii_digit() || ch == '$' || ch == '_' || + // PostgreSQL implements Unicode characters in identifiers. + !ch.is_ascii() } fn supports_unicode_string_literal(&self) -> bool { @@ -104,7 +105,7 @@ impl Dialect for PostgreSqlDialect { fn get_next_precedence(&self, parser: &Parser) -> Option> { let token = parser.peek_token(); - debug!("get_next_precedence() {:?}", token); + debug!("get_next_precedence() {token:?}"); // we only return some custom value here when the behaviour (not merely the numeric value) differs // from the default implementation @@ -258,4 +259,8 @@ impl Dialect for PostgreSqlDialect { fn supports_set_names(&self) -> bool { true } + + fn supports_alter_column_type_using(&self) -> bool { + true + } } diff --git a/src/dialect/redshift.rs b/src/dialect/redshift.rs index feccca5d..68e025d1 100644 --- a/src/dialect/redshift.rs +++ b/src/dialect/redshift.rs @@ -80,12 +80,14 @@ impl Dialect for RedshiftSqlDialect { } fn is_identifier_start(&self, ch: char) -> bool { - // Extends Postgres dialect with sharp + // UTF-8 multibyte characters are supported in identifiers via the PostgreSqlDialect. + // https://docs.aws.amazon.com/redshift/latest/dg/r_names.html PostgreSqlDialect {}.is_identifier_start(ch) || ch == '#' } fn is_identifier_part(&self, ch: char) -> bool { - // Extends Postgres dialect with sharp + // UTF-8 multibyte characters are supported in identifiers via the PostgreSqlDialect. + // https://docs.aws.amazon.com/redshift/latest/dg/r_names.html PostgreSqlDialect {}.is_identifier_part(ch) || ch == '#' } @@ -129,4 +131,12 @@ impl Dialect for RedshiftSqlDialect { fn supports_string_literal_backslash_escape(&self) -> bool { true } + + fn supports_select_wildcard_exclude(&self) -> bool { + true + } + + fn supports_select_exclude(&self) -> bool { + true + } } diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs index ea0b94a6..baf99b84 100644 --- a/src/dialect/snowflake.rs +++ b/src/dialect/snowflake.rs @@ -23,10 +23,10 @@ use crate::ast::helpers::stmt_data_loading::{ FileStagingCommand, StageLoadSelectItem, StageLoadSelectItemKind, StageParamsObject, }; use crate::ast::{ - ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, Ident, - IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, - IdentityPropertyOrder, ObjectName, RowAccessPolicy, ShowObjects, SqlOption, Statement, - TagsColumnOption, WrappedCollection, + ColumnOption, ColumnPolicy, ColumnPolicyProperty, CopyIntoSnowflakeKind, DollarQuotedString, + Ident, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind, + IdentityPropertyOrder, ObjectName, ObjectNamePart, RowAccessPolicy, ShowObjects, SqlOption, + Statement, TagsColumnOption, WrappedCollection, }; use crate::dialect::{Dialect, Precedence}; use crate::keywords::Keyword; @@ -131,6 +131,10 @@ impl Dialect for SnowflakeDialect { } fn parse_statement(&self, parser: &mut Parser) -> Option> { + if parser.parse_keyword(Keyword::BEGIN) { + return Some(parser.parse_begin_exception_end()); + } + if parser.parse_keywords(&[Keyword::ALTER, Keyword::SESSION]) { // ALTER SESSION let set = match parser.parse_one_of_keywords(&[Keyword::SET, Keyword::UNSET]) { @@ -279,6 +283,10 @@ impl Dialect for SnowflakeDialect { true } + fn supports_left_associative_joins_without_parens(&self) -> bool { + false + } + fn is_reserved_for_identifier(&self, kw: Keyword) -> bool { // Unreserve some keywords that Snowflake accepts as identifiers // See: https://docs.snowflake.com/en/sql-reference/reserved-keywords @@ -293,27 +301,28 @@ impl Dialect for SnowflakeDialect { true } - fn is_select_item_alias(&self, explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { - explicit - || match kw { + fn is_column_alias(&self, kw: &Keyword, parser: &mut Parser) -> bool { + match kw { // The following keywords can be considered an alias as long as // they are not followed by other tokens that may change their meaning // e.g. `SELECT * EXCEPT (col1) FROM tbl` Keyword::EXCEPT - // e.g. `SELECT 1 LIMIT 5` - | Keyword::LIMIT - // e.g. `SELECT 1 OFFSET 5 ROWS` - | Keyword::OFFSET // e.g. `INSERT INTO t SELECT 1 RETURNING *` | Keyword::RETURNING if !matches!(parser.peek_token_ref().token, Token::Comma | Token::EOF) => { false } + // e.g. `SELECT 1 LIMIT 5` - not an alias + // e.g. `SELECT 1 OFFSET 5 ROWS` - not an alias + Keyword::LIMIT | Keyword::OFFSET if peek_for_limit_options(parser) => false, + // `FETCH` can be considered an alias as long as it's not followed by `FIRST`` or `NEXT` - // which would give it a different meanings, for example: `SELECT 1 FETCH FIRST 10 ROWS` - not an alias - Keyword::FETCH - if parser.peek_keyword(Keyword::FIRST) || parser.peek_keyword(Keyword::NEXT) => + // which would give it a different meanings, for example: + // `SELECT 1 FETCH FIRST 10 ROWS` - not an alias + // `SELECT 1 FETCH 10` - not an alias + Keyword::FETCH if parser.peek_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT]).is_some() + || peek_for_limit_options(parser) => { false } @@ -338,6 +347,90 @@ impl Dialect for SnowflakeDialect { } } + fn is_table_alias(&self, kw: &Keyword, parser: &mut Parser) -> bool { + match kw { + // The following keywords can be considered an alias as long as + // they are not followed by other tokens that may change their meaning + Keyword::RETURNING + | Keyword::INNER + | Keyword::USING + | Keyword::PIVOT + | Keyword::UNPIVOT + | Keyword::EXCEPT + | Keyword::MATCH_RECOGNIZE + if !matches!(parser.peek_token_ref().token, Token::SemiColon | Token::EOF) => + { + false + } + + // `LIMIT` can be considered an alias as long as it's not followed by a value. For example: + // `SELECT * FROM tbl LIMIT WHERE 1=1` - alias + // `SELECT * FROM tbl LIMIT 3` - not an alias + Keyword::LIMIT | Keyword::OFFSET if peek_for_limit_options(parser) => false, + + // `FETCH` can be considered an alias as long as it's not followed by `FIRST`` or `NEXT` + // which would give it a different meanings, for example: + // `SELECT * FROM tbl FETCH FIRST 10 ROWS` - not an alias + // `SELECT * FROM tbl FETCH 10` - not an alias + Keyword::FETCH + if parser + .peek_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT]) + .is_some() + || peek_for_limit_options(parser) => + { + false + } + + // All sorts of join-related keywords can be considered aliases unless additional + // keywords change their meaning. + Keyword::RIGHT | Keyword::LEFT | Keyword::SEMI | Keyword::ANTI + if parser + .peek_one_of_keywords(&[Keyword::JOIN, Keyword::OUTER]) + .is_some() => + { + false + } + + Keyword::GLOBAL if parser.peek_keyword(Keyword::FULL) => false, + + // Reserved keywords by the Snowflake dialect, which seem to be less strictive + // than what is listed in `keywords::RESERVED_FOR_TABLE_ALIAS`. The following + // keywords were tested with the this statement: `SELECT .* FROM tbl `. + Keyword::WITH + | Keyword::ORDER + | Keyword::SELECT + | Keyword::WHERE + | Keyword::GROUP + | Keyword::HAVING + | Keyword::LATERAL + | Keyword::UNION + | Keyword::INTERSECT + | Keyword::MINUS + | Keyword::ON + | Keyword::JOIN + | Keyword::INNER + | Keyword::CROSS + | Keyword::FULL + | Keyword::LEFT + | Keyword::RIGHT + | Keyword::NATURAL + | Keyword::USING + | Keyword::ASOF + | Keyword::MATCH_CONDITION + | Keyword::SET + | Keyword::QUALIFY + | Keyword::FOR + | Keyword::START + | Keyword::CONNECT + | Keyword::SAMPLE + | Keyword::TABLESAMPLE + | Keyword::FROM => false, + + // Any other word is considered an alias + _ => true, + } + } + /// See: fn supports_timestamp_versioning(&self) -> bool { true @@ -352,6 +445,47 @@ impl Dialect for SnowflakeDialect { fn get_reserved_keywords_for_select_item_operator(&self) -> &[Keyword] { &RESERVED_KEYWORDS_FOR_SELECT_ITEM_OPERATOR } + + fn supports_space_separated_column_options(&self) -> bool { + true + } + + fn supports_comma_separated_drop_column_list(&self) -> bool { + true + } + + fn is_identifier_generating_function_name( + &self, + ident: &Ident, + name_parts: &[ObjectNamePart], + ) -> bool { + ident.quote_style.is_none() + && ident.value.to_lowercase() == "identifier" + && !name_parts + .iter() + .any(|p| matches!(p, ObjectNamePart::Function(_))) + } + + // For example: `SELECT IDENTIFIER('alias1').* FROM tbl AS alias1` + fn supports_select_expr_star(&self) -> bool { + true + } + + fn supports_select_wildcard_exclude(&self) -> bool { + true + } +} + +// Peeks ahead to identify tokens that are expected after +// a LIMIT/FETCH keyword. +fn peek_for_limit_options(parser: &Parser) -> bool { + match &parser.peek_token_ref().token { + Token::Number(_, _) | Token::Placeholder(_) => true, + Token::SingleQuotedString(val) if val.is_empty() => true, + Token::DollarQuotedString(DollarQuotedString { value, .. }) if value.is_empty() => true, + Token::Word(w) if w.keyword == Keyword::NULL => true, + _ => false, + } } fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result { @@ -437,17 +571,14 @@ pub fn parse_create_table( Keyword::AS => { let query = parser.parse_query()?; builder = builder.query(Some(query)); - break; } Keyword::CLONE => { let clone = parser.parse_object_name(false).ok(); builder = builder.clone_clause(clone); - break; } Keyword::LIKE => { let like = parser.parse_object_name(false).ok(); builder = builder.like(like); - break; } Keyword::CLUSTER => { parser.expect_keyword_is(Keyword::BY)?; @@ -573,7 +704,7 @@ pub fn parse_create_table( builder = builder.columns(columns).constraints(constraints); } Token::EOF => { - if builder.columns.is_empty() { + if !builder.validate_schema_info() { return Err(ParserError::ParserError( "unexpected end of input".to_string(), )); @@ -582,7 +713,7 @@ pub fn parse_create_table( break; } Token::SemiColon => { - if builder.columns.is_empty() { + if !builder.validate_schema_info() { return Err(ParserError::ParserError( "unexpected end of input".to_string(), )); @@ -703,6 +834,7 @@ pub fn parse_stage_name_identifier(parser: &mut Parser) -> Result ident.push('~'), Token::Mod => ident.push('%'), Token::Div => ident.push('/'), + Token::Plus => ident.push('+'), Token::Word(w) => ident.push_str(&w.to_string()), _ => return parser.expected("stage name identifier", parser.peek_token()), } @@ -1177,7 +1309,7 @@ fn parse_column_policy_property( parser: &mut Parser, with: bool, ) -> Result { - let policy_name = parser.parse_identifier()?; + let policy_name = parser.parse_object_name(false)?; let using_columns = if parser.parse_keyword(Keyword::USING) { parser.expect_token(&Token::LParen)?; let columns = parser.parse_comma_separated(|p| p.parse_identifier())?; diff --git a/src/keywords.rs b/src/keywords.rs index f5c5e567..9e689a6d 100644 --- a/src/keywords.rs +++ b/src/keywords.rs @@ -395,6 +395,7 @@ define_keywords!( FUNCTION, FUNCTIONS, FUSION, + FUTURE, GENERAL, GENERATE, GENERATED, @@ -646,6 +647,7 @@ define_keywords!( ORDER, ORDINALITY, ORGANIZATION, + OTHER, OUT, OUTER, OUTPUT, @@ -814,6 +816,7 @@ define_keywords!( SERDE, SERDEPROPERTIES, SERIALIZABLE, + SERVER, SERVICE, SESSION, SESSION_USER, @@ -934,6 +937,8 @@ define_keywords!( TRY, TRY_CAST, TRY_CONVERT, + TSQUERY, + TSVECTOR, TUPLE, TYPE, UBIGINT, @@ -977,6 +982,7 @@ define_keywords!( UUID, VACUUM, VALID, + VALIDATE, VALIDATION_MODE, VALUE, VALUES, @@ -1012,6 +1018,7 @@ define_keywords!( WITHOUT, WITHOUT_ARRAY_WRAPPER, WORK, + WRAPPER, WRITE, XML, XMLNAMESPACES, @@ -1112,6 +1119,7 @@ pub const RESERVED_FOR_COLUMN_ALIAS: &[Keyword] = &[ Keyword::FETCH, Keyword::UNION, Keyword::EXCEPT, + Keyword::EXCLUDE, Keyword::INTERSECT, Keyword::MINUS, Keyword::CLUSTER, diff --git a/src/parser/mod.rs b/src/parser/mod.rs index 73cc3e0e..3bb91311 100644 --- a/src/parser/mod.rs +++ b/src/parser/mod.rs @@ -222,6 +222,9 @@ pub struct ParserOptions { /// Controls how literal values are unescaped. See /// [`Tokenizer::with_unescape`] for more details. pub unescape: bool, + /// Controls if the parser expects a semi-colon token + /// between statements. Default is `true`. + pub require_semicolon_stmt_delimiter: bool, } impl Default for ParserOptions { @@ -229,6 +232,7 @@ impl Default for ParserOptions { Self { trailing_commas: false, unescape: true, + require_semicolon_stmt_delimiter: true, } } } @@ -436,7 +440,7 @@ impl<'a> Parser<'a> { /// /// See example on [`Parser::new()`] for an example pub fn try_with_sql(self, sql: &str) -> Result { - debug!("Parsing sql '{}'...", sql); + debug!("Parsing sql '{sql}'..."); let tokens = Tokenizer::new(self.dialect, sql) .with_unescape(self.options.unescape) .tokenize_with_location()?; @@ -467,6 +471,10 @@ impl<'a> Parser<'a> { expecting_statement_delimiter = false; } + if !self.options.require_semicolon_stmt_delimiter { + expecting_statement_delimiter = false; + } + match self.peek_token().token { Token::EOF => break, @@ -1226,10 +1234,10 @@ impl<'a> Parser<'a> { expr = self.parse_compound_expr(expr, vec![])?; - debug!("prefix: {:?}", expr); + debug!("prefix: {expr:?}"); loop { let next_precedence = self.get_next_precedence()?; - debug!("next precedence: {:?}", next_precedence); + debug!("next precedence: {next_precedence:?}"); if precedence >= next_precedence { break; @@ -1521,7 +1529,7 @@ impl<'a> Parser<'a> { DataType::Custom(..) => parser_err!("dummy", loc), data_type => Ok(Expr::TypedString { data_type, - value: parser.parse_value()?.value, + value: parser.parse_value()?, }), } })?; @@ -1631,8 +1639,7 @@ impl<'a> Parser<'a> { Token::QuestionPipe => UnaryOperator::QuestionPipe, _ => { return Err(ParserError::ParserError(format!( - "Unexpected token in unary operator parsing: {:?}", - tok + "Unexpected token in unary operator parsing: {tok:?}" ))) } }; @@ -1709,10 +1716,9 @@ impl<'a> Parser<'a> { } fn parse_geometric_type(&mut self, kind: GeometricTypeKind) -> Result { - let value: Value = self.parse_value()?.value; Ok(Expr::TypedString { data_type: DataType::GeometricType(kind), - value, + value: self.parse_value()?, }) } @@ -2577,7 +2583,7 @@ impl<'a> Parser<'a> { trim_characters: None, }) } else if self.consume_token(&Token::Comma) - && dialect_of!(self is SnowflakeDialect | BigQueryDialect | GenericDialect) + && dialect_of!(self is DuckDbDialect | SnowflakeDialect | BigQueryDialect | GenericDialect) { let characters = self.parse_comma_separated(Parser::parse_expr)?; self.expect_token(&Token::RParen)?; @@ -2771,7 +2777,7 @@ impl<'a> Parser<'a> { if self.dialect.supports_dictionary_syntax() { self.prev_token(); // Put back the '{' - return self.parse_duckdb_struct_literal(); + return self.parse_dictionary(); } self.expected("an expression", token) @@ -3034,7 +3040,6 @@ impl<'a> Parser<'a> { where F: FnMut(&mut Parser<'a>) -> Result<(StructField, MatchedTrailingBracket), ParserError>, { - let start_token = self.peek_token(); self.expect_keyword_is(Keyword::STRUCT)?; // Nothing to do if we have no type information. @@ -3047,16 +3052,10 @@ impl<'a> Parser<'a> { let trailing_bracket = loop { let (def, trailing_bracket) = elem_parser(self)?; field_defs.push(def); - if !self.consume_token(&Token::Comma) { + // The struct field definition is finished if it occurs `>>` or comma. + if trailing_bracket.0 || !self.consume_token(&Token::Comma) { break trailing_bracket; } - - // Angle brackets are balanced so we only expect the trailing `>>` after - // we've matched all field types for the current struct. - // e.g. this is invalid syntax `STRUCT>>, INT>(NULL)` - if trailing_bracket.0 { - return parser_err!("unmatched > in STRUCT definition", start_token.span.start); - } }; Ok(( @@ -3076,6 +3075,7 @@ impl<'a> Parser<'a> { Ok(StructField { field_name: Some(field_name), field_type, + options: None, }) }); self.expect_token(&Token::RParen)?; @@ -3109,10 +3109,12 @@ impl<'a> Parser<'a> { let (field_type, trailing_bracket) = self.parse_data_type_helper()?; + let options = self.maybe_parse_options(Keyword::OPTIONS)?; Ok(( StructField { field_name, field_type, + options, }, trailing_bracket, )) @@ -3144,7 +3146,7 @@ impl<'a> Parser<'a> { Ok(fields) } - /// DuckDB specific: Parse a duckdb [dictionary] + /// DuckDB and ClickHouse specific: Parse a duckdb [dictionary] or a clickhouse [map] setting /// /// Syntax: /// @@ -3153,18 +3155,18 @@ impl<'a> Parser<'a> { /// ``` /// /// [dictionary]: https://duckdb.org/docs/sql/data_types/struct#creating-structs - fn parse_duckdb_struct_literal(&mut self) -> Result { + /// [map]: https://clickhouse.com/docs/operations/settings/settings#additional_table_filters + fn parse_dictionary(&mut self) -> Result { self.expect_token(&Token::LBrace)?; - let fields = - self.parse_comma_separated0(Self::parse_duckdb_dictionary_field, Token::RBrace)?; + let fields = self.parse_comma_separated0(Self::parse_dictionary_field, Token::RBrace)?; self.expect_token(&Token::RBrace)?; Ok(Expr::Dictionary(fields)) } - /// Parse a field for a duckdb [dictionary] + /// Parse a field for a duckdb [dictionary] or a clickhouse [map] setting /// /// Syntax /// @@ -3173,7 +3175,8 @@ impl<'a> Parser<'a> { /// ``` /// /// [dictionary]: https://duckdb.org/docs/sql/data_types/struct#creating-structs - fn parse_duckdb_dictionary_field(&mut self) -> Result { + /// [map]: https://clickhouse.com/docs/operations/settings/settings#additional_table_filters + fn parse_dictionary_field(&mut self) -> Result { let key = self.parse_identifier()?; self.expect_token(&Token::Colon)?; @@ -3483,10 +3486,18 @@ impl<'a> Parser<'a> { | BinaryOperator::LtEq | BinaryOperator::Eq | BinaryOperator::NotEq + | BinaryOperator::PGRegexMatch + | BinaryOperator::PGRegexIMatch + | BinaryOperator::PGRegexNotMatch + | BinaryOperator::PGRegexNotIMatch + | BinaryOperator::PGLikeMatch + | BinaryOperator::PGILikeMatch + | BinaryOperator::PGNotLikeMatch + | BinaryOperator::PGNotILikeMatch ) { return parser_err!( format!( - "Expected one of [=, >, <, =>, =<, !=] as comparison operator, found: {op}" + "Expected one of [=, >, <, =>, =<, !=, ~, ~*, !~, !~*, ~~, ~~*, !~~, !~~*] as comparison operator, found: {op}" ), span.start ); @@ -3613,6 +3624,19 @@ impl<'a> Parser<'a> { self.expected("IN or BETWEEN after NOT", self.peek_token()) } } + Keyword::MEMBER => { + if self.parse_keyword(Keyword::OF) { + self.expect_token(&Token::LParen)?; + let array = self.parse_expr()?; + self.expect_token(&Token::RParen)?; + Ok(Expr::MemberOf(MemberOf { + value: Box::new(expr), + array: Box::new(array), + })) + } else { + self.expected("OF after MEMBER", self.peek_token()) + } + } // Can only happen if `get_next_precedence` got out of sync with this function _ => parser_err!( format!("No infix parser for token {:?}", tok.token), @@ -3646,9 +3670,9 @@ impl<'a> Parser<'a> { } /// Parse the `ESCAPE CHAR` portion of `LIKE`, `ILIKE`, and `SIMILAR TO` - pub fn parse_escape_char(&mut self) -> Result, ParserError> { + pub fn parse_escape_char(&mut self) -> Result, ParserError> { if self.parse_keyword(Keyword::ESCAPE) { - Ok(Some(self.parse_literal_string()?)) + Ok(Some(self.parse_value()?.into())) } else { Ok(None) } @@ -3821,7 +3845,7 @@ impl<'a> Parser<'a> { }); } self.expect_token(&Token::LParen)?; - let in_op = match self.maybe_parse(|p| p.parse_query_body(p.dialect.prec_unknown()))? { + let in_op = match self.maybe_parse(|p| p.parse_query())? { Some(subquery) => Expr::InSubquery { expr: Box::new(expr), subquery, @@ -4666,6 +4690,8 @@ impl<'a> Parser<'a> { self.parse_create_procedure(or_alter) } else if self.parse_keyword(Keyword::CONNECTOR) { self.parse_create_connector() + } else if self.parse_keyword(Keyword::SERVER) { + self.parse_pg_create_server() } else { self.expected("an object type after CREATE", self.peek_token()) } @@ -5648,6 +5674,14 @@ impl<'a> Parser<'a> { } } + fn parse_analyze_format_kind(&mut self) -> Result { + if self.consume_token(&Token::Eq) { + Ok(AnalyzeFormatKind::Assignment(self.parse_analyze_format()?)) + } else { + Ok(AnalyzeFormatKind::Keyword(self.parse_analyze_format()?)) + } + } + pub fn parse_analyze_format(&mut self) -> Result { let next_token = self.next_token(); match &next_token.token { @@ -6225,6 +6259,8 @@ impl<'a> Parser<'a> { ObjectType::Stage } else if self.parse_keyword(Keyword::TYPE) { ObjectType::Type + } else if self.parse_keyword(Keyword::USER) { + ObjectType::User } else if self.parse_keyword(Keyword::FUNCTION) { return self.parse_drop_function(); } else if self.parse_keyword(Keyword::POLICY) { @@ -6243,7 +6279,7 @@ impl<'a> Parser<'a> { return self.parse_drop_extension(); } else { return self.expected( - "CONNECTOR, DATABASE, EXTENSION, FUNCTION, INDEX, POLICY, PROCEDURE, ROLE, SCHEMA, SECRET, SEQUENCE, STAGE, TABLE, TRIGGER, TYPE, VIEW, or MATERIALIZED VIEW after DROP", + "CONNECTOR, DATABASE, EXTENSION, FUNCTION, INDEX, POLICY, PROCEDURE, ROLE, SCHEMA, SECRET, SEQUENCE, STAGE, TABLE, TRIGGER, TYPE, VIEW, MATERIALIZED VIEW or USER after DROP", self.peek_token(), ); }; @@ -6872,9 +6908,7 @@ impl<'a> Parser<'a> { None }; - self.expect_token(&Token::LParen)?; - let columns = self.parse_comma_separated(Parser::parse_create_index_expr)?; - self.expect_token(&Token::RParen)?; + let columns = self.parse_parenthesized_index_column_list()?; let include = if self.parse_keyword(Keyword::INCLUDE) { self.expect_token(&Token::LParen)?; @@ -7630,9 +7664,22 @@ impl<'a> Parser<'a> { } pub fn parse_procedure_param(&mut self) -> Result { + let mode = if self.parse_keyword(Keyword::IN) { + Some(ArgMode::In) + } else if self.parse_keyword(Keyword::OUT) { + Some(ArgMode::Out) + } else if self.parse_keyword(Keyword::INOUT) { + Some(ArgMode::InOut) + } else { + None + }; let name = self.parse_identifier()?; let data_type = self.parse_data_type()?; - Ok(ProcedureParam { name, data_type }) + Ok(ProcedureParam { + name, + data_type, + mode, + }) } pub fn parse_column_def(&mut self) -> Result { @@ -7863,7 +7910,7 @@ impl<'a> Parser<'a> { } pub(crate) fn parse_tag(&mut self) -> Result { - let name = self.parse_identifier()?; + let name = self.parse_object_name(false)?; self.expect_token(&Token::Eq)?; let value = self.parse_literal_string()?; @@ -8074,7 +8121,7 @@ impl<'a> Parser<'a> { let index_name = self.parse_optional_ident()?; let index_type = self.parse_optional_using_then_index_type()?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_index_column_list()?; let index_options = self.parse_index_options()?; let characteristics = self.parse_constraint_characteristics()?; Ok(Some(TableConstraint::Unique { @@ -8096,7 +8143,7 @@ impl<'a> Parser<'a> { let index_name = self.parse_optional_ident()?; let index_type = self.parse_optional_using_then_index_type()?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_index_column_list()?; let index_options = self.parse_index_options()?; let characteristics = self.parse_constraint_characteristics()?; Ok(Some(TableConstraint::PrimaryKey { @@ -8174,7 +8221,7 @@ impl<'a> Parser<'a> { }; let index_type = self.parse_optional_using_then_index_type()?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_index_column_list()?; Ok(Some(TableConstraint::Index { display_as_key, @@ -8203,7 +8250,7 @@ impl<'a> Parser<'a> { let opt_index_name = self.parse_optional_ident()?; - let columns = self.parse_parenthesized_column_list(Mandatory, false)?; + let columns = self.parse_parenthesized_index_column_list()?; Ok(Some(TableConstraint::FulltextOrSpatial { fulltext, @@ -8470,7 +8517,11 @@ impl<'a> Parser<'a> { pub fn parse_alter_table_operation(&mut self) -> Result { let operation = if self.parse_keyword(Keyword::ADD) { if let Some(constraint) = self.parse_optional_table_constraint()? { - AlterTableOperation::AddConstraint(constraint) + let not_valid = self.parse_keywords(&[Keyword::NOT, Keyword::VALID]); + AlterTableOperation::AddConstraint { + constraint, + not_valid, + } } else if dialect_of!(self is ClickHouseDialect|GenericDialect) && self.parse_keyword(Keyword::PROJECTION) { @@ -8650,11 +8701,15 @@ impl<'a> Parser<'a> { } else { let has_column_keyword = self.parse_keyword(Keyword::COLUMN); // [ COLUMN ] let if_exists = self.parse_keywords(&[Keyword::IF, Keyword::EXISTS]); - let column_name = self.parse_identifier()?; + let column_names = if self.dialect.supports_comma_separated_drop_column_list() { + self.parse_comma_separated(Parser::parse_identifier)? + } else { + vec![self.parse_identifier()?] + }; let drop_behavior = self.parse_optional_drop_behavior(); AlterTableOperation::DropColumn { has_column_keyword, - column_name, + column_names, if_exists, drop_behavior, } @@ -8727,16 +8782,10 @@ impl<'a> Parser<'a> { } } else if self.parse_keywords(&[Keyword::DROP, Keyword::DEFAULT]) { AlterColumnOperation::DropDefault {} - } else if self.parse_keywords(&[Keyword::SET, Keyword::DATA, Keyword::TYPE]) - || (is_postgresql && self.parse_keyword(Keyword::TYPE)) - { - let data_type = self.parse_data_type()?; - let using = if is_postgresql && self.parse_keyword(Keyword::USING) { - Some(self.parse_expr()?) - } else { - None - }; - AlterColumnOperation::SetDataType { data_type, using } + } else if self.parse_keywords(&[Keyword::SET, Keyword::DATA, Keyword::TYPE]) { + self.parse_set_data_type(true)? + } else if self.parse_keyword(Keyword::TYPE) { + self.parse_set_data_type(false)? } else if self.parse_keywords(&[Keyword::ADD, Keyword::GENERATED]) { let generated_as = if self.parse_keyword(Keyword::ALWAYS) { Some(GeneratedAs::Always) @@ -8885,6 +8934,9 @@ impl<'a> Parser<'a> { }; AlterTableOperation::ReplicaIdentity { identity } + } else if self.parse_keywords(&[Keyword::VALIDATE, Keyword::CONSTRAINT]) { + let name = self.parse_identifier()?; + AlterTableOperation::ValidateConstraint { name } } else { let options: Vec = self.parse_options_with_keywords(&[Keyword::SET, Keyword::TBLPROPERTIES])?; @@ -8902,6 +8954,22 @@ impl<'a> Parser<'a> { Ok(operation) } + fn parse_set_data_type(&mut self, had_set: bool) -> Result { + let data_type = self.parse_data_type()?; + let using = if self.dialect.supports_alter_column_type_using() + && self.parse_keyword(Keyword::USING) + { + Some(self.parse_expr()?) + } else { + None + }; + Ok(AlterColumnOperation::SetDataType { + data_type, + using, + had_set, + }) + } + fn parse_part_or_partition(&mut self) -> Result { let keyword = self.expect_one_of_keywords(&[Keyword::PART, Keyword::PARTITION])?; match keyword { @@ -9913,6 +9981,12 @@ impl<'a> Parser<'a> { Ok(DataType::Unsigned) } } + Keyword::TSVECTOR if dialect_is!(dialect is PostgreSqlDialect | GenericDialect) => { + Ok(DataType::TsVector) + } + Keyword::TSQUERY if dialect_is!(dialect is PostgreSqlDialect | GenericDialect) => { + Ok(DataType::TsQuery) + } _ => { self.prev_token(); let type_name = self.parse_object_name(false)?; @@ -9975,6 +10049,48 @@ impl<'a> Parser<'a> { Ok(IdentWithAlias { ident, alias }) } + /// Parse `identifier [AS] identifier` where the AS keyword is optional + fn parse_identifier_with_optional_alias(&mut self) -> Result { + let ident = self.parse_identifier()?; + let _after_as = self.parse_keyword(Keyword::AS); + let alias = self.parse_identifier()?; + Ok(IdentWithAlias { ident, alias }) + } + + /// Parse comma-separated list of parenthesized queries for pipe operators + fn parse_pipe_operator_queries(&mut self) -> Result, ParserError> { + self.parse_comma_separated(|parser| { + parser.expect_token(&Token::LParen)?; + let query = parser.parse_query()?; + parser.expect_token(&Token::RParen)?; + Ok(*query) + }) + } + + /// Parse set quantifier for pipe operators that require DISTINCT. E.g. INTERSECT and EXCEPT + fn parse_distinct_required_set_quantifier( + &mut self, + operator_name: &str, + ) -> Result { + let quantifier = self.parse_set_quantifier(&Some(SetOperator::Intersect)); + match quantifier { + SetQuantifier::Distinct | SetQuantifier::DistinctByName => Ok(quantifier), + _ => Err(ParserError::ParserError(format!( + "{operator_name} pipe operator requires DISTINCT modifier", + ))), + } + } + + /// Parse optional identifier alias (with or without AS keyword) + fn parse_identifier_optional_alias(&mut self) -> Result, ParserError> { + if self.parse_keyword(Keyword::AS) { + Ok(Some(self.parse_identifier()?)) + } else { + // Check if the next token is an identifier (implicit alias) + self.maybe_parse(|parser| parser.parse_identifier()) + } + } + /// Optionally parses an alias for a select list item fn maybe_parse_select_item_alias(&mut self) -> Result, ParserError> { fn validator(explicit: bool, kw: &Keyword, parser: &mut Parser) -> bool { @@ -10263,49 +10379,14 @@ impl<'a> Parser<'a> { } } - /// Parse a possibly qualified, possibly quoted identifier, optionally allowing for wildcards, - /// e.g. *, *.*, `foo`.*, or "foo"."bar" - fn parse_object_name_with_wildcards( - &mut self, - in_table_clause: bool, - allow_wildcards: bool, - ) -> Result { - let mut idents = vec![]; - - if dialect_of!(self is BigQueryDialect) && in_table_clause { - loop { - let (ident, end_with_period) = self.parse_unquoted_hyphenated_identifier()?; - idents.push(ident); - if !self.consume_token(&Token::Period) && !end_with_period { - break; - } - } - } else { - loop { - let ident = if allow_wildcards && self.peek_token().token == Token::Mul { - let span = self.next_token().span; - Ident { - value: Token::Mul.to_string(), - quote_style: None, - span, - } - } else { - if self.dialect.supports_object_name_double_dot_notation() - && idents.len() == 1 - && self.consume_token(&Token::Period) - { - // Empty string here means default schema - idents.push(Ident::new("")); - } - self.parse_identifier()? - }; - idents.push(ident); - if !self.consume_token(&Token::Period) { - break; - } - } - } - Ok(ObjectName::from(idents)) + /// Parse a possibly qualified, possibly quoted identifier, e.g. + /// `foo` or `myschema."table" + /// + /// The `in_table_clause` parameter indicates whether the object name is a table in a FROM, JOIN, + /// or similar table clause. Currently, this is used only to support unquoted hyphenated identifiers + /// in this context on BigQuery. + pub fn parse_object_name(&mut self, in_table_clause: bool) -> Result { + self.parse_object_name_inner(in_table_clause, false) } /// Parse a possibly qualified, possibly quoted identifier, e.g. @@ -10314,19 +10395,76 @@ impl<'a> Parser<'a> { /// The `in_table_clause` parameter indicates whether the object name is a table in a FROM, JOIN, /// or similar table clause. Currently, this is used only to support unquoted hyphenated identifiers /// in this context on BigQuery. - pub fn parse_object_name(&mut self, in_table_clause: bool) -> Result { - let ObjectName(mut idents) = - self.parse_object_name_with_wildcards(in_table_clause, false)?; + /// + /// The `allow_wildcards` parameter indicates whether to allow for wildcards in the object name + /// e.g. *, *.*, `foo`.*, or "foo"."bar" + fn parse_object_name_inner( + &mut self, + in_table_clause: bool, + allow_wildcards: bool, + ) -> Result { + let mut parts = vec![]; + if dialect_of!(self is BigQueryDialect) && in_table_clause { + loop { + let (ident, end_with_period) = self.parse_unquoted_hyphenated_identifier()?; + parts.push(ObjectNamePart::Identifier(ident)); + if !self.consume_token(&Token::Period) && !end_with_period { + break; + } + } + } else { + loop { + if allow_wildcards && self.peek_token().token == Token::Mul { + let span = self.next_token().span; + parts.push(ObjectNamePart::Identifier(Ident { + value: Token::Mul.to_string(), + quote_style: None, + span, + })); + } else if dialect_of!(self is BigQueryDialect) && in_table_clause { + let (ident, end_with_period) = self.parse_unquoted_hyphenated_identifier()?; + parts.push(ObjectNamePart::Identifier(ident)); + if !self.consume_token(&Token::Period) && !end_with_period { + break; + } + } else if self.dialect.supports_object_name_double_dot_notation() + && parts.len() == 1 + && matches!(self.peek_token().token, Token::Period) + { + // Empty string here means default schema + parts.push(ObjectNamePart::Identifier(Ident::new(""))); + } else { + let ident = self.parse_identifier()?; + let part = if self + .dialect + .is_identifier_generating_function_name(&ident, &parts) + { + self.expect_token(&Token::LParen)?; + let args: Vec = + self.parse_comma_separated0(Self::parse_function_args, Token::RParen)?; + self.expect_token(&Token::RParen)?; + ObjectNamePart::Function(ObjectNamePartFunction { name: ident, args }) + } else { + ObjectNamePart::Identifier(ident) + }; + parts.push(part); + } + + if !self.consume_token(&Token::Period) { + break; + } + } + } // BigQuery accepts any number of quoted identifiers of a table name. // https://cloud.google.com/bigquery/docs/reference/standard-sql/lexical#quoted_identifiers if dialect_of!(self is BigQueryDialect) - && idents.iter().any(|part| { + && parts.iter().any(|part| { part.as_ident() .is_some_and(|ident| ident.value.contains('.')) }) { - idents = idents + parts = parts .into_iter() .flat_map(|part| match part.as_ident() { Some(ident) => ident @@ -10345,7 +10483,7 @@ impl<'a> Parser<'a> { .collect() } - Ok(ObjectName(idents)) + Ok(ObjectName(parts)) } /// Parse identifiers @@ -10566,17 +10704,7 @@ impl<'a> Parser<'a> { /// Parses a column definition within a view. fn parse_view_column(&mut self) -> Result { let name = self.parse_identifier()?; - let options = if (dialect_of!(self is BigQueryDialect | GenericDialect) - && self.parse_keyword(Keyword::OPTIONS)) - || (dialect_of!(self is SnowflakeDialect | GenericDialect) - && self.parse_keyword(Keyword::COMMENT)) - { - self.prev_token(); - self.parse_optional_column_option()? - .map(|option| vec![option]) - } else { - None - }; + let options = self.parse_view_column_options()?; let data_type = if dialect_of!(self is ClickHouseDialect) { Some(self.parse_data_type()?) } else { @@ -10589,6 +10717,25 @@ impl<'a> Parser<'a> { }) } + fn parse_view_column_options(&mut self) -> Result, ParserError> { + let mut options = Vec::new(); + loop { + let option = self.parse_optional_column_option()?; + if let Some(option) = option { + options.push(option); + } else { + break; + } + } + if options.is_empty() { + Ok(None) + } else if self.dialect.supports_space_separated_column_options() { + Ok(Some(ColumnOptions::SpaceSeparated(options))) + } else { + Ok(Some(ColumnOptions::CommaSeparated(options))) + } + } + /// Parses a parenthesized comma-separated list of unqualified, possibly quoted identifiers. /// For example: `(col1, "col 2", ...)` pub fn parse_parenthesized_column_list( @@ -10599,6 +10746,14 @@ impl<'a> Parser<'a> { self.parse_parenthesized_column_list_inner(optional, allow_empty, |p| p.parse_identifier()) } + /// Parses a parenthesized comma-separated list of index columns, which can be arbitrary + /// expressions with ordering information (and an opclass in some dialects). + fn parse_parenthesized_index_column_list(&mut self) -> Result, ParserError> { + self.parse_parenthesized_column_list_inner(Mandatory, false, |p| { + p.parse_create_index_expr() + }) + } + /// Parses a parenthesized comma-separated list of qualified, possibly quoted identifiers. /// For example: `(db1.sc1.tbl1.col1, db1.sc1.tbl1."col 2", ...)` pub fn parse_parenthesized_qualified_column_list( @@ -10927,7 +11082,7 @@ impl<'a> Parser<'a> { analyze = self.parse_keyword(Keyword::ANALYZE); verbose = self.parse_keyword(Keyword::VERBOSE); if self.parse_keyword(Keyword::FORMAT) { - format = Some(self.parse_analyze_format()?); + format = Some(self.parse_analyze_format_kind()?); } } @@ -11104,6 +11259,19 @@ impl<'a> Parser<'a> { Keyword::AGGREGATE, Keyword::ORDER, Keyword::TABLESAMPLE, + Keyword::RENAME, + Keyword::UNION, + Keyword::INTERSECT, + Keyword::EXCEPT, + Keyword::CALL, + Keyword::PIVOT, + Keyword::UNPIVOT, + Keyword::JOIN, + Keyword::INNER, + Keyword::LEFT, + Keyword::RIGHT, + Keyword::FULL, + Keyword::CROSS, ])?; match kw { Keyword::SELECT => { @@ -11170,6 +11338,121 @@ impl<'a> Parser<'a> { let sample = self.parse_table_sample(TableSampleModifier::TableSample)?; pipe_operators.push(PipeOperator::TableSample { sample }); } + Keyword::RENAME => { + let mappings = + self.parse_comma_separated(Parser::parse_identifier_with_optional_alias)?; + pipe_operators.push(PipeOperator::Rename { mappings }); + } + Keyword::UNION => { + let set_quantifier = self.parse_set_quantifier(&Some(SetOperator::Union)); + let queries = self.parse_pipe_operator_queries()?; + pipe_operators.push(PipeOperator::Union { + set_quantifier, + queries, + }); + } + Keyword::INTERSECT => { + let set_quantifier = + self.parse_distinct_required_set_quantifier("INTERSECT")?; + let queries = self.parse_pipe_operator_queries()?; + pipe_operators.push(PipeOperator::Intersect { + set_quantifier, + queries, + }); + } + Keyword::EXCEPT => { + let set_quantifier = self.parse_distinct_required_set_quantifier("EXCEPT")?; + let queries = self.parse_pipe_operator_queries()?; + pipe_operators.push(PipeOperator::Except { + set_quantifier, + queries, + }); + } + Keyword::CALL => { + let function_name = self.parse_object_name(false)?; + let function_expr = self.parse_function(function_name)?; + if let Expr::Function(function) = function_expr { + let alias = self.parse_identifier_optional_alias()?; + pipe_operators.push(PipeOperator::Call { function, alias }); + } else { + return Err(ParserError::ParserError( + "Expected function call after CALL".to_string(), + )); + } + } + Keyword::PIVOT => { + self.expect_token(&Token::LParen)?; + let aggregate_functions = + self.parse_comma_separated(Self::parse_aliased_function_call)?; + self.expect_keyword_is(Keyword::FOR)?; + let value_column = self.parse_period_separated(|p| p.parse_identifier())?; + self.expect_keyword_is(Keyword::IN)?; + + self.expect_token(&Token::LParen)?; + let value_source = if self.parse_keyword(Keyword::ANY) { + let order_by = if self.parse_keywords(&[Keyword::ORDER, Keyword::BY]) { + self.parse_comma_separated(Parser::parse_order_by_expr)? + } else { + vec![] + }; + PivotValueSource::Any(order_by) + } else if self.peek_sub_query() { + PivotValueSource::Subquery(self.parse_query()?) + } else { + PivotValueSource::List( + self.parse_comma_separated(Self::parse_expr_with_alias)?, + ) + }; + self.expect_token(&Token::RParen)?; + self.expect_token(&Token::RParen)?; + + let alias = self.parse_identifier_optional_alias()?; + + pipe_operators.push(PipeOperator::Pivot { + aggregate_functions, + value_column, + value_source, + alias, + }); + } + Keyword::UNPIVOT => { + self.expect_token(&Token::LParen)?; + let value_column = self.parse_identifier()?; + self.expect_keyword(Keyword::FOR)?; + let name_column = self.parse_identifier()?; + self.expect_keyword(Keyword::IN)?; + + self.expect_token(&Token::LParen)?; + let unpivot_columns = self.parse_comma_separated(Parser::parse_identifier)?; + self.expect_token(&Token::RParen)?; + + self.expect_token(&Token::RParen)?; + + let alias = self.parse_identifier_optional_alias()?; + + pipe_operators.push(PipeOperator::Unpivot { + value_column, + name_column, + unpivot_columns, + alias, + }); + } + Keyword::JOIN + | Keyword::INNER + | Keyword::LEFT + | Keyword::RIGHT + | Keyword::FULL + | Keyword::CROSS => { + self.prev_token(); + let mut joins = self.parse_joins()?; + if joins.len() != 1 { + return Err(ParserError::ParserError( + "Join pipe operator must have a single join".to_string(), + )); + } + let join = joins.swap_remove(0); + pipe_operators.push(PipeOperator::Join(join)) + } unhandled => { return Err(ParserError::ParserError(format!( "`expect_one_of_keywords` further up allowed unhandled keyword: {unhandled:?}" @@ -11187,7 +11470,7 @@ impl<'a> Parser<'a> { let key_values = self.parse_comma_separated(|p| { let key = p.parse_identifier()?; p.expect_token(&Token::Eq)?; - let value = p.parse_value()?.value; + let value = p.parse_expr()?; Ok(Setting { key, value }) })?; Some(key_values) @@ -11483,6 +11766,7 @@ impl<'a> Parser<'a> { top: None, top_before_distinct: false, projection: vec![], + exclude: None, into: None, from, lateral_views: vec![], @@ -11525,6 +11809,12 @@ impl<'a> Parser<'a> { self.parse_projection()? }; + let exclude = if self.dialect.supports_select_exclude() { + self.parse_optional_select_item_exclude()? + } else { + None + }; + let into = if self.parse_keyword(Keyword::INTO) { Some(self.parse_select_into()?) } else { @@ -11658,6 +11948,7 @@ impl<'a> Parser<'a> { top, top_before_distinct, projection, + exclude, into, from, lateral_views, @@ -12465,7 +12756,11 @@ impl<'a> Parser<'a> { }; let mut relation = self.parse_table_factor()?; - if self.peek_parens_less_nested_join() { + if !self + .dialect + .supports_left_associative_joins_without_parens() + && self.peek_parens_less_nested_join() + { let joins = self.parse_joins()?; relation = TableFactor::NestedJoin { table_with_joins: Box::new(TableWithJoins { relation, joins }), @@ -13555,6 +13850,15 @@ impl<'a> Parser<'a> { let with_grant_option = self.parse_keywords(&[Keyword::WITH, Keyword::GRANT, Keyword::OPTION]); + let current_grants = + if self.parse_keywords(&[Keyword::COPY, Keyword::CURRENT, Keyword::GRANTS]) { + Some(CurrentGrantsKind::CopyCurrentGrants) + } else if self.parse_keywords(&[Keyword::REVOKE, Keyword::CURRENT, Keyword::GRANTS]) { + Some(CurrentGrantsKind::RevokeCurrentGrants) + } else { + None + }; + let as_grantor = if self.parse_keywords(&[Keyword::AS]) { Some(self.parse_identifier()?) } else { @@ -13574,6 +13878,7 @@ impl<'a> Parser<'a> { with_grant_option, as_grantor, granted_by, + current_grants, }) } @@ -13625,7 +13930,7 @@ impl<'a> Parser<'a> { let ident = self.parse_identifier()?; if let GranteeName::ObjectName(namespace) = name { name = GranteeName::ObjectName(ObjectName::from(vec![Ident::new( - format!("{}:{}", namespace, ident), + format!("{namespace}:{ident}"), )])); }; } @@ -13662,6 +13967,82 @@ impl<'a> Parser<'a> { Some(GrantObjects::AllTablesInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[ + Keyword::ALL, + Keyword::EXTERNAL, + Keyword::TABLES, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::AllExternalTablesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::ALL, + Keyword::VIEWS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::AllViewsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::ALL, + Keyword::MATERIALIZED, + Keyword::VIEWS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::AllMaterializedViewsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::SCHEMAS, + Keyword::IN, + Keyword::DATABASE, + ]) { + Some(GrantObjects::FutureSchemasInDatabase { + databases: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::TABLES, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureTablesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::EXTERNAL, + Keyword::TABLES, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureExternalTablesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::VIEWS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureViewsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::MATERIALIZED, + Keyword::VIEWS, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureMaterializedViewsInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) } else if self.parse_keywords(&[ Keyword::ALL, Keyword::SEQUENCES, @@ -13671,26 +14052,35 @@ impl<'a> Parser<'a> { Some(GrantObjects::AllSequencesInSchema { schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, }) + } else if self.parse_keywords(&[ + Keyword::FUTURE, + Keyword::SEQUENCES, + Keyword::IN, + Keyword::SCHEMA, + ]) { + Some(GrantObjects::FutureSequencesInSchema { + schemas: self.parse_comma_separated(|p| p.parse_object_name(false))?, + }) } else if self.parse_keywords(&[Keyword::RESOURCE, Keyword::MONITOR]) { - Some(GrantObjects::ResourceMonitors(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::ResourceMonitors( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else if self.parse_keywords(&[Keyword::COMPUTE, Keyword::POOL]) { - Some(GrantObjects::ComputePools(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::ComputePools( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else if self.parse_keywords(&[Keyword::FAILOVER, Keyword::GROUP]) { - Some(GrantObjects::FailoverGroup(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::FailoverGroup( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else if self.parse_keywords(&[Keyword::REPLICATION, Keyword::GROUP]) { - Some(GrantObjects::ReplicationGroup(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::ReplicationGroup( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else if self.parse_keywords(&[Keyword::EXTERNAL, Keyword::VOLUME]) { - Some(GrantObjects::ExternalVolumes(self.parse_comma_separated( - |p| p.parse_object_name_with_wildcards(false, true), - )?)) + Some(GrantObjects::ExternalVolumes( + self.parse_comma_separated(|p| p.parse_object_name(false))?, + )) } else { let object_type = self.parse_one_of_keywords(&[ Keyword::SEQUENCE, @@ -13705,9 +14095,11 @@ impl<'a> Parser<'a> { Keyword::INTEGRATION, Keyword::USER, Keyword::CONNECTION, + Keyword::PROCEDURE, + Keyword::FUNCTION, ]); let objects = - self.parse_comma_separated(|p| p.parse_object_name_with_wildcards(false, true)); + self.parse_comma_separated(|p| p.parse_object_name_inner(false, true)); match object_type { Some(Keyword::DATABASE) => Some(GrantObjects::Databases(objects?)), Some(Keyword::SCHEMA) => Some(GrantObjects::Schemas(objects?)), @@ -13717,6 +14109,13 @@ impl<'a> Parser<'a> { Some(Keyword::VIEW) => Some(GrantObjects::Views(objects?)), Some(Keyword::USER) => Some(GrantObjects::Users(objects?)), Some(Keyword::CONNECTION) => Some(GrantObjects::Connections(objects?)), + kw @ (Some(Keyword::PROCEDURE) | Some(Keyword::FUNCTION)) => { + if let Some(name) = objects?.first() { + self.parse_grant_procedure_or_function(name, &kw)? + } else { + self.expected("procedure or function name", self.peek_token())? + } + } Some(Keyword::TABLE) | None => Some(GrantObjects::Tables(objects?)), _ => unreachable!(), } @@ -13728,6 +14127,31 @@ impl<'a> Parser<'a> { Ok((privileges, objects)) } + fn parse_grant_procedure_or_function( + &mut self, + name: &ObjectName, + kw: &Option, + ) -> Result, ParserError> { + let arg_types = if self.consume_token(&Token::LParen) { + let list = self.parse_comma_separated0(Self::parse_data_type, Token::RParen)?; + self.expect_token(&Token::RParen)?; + list + } else { + vec![] + }; + match kw { + Some(Keyword::PROCEDURE) => Ok(Some(GrantObjects::Procedure { + name: name.clone(), + arg_types, + })), + Some(Keyword::FUNCTION) => Ok(Some(GrantObjects::Function { + name: name.clone(), + arg_types, + })), + _ => self.expected("procedure or function keywords", self.peek_token())?, + } + } + pub fn parse_grant_permission(&mut self) -> Result { fn parse_columns(parser: &mut Parser) -> Result>, ParserError> { let columns = parser.parse_parenthesized_column_list(Optional, false)?; @@ -14595,7 +15019,7 @@ impl<'a> Parser<'a> { self.dialect .get_reserved_keywords_for_select_item_operator(), ) - .map(|keyword| Ident::new(format!("{:?}", keyword))); + .map(|keyword| Ident::new(format!("{keyword:?}"))); match self.parse_wildcard_expr()? { Expr::QualifiedWildcard(prefix, token) => Ok(SelectItem::QualifiedWildcard( @@ -14662,8 +15086,7 @@ impl<'a> Parser<'a> { } else { None }; - let opt_exclude = if opt_ilike.is_none() - && dialect_of!(self is GenericDialect | DuckDbDialect | SnowflakeDialect) + let opt_exclude = if opt_ilike.is_none() && self.dialect.supports_select_wildcard_exclude() { self.parse_optional_select_item_exclude()? } else { @@ -15015,7 +15438,8 @@ impl<'a> Parser<'a> { /// Parse a FETCH clause pub fn parse_fetch(&mut self) -> Result { - self.expect_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT])?; + let _ = self.parse_one_of_keywords(&[Keyword::FIRST, Keyword::NEXT]); + let (quantity, percent) = if self .parse_one_of_keywords(&[Keyword::ROW, Keyword::ROWS]) .is_some() @@ -15024,16 +15448,16 @@ impl<'a> Parser<'a> { } else { let quantity = Expr::Value(self.parse_value()?); let percent = self.parse_keyword(Keyword::PERCENT); - self.expect_one_of_keywords(&[Keyword::ROW, Keyword::ROWS])?; + let _ = self.parse_one_of_keywords(&[Keyword::ROW, Keyword::ROWS]); (Some(quantity), percent) }; + let with_ties = if self.parse_keyword(Keyword::ONLY) { false - } else if self.parse_keywords(&[Keyword::WITH, Keyword::TIES]) { - true } else { - return self.expected("one of ONLY or WITH TIES", self.peek_token()); + self.parse_keywords(&[Keyword::WITH, Keyword::TIES]) }; + Ok(Fetch { with_ties, percent, @@ -15096,7 +15520,7 @@ impl<'a> Parser<'a> { transaction: Some(BeginTransactionKind::Transaction), modifier: None, statements: vec![], - exception_statements: None, + exception: None, has_end_keyword: false, }) } @@ -15128,11 +15552,56 @@ impl<'a> Parser<'a> { transaction, modifier, statements: vec![], - exception_statements: None, + exception: None, has_end_keyword: false, }) } + pub fn parse_begin_exception_end(&mut self) -> Result { + let statements = self.parse_statement_list(&[Keyword::EXCEPTION, Keyword::END])?; + + let exception = if self.parse_keyword(Keyword::EXCEPTION) { + let mut when = Vec::new(); + + // We can have multiple `WHEN` arms so we consume all cases until `END` + while !self.peek_keyword(Keyword::END) { + self.expect_keyword(Keyword::WHEN)?; + + // Each `WHEN` case can have one or more conditions, e.g. + // WHEN EXCEPTION_1 [OR EXCEPTION_2] THEN + // So we parse identifiers until the `THEN` keyword. + let mut idents = Vec::new(); + + while !self.parse_keyword(Keyword::THEN) { + let ident = self.parse_identifier()?; + idents.push(ident); + + self.maybe_parse(|p| p.expect_keyword(Keyword::OR))?; + } + + let statements = self.parse_statement_list(&[Keyword::WHEN, Keyword::END])?; + + when.push(ExceptionWhen { idents, statements }); + } + + Some(when) + } else { + None + }; + + self.expect_keyword(Keyword::END)?; + + Ok(Statement::StartTransaction { + begin: true, + statements, + exception, + has_end_keyword: true, + transaction: None, + modifier: None, + modes: Default::default(), + }) + } + pub fn parse_end(&mut self) -> Result { let modifier = if !self.dialect.supports_end_transaction_modifier() { None @@ -15283,10 +15752,11 @@ impl<'a> Parser<'a> { let has_parentheses = self.consume_token(&Token::LParen); + let end_kws = &[Keyword::USING, Keyword::OUTPUT, Keyword::DEFAULT]; let end_token = match (has_parentheses, self.peek_token().token) { (true, _) => Token::RParen, (false, Token::EOF) => Token::EOF, - (false, Token::Word(w)) if w.keyword == Keyword::USING => Token::Word(w), + (false, Token::Word(w)) if end_kws.contains(&w.keyword) => Token::Word(w), (false, _) => Token::SemiColon, }; @@ -15308,6 +15778,10 @@ impl<'a> Parser<'a> { vec![] }; + let output = self.parse_keyword(Keyword::OUTPUT); + + let default = self.parse_keyword(Keyword::DEFAULT); + Ok(Statement::Execute { immediate: name.is_none(), name, @@ -15315,6 +15789,8 @@ impl<'a> Parser<'a> { has_parentheses, into, using, + output, + default, }) } @@ -15703,6 +16179,49 @@ impl<'a> Parser<'a> { Ok(sequence_options) } + /// Parse a `CREATE SERVER` statement. + /// + /// See [Statement::CreateServer] + pub fn parse_pg_create_server(&mut self) -> Result { + let ine = self.parse_keywords(&[Keyword::IF, Keyword::NOT, Keyword::EXISTS]); + let name = self.parse_object_name(false)?; + + let server_type = if self.parse_keyword(Keyword::TYPE) { + Some(self.parse_identifier()?) + } else { + None + }; + + let version = if self.parse_keyword(Keyword::VERSION) { + Some(self.parse_identifier()?) + } else { + None + }; + + self.expect_keywords(&[Keyword::FOREIGN, Keyword::DATA, Keyword::WRAPPER])?; + let foreign_data_wrapper = self.parse_object_name(false)?; + + let mut options = None; + if self.parse_keyword(Keyword::OPTIONS) { + self.expect_token(&Token::LParen)?; + options = Some(self.parse_comma_separated(|p| { + let key = p.parse_identifier()?; + let value = p.parse_identifier()?; + Ok(CreateServerOption { key, value }) + })?); + self.expect_token(&Token::RParen)?; + } + + Ok(Statement::CreateServer(CreateServerStatement { + name, + if_not_exists: ine, + server_type, + version, + foreign_data_wrapper, + options, + })) + } + /// The index of the first unprocessed token. pub fn index(&self) -> usize { self.index @@ -15726,6 +16245,13 @@ impl<'a> Parser<'a> { pub fn parse_create_procedure(&mut self, or_alter: bool) -> Result { let name = self.parse_object_name(false)?; let params = self.parse_optional_procedure_parameters()?; + + let language = if self.parse_keyword(Keyword::LANGUAGE) { + Some(self.parse_identifier()?) + } else { + None + }; + self.expect_keyword_is(Keyword::AS)?; let body = self.parse_conditional_statements(&[Keyword::END])?; @@ -15734,6 +16260,7 @@ impl<'a> Parser<'a> { name, or_alter, params, + language, body, }) } @@ -15832,9 +16359,9 @@ impl<'a> Parser<'a> { fn parse_parenthesized_identifiers(&mut self) -> Result, ParserError> { self.expect_token(&Token::LParen)?; - let partitions = self.parse_comma_separated(|p| p.parse_identifier())?; + let idents = self.parse_comma_separated0(|p| p.parse_identifier(), Token::RParen)?; self.expect_token(&Token::RParen)?; - Ok(partitions) + Ok(idents) } fn parse_column_position(&mut self) -> Result, ParserError> { @@ -16480,6 +17007,20 @@ mod tests { }}; } + fn mk_expected_col(name: &str) -> IndexColumn { + IndexColumn { + column: OrderByExpr { + expr: Expr::Identifier(name.into()), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + }, + operator_class: None, + } + } + let dialect = TestedDialects::new(vec![Box::new(GenericDialect {}), Box::new(MySqlDialect {})]); @@ -16490,7 +17031,7 @@ mod tests { display_as_key: false, name: None, index_type: None, - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16501,7 +17042,7 @@ mod tests { display_as_key: true, name: None, index_type: None, - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16512,7 +17053,7 @@ mod tests { display_as_key: false, name: Some(Ident::with_quote('\'', "index")), index_type: None, - columns: vec![Ident::new("c1"), Ident::new("c2")], + columns: vec![mk_expected_col("c1"), mk_expected_col("c2")], } ); @@ -16523,7 +17064,7 @@ mod tests { display_as_key: false, name: None, index_type: Some(IndexType::BTree), - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16534,7 +17075,7 @@ mod tests { display_as_key: false, name: None, index_type: Some(IndexType::Hash), - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16545,7 +17086,7 @@ mod tests { display_as_key: false, name: Some(Ident::new("idx_name")), index_type: Some(IndexType::BTree), - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); @@ -16556,7 +17097,7 @@ mod tests { display_as_key: false, name: Some(Ident::new("idx_name")), index_type: Some(IndexType::Hash), - columns: vec![Ident::new("c1")], + columns: vec![mk_expected_col("c1")], } ); } diff --git a/src/test_utils.rs b/src/test_utils.rs index 24c0ca57..654f2723 100644 --- a/src/test_utils.rs +++ b/src/test_utils.rs @@ -270,7 +270,7 @@ impl TestedDialects { tokenizer = tokenizer.with_unescape(options.unescape); } let tokens = tokenizer.tokenize().unwrap(); - assert_eq!(expected, tokens, "Tokenized differently for {:?}", dialect); + assert_eq!(expected, tokens, "Tokenized differently for {dialect:?}"); }); } } @@ -294,6 +294,11 @@ pub fn all_dialects() -> TestedDialects { ]) } +// Returns all available dialects with the specified parser options +pub fn all_dialects_with_options(options: ParserOptions) -> TestedDialects { + TestedDialects::new_with_options(all_dialects().dialects, options) +} + /// Returns all dialects matching the given predicate. pub fn all_dialects_where(predicate: F) -> TestedDialects where @@ -366,6 +371,11 @@ pub fn number(n: &str) -> Value { Value::Number(n.parse().unwrap(), false) } +/// Creates a [Value::SingleQuotedString] +pub fn single_quoted_string(s: impl Into) -> Value { + Value::SingleQuotedString(s.into()) +} + pub fn table_alias(name: impl Into) -> Option { Some(TableAlias { name: Ident::new(name), @@ -448,3 +458,52 @@ pub fn call(function: &str, args: impl IntoIterator) -> Expr { within_group: vec![], }) } + +/// Gets the first index column (mysql calls it a key part) of the first index found in a +/// [`Statement::CreateIndex`], [`Statement::CreateTable`], or [`Statement::AlterTable`]. +pub fn index_column(stmt: Statement) -> Expr { + match stmt { + Statement::CreateIndex(CreateIndex { columns, .. }) => { + columns.first().unwrap().column.expr.clone() + } + Statement::CreateTable(CreateTable { constraints, .. }) => { + match constraints.first().unwrap() { + TableConstraint::Index { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::Unique { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::PrimaryKey { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::FulltextOrSpatial { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), + } + } + Statement::AlterTable { operations, .. } => match operations.first().unwrap() { + AlterTableOperation::AddConstraint { constraint, .. } => { + match constraint { + TableConstraint::Index { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::Unique { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::PrimaryKey { columns, .. } => { + columns.first().unwrap().column.expr.clone() + } + TableConstraint::FulltextOrSpatial { + columns, + .. + } => columns.first().unwrap().column.expr.clone(), + _ => panic!("Expected an index, unique, primary, full text, or spatial constraint (foreign key does not support general key part expressions)"), + } + } + _ => panic!("Expected a constraint"), + }, + _ => panic!("Expected CREATE INDEX, ALTER TABLE, or CREATE TABLE, got: {stmt:?}"), + } +} diff --git a/src/tokenizer.rs b/src/tokenizer.rs index afe1e35c..8382a534 100644 --- a/src/tokenizer.rs +++ b/src/tokenizer.rs @@ -1751,7 +1751,7 @@ impl<'a> Tokenizer<'a> { (None, Some(tok)) => Ok(Some(tok)), (None, None) => self.tokenizer_error( chars.location(), - format!("Expected a valid binary operator after '{}'", prefix), + format!("Expected a valid binary operator after '{prefix}'"), ), } } @@ -1809,7 +1809,7 @@ impl<'a> Tokenizer<'a> { chars.next(); let mut temp = String::new(); - let end_delimiter = format!("${}$", value); + let end_delimiter = format!("${value}$"); loop { match chars.next() { @@ -2402,13 +2402,13 @@ fn take_char_from_hex_digits( location: chars.location(), })?; let digit = next_char.to_digit(16).ok_or_else(|| TokenizerError { - message: format!("Invalid hex digit in escaped unicode string: {}", next_char), + message: format!("Invalid hex digit in escaped unicode string: {next_char}"), location: chars.location(), })?; result = result * 16 + digit; } char::from_u32(result).ok_or_else(|| TokenizerError { - message: format!("Invalid unicode character: {:x}", result), + message: format!("Invalid unicode character: {result:x}"), location: chars.location(), }) } @@ -3504,7 +3504,7 @@ mod tests { } fn check_unescape(s: &str, expected: Option<&str>) { - let s = format!("'{}'", s); + let s = format!("'{s}'"); let mut state = State { peekable: s.chars().peekable(), line: 0, diff --git a/tests/sqlparser_bigquery.rs b/tests/sqlparser_bigquery.rs index 6a303577..2ba54d3e 100644 --- a/tests/sqlparser_bigquery.rs +++ b/tests/sqlparser_bigquery.rs @@ -261,10 +261,10 @@ fn parse_at_at_identifier() { #[test] fn parse_begin() { - let sql = r#"BEGIN SELECT 1; EXCEPTION WHEN ERROR THEN SELECT 2; END"#; + let sql = r#"BEGIN SELECT 1; EXCEPTION WHEN ERROR THEN SELECT 2; RAISE USING MESSAGE = FORMAT('ERR: %s', 'Bad'); END"#; let Statement::StartTransaction { statements, - exception_statements, + exception, has_end_keyword, .. } = bigquery().verified_stmt(sql) @@ -272,7 +272,10 @@ fn parse_begin() { unreachable!(); }; assert_eq!(1, statements.len()); - assert_eq!(1, exception_statements.unwrap().len()); + assert!(exception.is_some()); + + let exception = exception.unwrap(); + assert_eq!(1, exception.len()); assert!(has_end_keyword); bigquery().verified_stmt( @@ -352,14 +355,16 @@ fn parse_create_view_with_options() { ViewColumnDef { name: Ident::new("age"), data_type: None, - options: Some(vec![ColumnOption::Options(vec![SqlOption::KeyValue { - key: Ident::new("description"), - value: Expr::Value( - Value::DoubleQuotedString("field age".to_string()).with_span( - Span::new(Location::new(1, 42), Location::new(1, 52)) - ) - ), - }])]), + options: Some(ColumnOptions::CommaSeparated(vec![ColumnOption::Options( + vec![SqlOption::KeyValue { + key: Ident::new("description"), + value: Expr::Value( + Value::DoubleQuotedString("field age".to_string()).with_span( + Span::new(Location::new(1, 42), Location::new(1, 52)) + ) + ), + }] + )])), }, ], columns @@ -601,11 +606,13 @@ fn parse_nested_data_types() { field_name: Some("a".into()), field_type: DataType::Array(ArrayElemTypeDef::AngleBracket( Box::new(DataType::Int64,) - )) + )), + options: None, }, StructField { field_name: Some("b".into()), - field_type: DataType::Bytes(Some(42)) + field_type: DataType::Bytes(Some(42)), + options: None, }, ], StructBracketKind::AngleBrackets @@ -619,6 +626,7 @@ fn parse_nested_data_types() { vec![StructField { field_name: None, field_type: DataType::Int64, + options: None, }], StructBracketKind::AngleBrackets ), @@ -632,35 +640,6 @@ fn parse_nested_data_types() { } } -#[test] -fn parse_invalid_brackets() { - let sql = "SELECT STRUCT>(NULL)"; - assert_eq!( - bigquery_and_generic() - .parse_sql_statements(sql) - .unwrap_err(), - ParserError::ParserError("unmatched > in STRUCT literal".to_string()) - ); - - let sql = "SELECT STRUCT>>(NULL)"; - assert_eq!( - bigquery_and_generic() - .parse_sql_statements(sql) - .unwrap_err(), - ParserError::ParserError("Expected: (, found: >".to_string()) - ); - - let sql = "CREATE TABLE table (x STRUCT>>)"; - assert_eq!( - bigquery_and_generic() - .parse_sql_statements(sql) - .unwrap_err(), - ParserError::ParserError( - "Expected: ',' or ')' after column definition, found: >".to_string() - ) - ); -} - #[test] fn parse_tuple_struct_literal() { // tuple syntax: https://cloud.google.com/bigquery/docs/reference/standard-sql/data-types#tuple_syntax @@ -771,6 +750,7 @@ fn parse_typed_struct_syntax_bigquery() { fields: vec![StructField { field_name: None, field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -799,7 +779,8 @@ fn parse_typed_struct_syntax_bigquery() { quote_style: None, span: Span::empty(), }), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }, StructField { field_name: Some(Ident { @@ -807,7 +788,8 @@ fn parse_typed_struct_syntax_bigquery() { quote_style: None, span: Span::empty(), }), - field_type: DataType::String(None) + field_type: DataType::String(None), + options: None, }, ] }, @@ -825,17 +807,20 @@ fn parse_typed_struct_syntax_bigquery() { field_name: Some("arr".into()), field_type: DataType::Array(ArrayElemTypeDef::AngleBracket(Box::new( DataType::Float64 - ))) + ))), + options: None, }, StructField { field_name: Some("str".into()), field_type: DataType::Struct( vec![StructField { field_name: None, - field_type: DataType::Bool + field_type: DataType::Bool, + options: None, }], StructBracketKind::AngleBrackets - ) + ), + options: None, }, ] }, @@ -858,13 +843,15 @@ fn parse_typed_struct_syntax_bigquery() { field_type: DataType::Struct( Default::default(), StructBracketKind::AngleBrackets - ) + ), + options: None, }, StructField { field_name: Some("y".into()), field_type: DataType::Array(ArrayElemTypeDef::AngleBracket(Box::new( DataType::Struct(Default::default(), StructBracketKind::AngleBrackets) - ))) + ))), + options: None, }, ] }, @@ -879,7 +866,8 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::Value(Value::Boolean(true).with_empty_span())], fields: vec![StructField { field_name: None, - field_type: DataType::Bool + field_type: DataType::Bool, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -891,7 +879,8 @@ fn parse_typed_struct_syntax_bigquery() { )], fields: vec![StructField { field_name: None, - field_type: DataType::Bytes(Some(42)) + field_type: DataType::Bytes(Some(42)), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -907,7 +896,8 @@ fn parse_typed_struct_syntax_bigquery() { )], fields: vec![StructField { field_name: None, - field_type: DataType::Date + field_type: DataType::Date, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -916,11 +906,15 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Datetime(None), - value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), + span: Span::empty(), + }, }], fields: vec![StructField { field_name: None, - field_type: DataType::Datetime(None) + field_type: DataType::Datetime(None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -930,7 +924,8 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::value(number("5.0"))], fields: vec![StructField { field_name: None, - field_type: DataType::Float64 + field_type: DataType::Float64, + options: None, }] }, expr_from_projection(&select.projection[2]) @@ -940,7 +935,8 @@ fn parse_typed_struct_syntax_bigquery() { values: vec![Expr::value(number("1"))], fields: vec![StructField { field_name: None, - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[3]) @@ -962,7 +958,8 @@ fn parse_typed_struct_syntax_bigquery() { })], fields: vec![StructField { field_name: None, - field_type: DataType::Interval + field_type: DataType::Interval, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -971,13 +968,17 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::JSON, - value: Value::SingleQuotedString( - r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() - ) + value: ValueWithSpan { + value: Value::SingleQuotedString( + r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() + ), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::JSON + field_type: DataType::JSON, + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -993,7 +994,8 @@ fn parse_typed_struct_syntax_bigquery() { )], fields: vec![StructField { field_name: None, - field_type: DataType::String(Some(42)) + field_type: DataType::String(Some(42)), + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1002,11 +1004,17 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: Value::SingleQuotedString("2008-12-25 15:30:00 America/Los_Angeles".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString( + "2008-12-25 15:30:00 America/Los_Angeles".into() + ), + span: Span::empty(), + }, }], fields: vec![StructField { field_name: None, - field_type: DataType::Timestamp(None, TimezoneInfo::None) + field_type: DataType::Timestamp(None, TimezoneInfo::None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1016,11 +1024,15 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: Value::SingleQuotedString("15:30:00".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("15:30:00".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::Time(None, TimezoneInfo::None) + field_type: DataType::Time(None, TimezoneInfo::None), + options: None, }] }, expr_from_projection(&select.projection[2]) @@ -1033,11 +1045,15 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), - value: Value::SingleQuotedString("1".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::Numeric(ExactNumberInfo::None) + field_type: DataType::Numeric(ExactNumberInfo::None), + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1046,11 +1062,15 @@ fn parse_typed_struct_syntax_bigquery() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString("1".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::BigNumeric(ExactNumberInfo::None) + field_type: DataType::BigNumeric(ExactNumberInfo::None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1067,10 +1087,12 @@ fn parse_typed_struct_syntax_bigquery() { StructField { field_name: Some("key".into()), field_type: DataType::Int64, + options: None, }, StructField { field_name: Some("value".into()), field_type: DataType::Int64, + options: None, }, ] }, @@ -1092,6 +1114,7 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { fields: vec![StructField { field_name: None, field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1120,7 +1143,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { quote_style: None, span: Span::empty(), }), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }, StructField { field_name: Some(Ident { @@ -1128,7 +1152,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { quote_style: None, span: Span::empty(), }), - field_type: DataType::String(None) + field_type: DataType::String(None), + options: None, }, ] }, @@ -1151,13 +1176,15 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { field_type: DataType::Struct( Default::default(), StructBracketKind::AngleBrackets - ) + ), + options: None, }, StructField { field_name: Some("y".into()), field_type: DataType::Array(ArrayElemTypeDef::AngleBracket(Box::new( DataType::Struct(Default::default(), StructBracketKind::AngleBrackets) - ))) + ))), + options: None, }, ] }, @@ -1172,7 +1199,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::Value(Value::Boolean(true).with_empty_span())], fields: vec![StructField { field_name: None, - field_type: DataType::Bool + field_type: DataType::Bool, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1184,7 +1212,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { )], fields: vec![StructField { field_name: None, - field_type: DataType::Bytes(Some(42)) + field_type: DataType::Bytes(Some(42)), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1200,7 +1229,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { )], fields: vec![StructField { field_name: None, - field_type: DataType::Date + field_type: DataType::Date, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1209,11 +1239,15 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Datetime(None), - value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::Datetime(None) + field_type: DataType::Datetime(None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1223,7 +1257,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::value(number("5.0"))], fields: vec![StructField { field_name: None, - field_type: DataType::Float64 + field_type: DataType::Float64, + options: None, }] }, expr_from_projection(&select.projection[2]) @@ -1233,7 +1268,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { values: vec![Expr::value(number("1"))], fields: vec![StructField { field_name: None, - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[3]) @@ -1255,7 +1291,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { })], fields: vec![StructField { field_name: None, - field_type: DataType::Interval + field_type: DataType::Interval, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1264,13 +1301,17 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::JSON, - value: Value::SingleQuotedString( - r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() - ) + value: ValueWithSpan { + value: Value::SingleQuotedString( + r#"{"class" : {"students" : [{"name" : "Jane"}]}}"#.into() + ), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::JSON + field_type: DataType::JSON, + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1286,7 +1327,8 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { )], fields: vec![StructField { field_name: None, - field_type: DataType::String(Some(42)) + field_type: DataType::String(Some(42)), + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1295,11 +1337,17 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: Value::SingleQuotedString("2008-12-25 15:30:00 America/Los_Angeles".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString( + "2008-12-25 15:30:00 America/Los_Angeles".into() + ), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::Timestamp(None, TimezoneInfo::None) + field_type: DataType::Timestamp(None, TimezoneInfo::None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1309,11 +1357,15 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: Value::SingleQuotedString("15:30:00".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("15:30:00".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::Time(None, TimezoneInfo::None) + field_type: DataType::Time(None, TimezoneInfo::None), + options: None, }] }, expr_from_projection(&select.projection[2]) @@ -1326,11 +1378,15 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::Numeric(ExactNumberInfo::None), - value: Value::SingleQuotedString("1".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::Numeric(ExactNumberInfo::None) + field_type: DataType::Numeric(ExactNumberInfo::None), + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1339,11 +1395,15 @@ fn parse_typed_struct_syntax_bigquery_and_generic() { &Expr::Struct { values: vec![Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString("1".into()) + value: ValueWithSpan { + value: Value::SingleQuotedString("1".into()), + span: Span::empty(), + } }], fields: vec![StructField { field_name: None, - field_type: DataType::BigNumeric(ExactNumberInfo::None) + field_type: DataType::BigNumeric(ExactNumberInfo::None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1360,7 +1420,8 @@ fn parse_typed_struct_with_field_name_bigquery() { values: vec![Expr::value(number("5"))], fields: vec![StructField { field_name: Some(Ident::from("x")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1372,7 +1433,8 @@ fn parse_typed_struct_with_field_name_bigquery() { )], fields: vec![StructField { field_name: Some(Ident::from("y")), - field_type: DataType::String(None) + field_type: DataType::String(None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1387,11 +1449,13 @@ fn parse_typed_struct_with_field_name_bigquery() { fields: vec![ StructField { field_name: Some(Ident::from("x")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }, StructField { field_name: Some(Ident::from("y")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, } ] }, @@ -1409,7 +1473,8 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { values: vec![Expr::value(number("5"))], fields: vec![StructField { field_name: Some(Ident::from("x")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }] }, expr_from_projection(&select.projection[0]) @@ -1421,7 +1486,8 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { )], fields: vec![StructField { field_name: Some(Ident::from("y")), - field_type: DataType::String(None) + field_type: DataType::String(None), + options: None, }] }, expr_from_projection(&select.projection[1]) @@ -1436,11 +1502,13 @@ fn parse_typed_struct_with_field_name_bigquery_and_generic() { fields: vec![ StructField { field_name: Some(Ident::from("x")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, }, StructField { field_name: Some(Ident::from("y")), - field_type: DataType::Int64 + field_type: DataType::Int64, + options: None, } ] }, @@ -2365,7 +2433,10 @@ fn test_triple_quote_typed_strings() { assert_eq!( Expr::TypedString { data_type: DataType::JSON, - value: Value::TripleDoubleQuotedString(r#"{"foo":"bar's"}"#.into()) + value: ValueWithSpan { + value: Value::TripleDoubleQuotedString(r#"{"foo":"bar's"}"#.into()), + span: Span::empty(), + } }, expr ); @@ -2407,3 +2478,91 @@ fn test_any_type() { fn test_any_type_dont_break_custom_type() { bigquery_and_generic().verified_stmt("CREATE TABLE foo (x ANY)"); } + +#[test] +fn test_struct_field_options() { + bigquery().verified_stmt(concat!( + "CREATE TABLE my_table (", + "f0 STRUCT, ", + "f1 STRUCT<", + "a STRING OPTIONS(description = 'This is a string', type = 'string'), ", + "b INT64", + "> OPTIONS(description = 'This is a struct field')", + ")", + )); +} + +#[test] +fn test_struct_trailing_and_nested_bracket() { + bigquery().verified_stmt(concat!( + "CREATE TABLE my_table (", + "f0 STRING, ", + "f1 STRUCT>, ", + "f2 STRING", + ")", + )); + + // More complex nested structs + bigquery().verified_stmt(concat!( + "CREATE TABLE my_table (", + "f0 STRING, ", + "f1 STRUCT>>, ", + "f2 STRUCT>>>, ", + "f3 STRUCT>", + ")", + )); + + // Bad case with missing closing bracket + assert_eq!( + ParserError::ParserError("Expected: >, found: )".to_owned()), + bigquery() + .parse_sql_statements("CREATE TABLE my_table(f1 STRUCT after parsing data type STRUCT)".to_owned() + ), + bigquery() + .parse_sql_statements("CREATE TABLE my_table(f1 STRUCT>)") + .unwrap_err() + ); + + // Base case with redundant closing bracket in nested struct + assert_eq!( + ParserError::ParserError( + "Expected: ',' or ')' after column definition, found: >".to_owned() + ), + bigquery() + .parse_sql_statements("CREATE TABLE my_table(f1 STRUCT>>, c INT64)") + .unwrap_err() + ); + + let sql = "SELECT STRUCT>(NULL)"; + assert_eq!( + bigquery_and_generic() + .parse_sql_statements(sql) + .unwrap_err(), + ParserError::ParserError("unmatched > in STRUCT literal".to_string()) + ); + + let sql = "SELECT STRUCT>>(NULL)"; + assert_eq!( + bigquery_and_generic() + .parse_sql_statements(sql) + .unwrap_err(), + ParserError::ParserError("Expected: (, found: >".to_string()) + ); + + let sql = "CREATE TABLE table (x STRUCT>>)"; + assert_eq!( + bigquery_and_generic() + .parse_sql_statements(sql) + .unwrap_err(), + ParserError::ParserError( + "Expected: ',' or ')' after column definition, found: >".to_string() + ) + ); +} diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs index d0218b6c..9e5b6ce8 100644 --- a/tests/sqlparser_clickhouse.rs +++ b/tests/sqlparser_clickhouse.rs @@ -28,7 +28,7 @@ use test_utils::*; use sqlparser::ast::Expr::{BinaryOp, Identifier}; use sqlparser::ast::SelectItem::UnnamedExpr; use sqlparser::ast::TableFactor::Table; -use sqlparser::ast::Value::Number; +use sqlparser::ast::Value::Boolean; use sqlparser::ast::*; use sqlparser::dialect::ClickHouseDialect; use sqlparser::dialect::GenericDialect; @@ -60,6 +60,7 @@ fn parse_map_access_expr() { ), })], })], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("foos")])), @@ -224,6 +225,10 @@ fn parse_create_table() { clickhouse().verified_stmt( r#"CREATE TABLE "x" ("a" "int") ENGINE = MergeTree ORDER BY "x" AS SELECT * FROM "t" WHERE true"#, ); + clickhouse().one_statement_parses_to( + "CREATE TABLE x (a int) ENGINE = MergeTree() ORDER BY a", + "CREATE TABLE x (a INT) ENGINE = MergeTree ORDER BY a", + ); } #[test] @@ -669,11 +674,13 @@ fn parse_create_table_with_nested_data_types() { DataType::Tuple(vec![ StructField { field_name: None, - field_type: DataType::FixedString(128) + field_type: DataType::FixedString(128), + options: None, }, StructField { field_name: None, - field_type: DataType::Int128 + field_type: DataType::Int128, + options: None, } ]) ))), @@ -685,12 +692,14 @@ fn parse_create_table_with_nested_data_types() { StructField { field_name: Some("a".into()), field_type: DataType::Datetime64(9, None), + options: None, }, StructField { field_name: Some("b".into()), field_type: DataType::Array(ArrayElemTypeDef::Parenthesis( Box::new(DataType::Uuid) - )) + )), + options: None, }, ]), options: vec![], @@ -910,7 +919,7 @@ fn parse_create_view_with_fields_data_types() { }]), vec![] )), - options: None + options: None, }, ViewColumnDef { name: "f".into(), @@ -922,7 +931,7 @@ fn parse_create_view_with_fields_data_types() { }]), vec![] )), - options: None + options: None, }, ] ); @@ -961,38 +970,103 @@ fn parse_limit_by() { #[test] fn parse_settings_in_query() { - match clickhouse_and_generic() - .verified_stmt(r#"SELECT * FROM t SETTINGS max_threads = 1, max_block_size = 10000"#) - { - Statement::Query(query) => { - assert_eq!( - query.settings, - Some(vec![ - Setting { - key: Ident::new("max_threads"), - value: Number("1".parse().unwrap(), false) - }, - Setting { - key: Ident::new("max_block_size"), - value: Number("10000".parse().unwrap(), false) - }, - ]) - ); + fn check_settings(sql: &str, expected: Vec) { + match clickhouse_and_generic().verified_stmt(sql) { + Statement::Query(q) => { + assert_eq!(q.settings, Some(expected)); + } + _ => unreachable!(), } - _ => unreachable!(), + } + + for (sql, expected_settings) in [ + ( + r#"SELECT * FROM t SETTINGS max_threads = 1, max_block_size = 10000"#, + vec![ + Setting { + key: Ident::new("max_threads"), + value: Expr::value(number("1")), + }, + Setting { + key: Ident::new("max_block_size"), + value: Expr::value(number("10000")), + }, + ], + ), + ( + r#"SELECT * FROM t SETTINGS additional_table_filters = {'table_1': 'x != 2'}"#, + vec![Setting { + key: Ident::new("additional_table_filters"), + value: Expr::Dictionary(vec![DictionaryField { + key: Ident::with_quote('\'', "table_1"), + value: Expr::value(single_quoted_string("x != 2")).into(), + }]), + }], + ), + ( + r#"SELECT * FROM t SETTINGS additional_result_filter = 'x != 2', query_plan_optimize_lazy_materialization = false"#, + vec![ + Setting { + key: Ident::new("additional_result_filter"), + value: Expr::value(single_quoted_string("x != 2")), + }, + Setting { + key: Ident::new("query_plan_optimize_lazy_materialization"), + value: Expr::value(Boolean(false)), + }, + ], + ), + ] { + check_settings(sql, expected_settings); } let invalid_cases = vec![ - "SELECT * FROM t SETTINGS a", - "SELECT * FROM t SETTINGS a=", - "SELECT * FROM t SETTINGS a=1, b", - "SELECT * FROM t SETTINGS a=1, b=", - "SELECT * FROM t SETTINGS a=1, b=c", + ("SELECT * FROM t SETTINGS a", "Expected: =, found: EOF"), + ( + "SELECT * FROM t SETTINGS a=", + "Expected: an expression, found: EOF", + ), + ("SELECT * FROM t SETTINGS a=1, b", "Expected: =, found: EOF"), + ( + "SELECT * FROM t SETTINGS a=1, b=", + "Expected: an expression, found: EOF", + ), + ( + "SELECT * FROM t SETTINGS a = {", + "Expected: identifier, found: EOF", + ), + ( + "SELECT * FROM t SETTINGS a = {'b'", + "Expected: :, found: EOF", + ), + ( + "SELECT * FROM t SETTINGS a = {'b': ", + "Expected: an expression, found: EOF", + ), + ( + "SELECT * FROM t SETTINGS a = {'b': 'c',}", + "Expected: identifier, found: }", + ), + ( + "SELECT * FROM t SETTINGS a = {'b': 'c', 'd'}", + "Expected: :, found: }", + ), + ( + "SELECT * FROM t SETTINGS a = {'b': 'c', 'd': }", + "Expected: an expression, found: }", + ), + ( + "SELECT * FROM t SETTINGS a = {ANY(b)}", + "Expected: :, found: (", + ), ]; - for sql in invalid_cases { - clickhouse_and_generic() - .parse_sql_statements(sql) - .expect_err("Expected: SETTINGS key = value, found: "); + for (sql, error_msg) in invalid_cases { + assert_eq!( + clickhouse_and_generic() + .parse_sql_statements(sql) + .unwrap_err(), + ParserError(error_msg.to_string()) + ); } } #[test] @@ -1341,7 +1415,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - clickhouse().verified_stmt(&format!("USE {}", object_name)), + clickhouse().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -1349,7 +1423,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - clickhouse().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), + clickhouse().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -1363,7 +1437,7 @@ fn parse_use() { fn test_query_with_format_clause() { let format_options = vec!["TabSeparated", "JSONCompact", "NULL"]; for format in &format_options { - let sql = format!("SELECT * FROM t FORMAT {}", format); + let sql = format!("SELECT * FROM t FORMAT {format}"); match clickhouse_and_generic().verified_stmt(&sql) { Statement::Query(query) => { if *format == "NULL" { @@ -1546,11 +1620,11 @@ fn parse_select_table_function_settings() { settings: Some(vec![ Setting { key: "s0".into(), - value: Value::Number("3".parse().unwrap(), false), + value: Expr::value(number("3")), }, Setting { key: "s1".into(), - value: Value::SingleQuotedString("s".into()), + value: Expr::value(single_quoted_string("s")), }, ]), }, @@ -1571,11 +1645,11 @@ fn parse_select_table_function_settings() { settings: Some(vec![ Setting { key: "s0".into(), - value: Value::Number("3".parse().unwrap(), false), + value: Expr::value(number("3")), }, Setting { key: "s1".into(), - value: Value::SingleQuotedString("s".into()), + value: Expr::value(single_quoted_string("s")), }, ]), }, @@ -1585,7 +1659,6 @@ fn parse_select_table_function_settings() { "SELECT * FROM t(SETTINGS a=)", "SELECT * FROM t(SETTINGS a=1, b)", "SELECT * FROM t(SETTINGS a=1, b=)", - "SELECT * FROM t(SETTINGS a=1, b=c)", ]; for sql in invalid_cases { clickhouse_and_generic() diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs index abcadb45..4183c553 100644 --- a/tests/sqlparser_common.rs +++ b/tests/sqlparser_common.rs @@ -40,8 +40,9 @@ use sqlparser::parser::{Parser, ParserError, ParserOptions}; use sqlparser::tokenizer::Tokenizer; use sqlparser::tokenizer::{Location, Span}; use test_utils::{ - all_dialects, all_dialects_where, alter_table_op, assert_eq_vec, call, expr_from_projection, - join, number, only, table, table_alias, table_from_name, TestedDialects, + all_dialects, all_dialects_where, all_dialects_with_options, alter_table_op, assert_eq_vec, + call, expr_from_projection, join, number, only, table, table_alias, table_from_name, + TestedDialects, }; #[macro_use] @@ -459,6 +460,7 @@ fn parse_update_set_from() { SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("name"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("id"))), ], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("t1")])), @@ -1232,7 +1234,6 @@ fn parse_select_expr_star() { "SELECT 2. * 3 FROM T", ); dialects.verified_only_select("SELECT myfunc().* FROM T"); - dialects.verified_only_select("SELECT myfunc().* EXCEPT (foo) FROM T"); // Invalid let res = dialects.parse_sql_statements("SELECT foo.*.* FROM T"); @@ -1240,6 +1241,11 @@ fn parse_select_expr_star() { ParserError::ParserError("Expected: end of statement, found: .".to_string()), res.unwrap_err() ); + + let dialects = all_dialects_where(|d| { + d.supports_select_expr_star() && d.supports_select_wildcard_except() + }); + dialects.verified_only_select("SELECT myfunc().* EXCEPT (foo) FROM T"); } #[test] @@ -2040,7 +2046,7 @@ fn parse_ilike() { pattern: Box::new(Expr::Value( (Value::SingleQuotedString("%a".to_string())).with_empty_span() )), - escape_char: Some('^'.to_string()), + escape_char: Some(Value::SingleQuotedString('^'.to_string())), any: false, }, select.selection.unwrap() @@ -2104,7 +2110,7 @@ fn parse_like() { pattern: Box::new(Expr::Value( (Value::SingleQuotedString("%a".to_string())).with_empty_span() )), - escape_char: Some('^'.to_string()), + escape_char: Some(Value::SingleQuotedString('^'.to_string())), any: false, }, select.selection.unwrap() @@ -2167,7 +2173,24 @@ fn parse_similar_to() { pattern: Box::new(Expr::Value( (Value::SingleQuotedString("%a".to_string())).with_empty_span() )), - escape_char: Some('^'.to_string()), + escape_char: Some(Value::SingleQuotedString('^'.to_string())), + }, + select.selection.unwrap() + ); + + let sql = &format!( + "SELECT * FROM customers WHERE name {}SIMILAR TO '%a' ESCAPE NULL", + if negated { "NOT " } else { "" } + ); + let select = verified_only_select(sql); + assert_eq!( + Expr::SimilarTo { + expr: Box::new(Expr::Identifier(Ident::new("name"))), + negated, + pattern: Box::new(Expr::Value( + (Value::SingleQuotedString("%a".to_string())).with_empty_span() + )), + escape_char: Some(Value::Null), }, select.selection.unwrap() ); @@ -2185,7 +2208,7 @@ fn parse_similar_to() { pattern: Box::new(Expr::Value( (Value::SingleQuotedString("%a".to_string())).with_empty_span() )), - escape_char: Some('^'.to_string()), + escape_char: Some(Value::SingleQuotedString('^'.to_string())), })), select.selection.unwrap() ); @@ -2225,7 +2248,7 @@ fn parse_in_subquery() { assert_eq!( Expr::InSubquery { expr: Box::new(Expr::Identifier(Ident::new("segment"))), - subquery: verified_query("SELECT segm FROM bar").body, + subquery: Box::new(verified_query("SELECT segm FROM bar")), negated: false, }, select.selection.unwrap() @@ -2239,7 +2262,9 @@ fn parse_in_union() { assert_eq!( Expr::InSubquery { expr: Box::new(Expr::Identifier(Ident::new("segment"))), - subquery: verified_query("(SELECT segm FROM bar) UNION (SELECT segm FROM bar2)").body, + subquery: Box::new(verified_query( + "(SELECT segm FROM bar) UNION (SELECT segm FROM bar2)" + )), negated: false, }, select.selection.unwrap() @@ -3561,7 +3586,7 @@ fn test_double_value() { for (input, expected) in test_cases { for (i, expr) in input.iter().enumerate() { if let Statement::Query(query) = - dialects.one_statement_parses_to(&format!("SELECT {}", expr), "") + dialects.one_statement_parses_to(&format!("SELECT {expr}"), "") { if let SetExpr::Select(select) = *query.body { assert_eq!(expected[i], select.projection[0]); @@ -4021,13 +4046,13 @@ fn parse_create_table_column_constraint_characteristics() { syntax }; - let sql = format!("CREATE TABLE t (a int UNIQUE {})", syntax); + let sql = format!("CREATE TABLE t (a int UNIQUE {syntax})"); let expected_clause = if syntax.is_empty() { String::new() } else { format!(" {syntax}") }; - let expected = format!("CREATE TABLE t (a INT UNIQUE{})", expected_clause); + let expected = format!("CREATE TABLE t (a INT UNIQUE{expected_clause})"); let ast = one_statement_parses_to(&sql, &expected); let expected_value = if deferrable.is_some() || initially.is_some() || enforced.is_some() { @@ -4322,8 +4347,9 @@ fn parse_create_table_as() { // BigQuery allows specifying table schema in CTAS // ANSI SQL and PostgreSQL let you only specify the list of columns // (without data types) in a CTAS, but we have yet to support that. + let dialects = all_dialects_where(|d| d.supports_create_table_multi_schema_info_sources()); let sql = "CREATE TABLE t (a INT, b INT) AS SELECT 1 AS b, 2 AS a"; - match verified_stmt(sql) { + match dialects.verified_stmt(sql) { Statement::CreateTable(CreateTable { columns, query, .. }) => { assert_eq!(columns.len(), 2); assert_eq!(columns[0].to_string(), "a INT".to_string()); @@ -4428,20 +4454,6 @@ fn parse_create_or_replace_table() { } _ => unreachable!(), } - - let sql = "CREATE TABLE t (a INT, b INT) AS SELECT 1 AS b, 2 AS a"; - match verified_stmt(sql) { - Statement::CreateTable(CreateTable { columns, query, .. }) => { - assert_eq!(columns.len(), 2); - assert_eq!(columns[0].to_string(), "a INT".to_string()); - assert_eq!(columns[1].to_string(), "b INT".to_string()); - assert_eq!( - query, - Some(Box::new(verified_query("SELECT 1 AS b, 2 AS a"))) - ); - } - _ => unreachable!(), - } } #[test] @@ -4954,7 +4966,7 @@ fn parse_alter_table_constraints() { match alter_table_op(verified_stmt(&format!( "ALTER TABLE tab ADD {constraint_text}" ))) { - AlterTableOperation::AddConstraint(constraint) => { + AlterTableOperation::AddConstraint { constraint, .. } => { assert_eq!(constraint_text, constraint.to_string()); } _ => unreachable!(), @@ -4977,15 +4989,18 @@ fn parse_alter_table_drop_column() { "ALTER TABLE tab DROP is_active CASCADE", ); + let dialects = all_dialects_where(|d| d.supports_comma_separated_drop_column_list()); + dialects.verified_stmt("ALTER TABLE tbl DROP COLUMN c1, c2, c3"); + fn check_one(constraint_text: &str) { match alter_table_op(verified_stmt(&format!("ALTER TABLE tab {constraint_text}"))) { AlterTableOperation::DropColumn { has_column_keyword: true, - column_name, + column_names, if_exists, drop_behavior, } => { - assert_eq!("is_active", column_name.to_string()); + assert_eq!("is_active", column_names.first().unwrap().to_string()); assert!(if_exists); match drop_behavior { None => assert!(constraint_text.ends_with(" is_active")), @@ -5055,22 +5070,21 @@ fn parse_alter_table_alter_column_type() { AlterColumnOperation::SetDataType { data_type: DataType::Text, using: None, + had_set: true, } ); } _ => unreachable!(), } + verified_stmt(&format!("{alter_stmt} ALTER COLUMN is_active TYPE TEXT")); - let dialect = TestedDialects::new(vec![Box::new(GenericDialect {})]); + let dialects = all_dialects_where(|d| d.supports_alter_column_type_using()); + dialects.verified_stmt(&format!( + "{alter_stmt} ALTER COLUMN is_active SET DATA TYPE TEXT USING 'text'" + )); - let res = - dialect.parse_sql_statements(&format!("{alter_stmt} ALTER COLUMN is_active TYPE TEXT")); - assert_eq!( - ParserError::ParserError("Expected: SET/DROP NOT NULL, SET DEFAULT, or SET DATA TYPE after ALTER COLUMN, found: TYPE".to_string()), - res.unwrap_err() - ); - - let res = dialect.parse_sql_statements(&format!( + let dialects = all_dialects_except(|d| d.supports_alter_column_type_using()); + let res = dialects.parse_sql_statements(&format!( "{alter_stmt} ALTER COLUMN is_active SET DATA TYPE TEXT USING 'text'" )); assert_eq!( @@ -5147,7 +5161,7 @@ fn run_explain_analyze( query: &str, expected_verbose: bool, expected_analyze: bool, - expected_format: Option, + expected_format: Option, expected_options: Option>, ) { match dialect.verified_stmt(query) { @@ -5258,7 +5272,7 @@ fn parse_explain_analyze_with_simple_select() { "EXPLAIN ANALYZE FORMAT GRAPHVIZ SELECT sqrt(id) FROM foo", false, true, - Some(AnalyzeFormat::GRAPHVIZ), + Some(AnalyzeFormatKind::Keyword(AnalyzeFormat::GRAPHVIZ)), None, ); @@ -5267,7 +5281,16 @@ fn parse_explain_analyze_with_simple_select() { "EXPLAIN ANALYZE VERBOSE FORMAT JSON SELECT sqrt(id) FROM foo", true, true, - Some(AnalyzeFormat::JSON), + Some(AnalyzeFormatKind::Keyword(AnalyzeFormat::JSON)), + None, + ); + + run_explain_analyze( + all_dialects(), + "EXPLAIN ANALYZE VERBOSE FORMAT=JSON SELECT sqrt(id) FROM foo", + true, + true, + Some(AnalyzeFormatKind::Assignment(AnalyzeFormat::JSON)), None, ); @@ -5276,7 +5299,7 @@ fn parse_explain_analyze_with_simple_select() { "EXPLAIN VERBOSE FORMAT TEXT SELECT sqrt(id) FROM foo", true, false, - Some(AnalyzeFormat::TEXT), + Some(AnalyzeFormatKind::Keyword(AnalyzeFormat::TEXT)), None, ); } @@ -5523,7 +5546,8 @@ fn parse_named_window_functions() { WINDOW w AS (PARTITION BY x), win AS (ORDER BY y)"; supported_dialects.verified_stmt(sql); - let select = verified_only_select(sql); + let select = all_dialects_except(|d| d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d))) + .verified_only_select(sql); const EXPECTED_PROJ_QTY: usize = 2; assert_eq!(EXPECTED_PROJ_QTY, select.projection.len()); @@ -5553,6 +5577,7 @@ fn parse_named_window_functions() { #[test] fn parse_window_clause() { + let dialects = all_dialects_except(|d| d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d))); let sql = "SELECT * \ FROM mytable \ WINDOW \ @@ -5565,10 +5590,14 @@ fn parse_window_clause() { window7 AS (window1 ROWS UNBOUNDED PRECEDING), \ window8 AS (window1 PARTITION BY a ORDER BY b ROWS UNBOUNDED PRECEDING) \ ORDER BY C3"; - verified_only_select(sql); + dialects.verified_only_select(sql); let sql = "SELECT * from mytable WINDOW window1 AS window2"; - let dialects = all_dialects_except(|d| d.is::() || d.is::()); + let dialects = all_dialects_except(|d| { + d.is::() + || d.is::() + || d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d)) + }); let res = dialects.parse_sql_statements(sql); assert_eq!( ParserError::ParserError("Expected: (, found: window2".to_string()), @@ -5578,6 +5607,7 @@ fn parse_window_clause() { #[test] fn test_parse_named_window() { + let dialects = all_dialects_except(|d| d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d))); let sql = "SELECT \ MIN(c12) OVER window1 AS min1, \ MAX(c12) OVER window2 AS max1 \ @@ -5585,7 +5615,7 @@ fn test_parse_named_window() { WINDOW window1 AS (ORDER BY C12), \ window2 AS (PARTITION BY C11) \ ORDER BY C3"; - let actual_select_only = verified_only_select(sql); + let actual_select_only = dialects.verified_only_select(sql); let expected = Select { select_token: AttachedToken::empty(), distinct: None, @@ -5663,6 +5693,7 @@ fn test_parse_named_window() { }, }, ], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident { @@ -5734,6 +5765,10 @@ fn test_parse_named_window() { #[test] fn parse_window_and_qualify_clause() { + let dialects = all_dialects_except(|d| { + d.is_table_alias(&Keyword::WINDOW, &mut Parser::new(d)) + || d.is_table_alias(&Keyword::QUALIFY, &mut Parser::new(d)) + }); let sql = "SELECT \ MIN(c12) OVER window1 AS min1 \ FROM aggregate_test_100 \ @@ -5741,7 +5776,7 @@ fn parse_window_and_qualify_clause() { WINDOW window1 AS (ORDER BY C12), \ window2 AS (PARTITION BY C11) \ ORDER BY C3"; - verified_only_select(sql); + dialects.verified_only_select(sql); let sql = "SELECT \ MIN(c12) OVER window1 AS min1 \ @@ -5750,7 +5785,7 @@ fn parse_window_and_qualify_clause() { window2 AS (PARTITION BY C11) \ QUALIFY ROW_NUMBER() OVER my_window \ ORDER BY C3"; - verified_only_select(sql); + dialects.verified_only_select(sql); } #[test] @@ -5850,7 +5885,10 @@ fn parse_literal_date() { assert_eq!( &Expr::TypedString { data_type: DataType::Date, - value: Value::SingleQuotedString("1999-01-01".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01".into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -5863,7 +5901,10 @@ fn parse_literal_time() { assert_eq!( &Expr::TypedString { data_type: DataType::Time(None, TimezoneInfo::None), - value: Value::SingleQuotedString("01:23:34".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("01:23:34".into()), + span: Span::empty(), + }, }, expr_from_projection(only(&select.projection)), ); @@ -5876,7 +5917,10 @@ fn parse_literal_datetime() { assert_eq!( &Expr::TypedString { data_type: DataType::Datetime(None), - value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34.45".into()), + span: Span::empty(), + }, }, expr_from_projection(only(&select.projection)), ); @@ -5889,7 +5933,10 @@ fn parse_literal_timestamp_without_time_zone() { assert_eq!( &Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: Value::SingleQuotedString("1999-01-01 01:23:34".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34".into()), + span: Span::empty(), + }, }, expr_from_projection(only(&select.projection)), ); @@ -5904,7 +5951,10 @@ fn parse_literal_timestamp_with_time_zone() { assert_eq!( &Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::Tz), - value: Value::SingleQuotedString("1999-01-01 01:23:34Z".into()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1999-01-01 01:23:34Z".into()), + span: Span::empty(), + }, }, expr_from_projection(only(&select.projection)), ); @@ -6300,6 +6350,7 @@ fn parse_interval_and_or_xor() { quote_style: None, span: Span::empty(), }))], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident { @@ -6476,8 +6527,9 @@ fn parse_json_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::JSON, - value: Value::SingleQuotedString( - r#"{ + value: ValueWithSpan { + value: Value::SingleQuotedString( + r#"{ "id": 10, "type": "fruit", "name": "apple", @@ -6497,8 +6549,10 @@ fn parse_json_keyword() { ] } }"# - .to_string() - ) + .to_string() + ), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6510,7 +6564,10 @@ fn parse_typed_strings() { assert_eq!( Expr::TypedString { data_type: DataType::JSON, - value: Value::SingleQuotedString(r#"{"foo":"bar"}"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"{"foo":"bar"}"#.into()), + span: Span::empty(), + } }, expr ); @@ -6528,7 +6585,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"0"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"0"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6539,7 +6599,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"123456"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"123456"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6550,7 +6613,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"-3.14"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"-3.14"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6561,7 +6627,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"-0.54321"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"-0.54321"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6572,7 +6641,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"1.23456e05"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"1.23456e05"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -6583,7 +6655,10 @@ fn parse_bignumeric_keyword() { assert_eq!( &Expr::TypedString { data_type: DataType::BigNumeric(ExactNumberInfo::None), - value: Value::SingleQuotedString(r#"-9.876e-3"#.into()) + value: ValueWithSpan { + value: Value::SingleQuotedString(r#"-9.876e-3"#.into()), + span: Span::empty(), + } }, expr_from_projection(only(&select.projection)), ); @@ -7379,7 +7454,8 @@ fn parse_join_syntax_variants() { "SELECT c1 FROM t1 FULL JOIN t2 USING(c1)", ); - let res = parse_sql_statements("SELECT * FROM a OUTER JOIN b ON 1"); + let dialects = all_dialects_except(|d| d.is_table_alias(&Keyword::OUTER, &mut Parser::new(d))); + let res = dialects.parse_sql_statements("SELECT * FROM a OUTER JOIN b ON 1"); assert_eq!( ParserError::ParserError("Expected: APPLY, found: JOIN".to_string()), res.unwrap_err() @@ -7497,7 +7573,7 @@ fn parse_cte_in_data_modification_statements() { assert_eq!(query.with.unwrap().to_string(), "WITH x AS (SELECT 1)"); assert!(matches!(*query.body, SetExpr::Update(_))); } - other => panic!("Expected: UPDATE, got: {:?}", other), + other => panic!("Expected: UPDATE, got: {other:?}"), } match verified_stmt("WITH t (x) AS (SELECT 9) DELETE FROM q WHERE id IN (SELECT x FROM t)") { @@ -7505,7 +7581,7 @@ fn parse_cte_in_data_modification_statements() { assert_eq!(query.with.unwrap().to_string(), "WITH t (x) AS (SELECT 9)"); assert!(matches!(*query.body, SetExpr::Delete(_))); } - other => panic!("Expected: DELETE, got: {:?}", other), + other => panic!("Expected: DELETE, got: {other:?}"), } match verified_stmt("WITH x AS (SELECT 42) INSERT INTO t SELECT foo FROM x") { @@ -7513,7 +7589,7 @@ fn parse_cte_in_data_modification_statements() { assert_eq!(query.with.unwrap().to_string(), "WITH x AS (SELECT 42)"); assert!(matches!(*query.body, SetExpr::Insert(_))); } - other => panic!("Expected: INSERT, got: {:?}", other), + other => panic!("Expected: INSERT, got: {other:?}"), } } @@ -7760,7 +7836,6 @@ fn parse_trim() { Box::new(MySqlDialect {}), //Box::new(BigQueryDialect {}), Box::new(SQLiteDialect {}), - Box::new(DuckDbDialect {}), ]); assert_eq!( @@ -7988,7 +8063,7 @@ fn parse_create_view_with_columns() { .map(|name| ViewColumnDef { name, data_type: None, - options: None + options: None, }) .collect::>() ); @@ -8288,6 +8363,24 @@ fn parse_drop_view() { verified_stmt("DROP MATERIALIZED VIEW IF EXISTS a.b.c"); } +#[test] +fn parse_drop_user() { + let sql = "DROP USER u1"; + match verified_stmt(sql) { + Statement::Drop { + names, object_type, .. + } => { + assert_eq!( + vec!["u1"], + names.iter().map(ToString::to_string).collect::>() + ); + assert_eq!(ObjectType::User, object_type); + } + _ => unreachable!(), + } + verified_stmt("DROP USER IF EXISTS u1"); +} + #[test] fn parse_invalid_subquery_without_parens() { let res = parse_sql_statements("SELECT SELECT 1 FROM bar WHERE 1=1 FROM baz"); @@ -8545,6 +8638,7 @@ fn lateral_function() { distinct: None, top: None, projection: vec![SelectItem::Wildcard(WildcardAdditionalOptions::default())], + exclude: None, top_before_distinct: false, into: None, from: vec![TableWithJoins { @@ -8592,8 +8686,11 @@ fn lateral_function() { #[test] fn parse_start_transaction() { let dialects = all_dialects_except(|d| - // BigQuery does not support this syntax - d.is::()); + // BigQuery and Snowflake does not support this syntax + // + // BigQuery: + // Snowflake: + d.is::() || d.is::()); match dialects .verified_stmt("START TRANSACTION READ ONLY, READ WRITE, ISOLATION LEVEL SERIALIZABLE") { @@ -9372,18 +9469,30 @@ fn parse_grant() { verified_stmt("GRANT SELECT ON ALL TABLES IN SCHEMA db1.sc1 TO APPLICATION role1"); verified_stmt("GRANT SELECT ON ALL TABLES IN SCHEMA db1.sc1 TO APPLICATION ROLE role1"); verified_stmt("GRANT SELECT ON ALL TABLES IN SCHEMA db1.sc1 TO SHARE share1"); + verified_stmt("GRANT SELECT ON ALL VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON ALL MATERIALIZED VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON ALL EXTERNAL TABLES IN SCHEMA db1.sc1 TO ROLE role1"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO a:b"); verified_stmt("GRANT USAGE ON SCHEMA sc1 TO GROUP group1"); verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST"); + verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST COPY CURRENT GRANTS"); + verified_stmt("GRANT OWNERSHIP ON ALL TABLES IN SCHEMA DEV_STAS_ROGOZHIN TO ROLE ANALYST REVOKE CURRENT GRANTS"); verified_stmt("GRANT USAGE ON DATABASE db1 TO ROLE role1"); verified_stmt("GRANT USAGE ON WAREHOUSE wh1 TO ROLE role1"); verified_stmt("GRANT OWNERSHIP ON INTEGRATION int1 TO ROLE role1"); verified_stmt("GRANT SELECT ON VIEW view1 TO ROLE role1"); verified_stmt("GRANT EXEC ON my_sp TO runner"); verified_stmt("GRANT UPDATE ON my_table TO updater_role AS dbo"); - all_dialects_where(|d| d.identifier_quote_style("none") == Some('[')) .verified_stmt("GRANT SELECT ON [my_table] TO [public]"); + verified_stmt("GRANT SELECT ON FUTURE SCHEMAS IN DATABASE db1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE TABLES IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE EXTERNAL TABLES IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE MATERIALIZED VIEWS IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT SELECT ON FUTURE SEQUENCES IN SCHEMA db1.sc1 TO ROLE role1"); + verified_stmt("GRANT USAGE ON PROCEDURE db1.sc1.foo(INT) TO ROLE role1"); + verified_stmt("GRANT USAGE ON FUNCTION db1.sc1.foo(INT) TO ROLE role1"); } #[test] @@ -9526,6 +9635,7 @@ fn parse_merge() { projection: vec![SelectItem::Wildcard( WildcardAdditionalOptions::default() )], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![ @@ -10038,7 +10148,7 @@ fn parse_offset_and_limit() { #[test] fn parse_time_functions() { fn test_time_function(func_name: &'static str) { - let sql = format!("SELECT {}()", func_name); + let sql = format!("SELECT {func_name}()"); let select = verified_only_select(&sql); let select_localtime_func_call_ast = Function { name: ObjectName::from(vec![Ident::new(func_name)]), @@ -10060,7 +10170,7 @@ fn parse_time_functions() { ); // Validating Parenthesis - let sql_without_parens = format!("SELECT {}", func_name); + let sql_without_parens = format!("SELECT {func_name}"); let mut ast_without_parens = select_localtime_func_call_ast; ast_without_parens.args = FunctionArguments::None; assert_eq!( @@ -11061,10 +11171,15 @@ fn parse_non_latin_identifiers() { Box::new(RedshiftSqlDialect {}), Box::new(MySqlDialect {}), ]); - supported_dialects.verified_stmt("SELECT a.説明 FROM test.public.inter01 AS a"); supported_dialects.verified_stmt("SELECT a.説明 FROM inter01 AS a, inter01_transactions AS b WHERE a.説明 = b.取引 GROUP BY a.説明"); supported_dialects.verified_stmt("SELECT 説明, hühnervögel, garçon, Москва, 東京 FROM inter01"); + + let supported_dialects = TestedDialects::new(vec![ + Box::new(GenericDialect {}), + Box::new(DuckDbDialect {}), + Box::new(MsSqlDialect {}), + ]); assert!(supported_dialects .parse_sql_statements("SELECT 💝 FROM table1") .is_err()); @@ -11113,7 +11228,7 @@ fn parse_trailing_comma() { trailing_commas.verified_stmt(r#"SELECT "from" FROM "from""#); // doesn't allow any trailing commas - let trailing_commas = TestedDialects::new(vec![Box::new(GenericDialect {})]); + let trailing_commas = TestedDialects::new(vec![Box::new(PostgreSqlDialect {})]); assert_eq!( trailing_commas @@ -11303,6 +11418,8 @@ fn parse_execute_stored_procedure() { immediate: false, using: vec![], into: vec![], + output: false, + default: false, }; assert_eq!( // Microsoft SQL Server does not use parentheses around arguments for EXECUTE @@ -11317,6 +11434,18 @@ fn parse_execute_stored_procedure() { ), expected ); + match ms_and_generic().verified_stmt("EXECUTE dbo.proc1 @ReturnVal = @X OUTPUT") { + Statement::Execute { output, .. } => { + assert!(output); + } + _ => unreachable!(), + } + match ms_and_generic().verified_stmt("EXECUTE dbo.proc1 DEFAULT") { + Statement::Execute { default, .. } => { + assert!(default); + } + _ => unreachable!(), + } } #[test] @@ -11335,6 +11464,8 @@ fn parse_execute_immediate() { into: vec![Ident::new("a")], name: None, has_parentheses: false, + output: false, + default: false, }; let stmt = dialects.verified_stmt("EXECUTE IMMEDIATE 'SELECT 1' INTO a USING 1 AS b"); @@ -11437,6 +11568,7 @@ fn parse_unload() { top: None, top_before_distinct: false, projection: vec![UnnamedExpr(Expr::Identifier(Ident::new("cola"))),], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("tab")])), @@ -11637,6 +11769,7 @@ fn parse_connect_by() { SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("manager_id"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("title"))), ], + exclude: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("employees")])), joins: vec![], @@ -11718,6 +11851,7 @@ fn parse_connect_by() { SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("manager_id"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("title"))), ], + exclude: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::new("employees")])), joins: vec![], @@ -12651,6 +12785,7 @@ fn test_extract_seconds_ok() { format: None, }), })], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -14301,7 +14436,7 @@ fn overflow() { let expr = std::iter::repeat_n("1", 1000) .collect::>() .join(" + "); - let sql = format!("SELECT {}", expr); + let sql = format!("SELECT {expr}"); let mut statements = Parser::parse_sql(&GenericDialect {}, sql.as_str()).unwrap(); let statement = statements.pop().unwrap(); @@ -14601,7 +14736,7 @@ fn test_conditional_statement_span() { else_block.unwrap().span() ); } - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } } @@ -14723,6 +14858,7 @@ fn test_select_from_first() { distinct: None, top: None, projection, + exclude: None, top_before_distinct: false, into: None, from: vec![TableWithJoins { @@ -14821,7 +14957,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Point), - value: Value::SingleQuotedString("1,2".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2".to_string()), + span: Span::empty(), + }, } ); @@ -14830,7 +14969,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Line), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); @@ -14839,7 +14981,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::GeometricPath), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); let sql = "box '1,2,3,4'"; @@ -14847,7 +14992,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::GeometricBox), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); @@ -14856,7 +15004,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Circle), - value: Value::SingleQuotedString("1,2,3".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3".to_string()), + span: Span::empty(), + }, } ); @@ -14865,7 +15016,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::Polygon), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); let sql = "lseg '1,2,3,4'"; @@ -14873,7 +15027,10 @@ fn test_geometry_type() { all_dialects_where(|d| d.supports_geometric_types()).verified_expr(sql), Expr::TypedString { data_type: DataType::GeometricType(GeometricTypeKind::LineSegment), - value: Value::SingleQuotedString("1,2,3,4".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("1,2,3,4".to_string()), + span: Span::empty(), + }, } ); } @@ -15210,10 +15367,426 @@ fn parse_pipeline_operator() { dialects.verified_stmt("SELECT * FROM tbl |> TABLESAMPLE SYSTEM (50 PERCENT)"); dialects.verified_stmt("SELECT * FROM tbl |> TABLESAMPLE SYSTEM (50) REPEATABLE (10)"); + // rename pipe operator + dialects.verified_stmt("SELECT * FROM users |> RENAME old_name AS new_name"); + dialects.verified_stmt("SELECT * FROM users |> RENAME id AS user_id, name AS user_name"); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> RENAME id user_id", + "SELECT * FROM users |> RENAME id AS user_id", + ); + + // union pipe operator + dialects.verified_stmt("SELECT * FROM users |> UNION ALL (SELECT * FROM admins)"); + dialects.verified_stmt("SELECT * FROM users |> UNION DISTINCT (SELECT * FROM admins)"); + dialects.verified_stmt("SELECT * FROM users |> UNION (SELECT * FROM admins)"); + + // union pipe operator with multiple queries + dialects.verified_stmt( + "SELECT * FROM users |> UNION ALL (SELECT * FROM admins), (SELECT * FROM guests)", + ); + dialects.verified_stmt("SELECT * FROM users |> UNION DISTINCT (SELECT * FROM admins), (SELECT * FROM guests), (SELECT * FROM employees)"); + dialects.verified_stmt( + "SELECT * FROM users |> UNION (SELECT * FROM admins), (SELECT * FROM guests)", + ); + + // union pipe operator with BY NAME modifier + dialects.verified_stmt("SELECT * FROM users |> UNION BY NAME (SELECT * FROM admins)"); + dialects.verified_stmt("SELECT * FROM users |> UNION ALL BY NAME (SELECT * FROM admins)"); + dialects.verified_stmt("SELECT * FROM users |> UNION DISTINCT BY NAME (SELECT * FROM admins)"); + + // union pipe operator with BY NAME and multiple queries + dialects.verified_stmt( + "SELECT * FROM users |> UNION BY NAME (SELECT * FROM admins), (SELECT * FROM guests)", + ); + + // intersect pipe operator (BigQuery requires DISTINCT modifier for INTERSECT) + dialects.verified_stmt("SELECT * FROM users |> INTERSECT DISTINCT (SELECT * FROM admins)"); + + // intersect pipe operator with BY NAME modifier + dialects + .verified_stmt("SELECT * FROM users |> INTERSECT DISTINCT BY NAME (SELECT * FROM admins)"); + + // intersect pipe operator with multiple queries + dialects.verified_stmt( + "SELECT * FROM users |> INTERSECT DISTINCT (SELECT * FROM admins), (SELECT * FROM guests)", + ); + + // intersect pipe operator with BY NAME and multiple queries + dialects.verified_stmt("SELECT * FROM users |> INTERSECT DISTINCT BY NAME (SELECT * FROM admins), (SELECT * FROM guests)"); + + // except pipe operator (BigQuery requires DISTINCT modifier for EXCEPT) + dialects.verified_stmt("SELECT * FROM users |> EXCEPT DISTINCT (SELECT * FROM admins)"); + + // except pipe operator with BY NAME modifier + dialects.verified_stmt("SELECT * FROM users |> EXCEPT DISTINCT BY NAME (SELECT * FROM admins)"); + + // except pipe operator with multiple queries + dialects.verified_stmt( + "SELECT * FROM users |> EXCEPT DISTINCT (SELECT * FROM admins), (SELECT * FROM guests)", + ); + + // except pipe operator with BY NAME and multiple queries + dialects.verified_stmt("SELECT * FROM users |> EXCEPT DISTINCT BY NAME (SELECT * FROM admins), (SELECT * FROM guests)"); + + // call pipe operator + dialects.verified_stmt("SELECT * FROM users |> CALL my_function()"); + dialects.verified_stmt("SELECT * FROM users |> CALL process_data(5, 'test')"); + dialects.verified_stmt( + "SELECT * FROM users |> CALL namespace.function_name(col1, col2, 'literal')", + ); + + // call pipe operator with complex arguments + dialects.verified_stmt("SELECT * FROM users |> CALL transform_data(col1 + col2)"); + dialects.verified_stmt("SELECT * FROM users |> CALL analyze_data('param1', 100, true)"); + + // call pipe operator with aliases + dialects.verified_stmt("SELECT * FROM input_table |> CALL tvf1(arg1) AS al"); + dialects.verified_stmt("SELECT * FROM users |> CALL process_data(5) AS result_table"); + dialects.verified_stmt("SELECT * FROM users |> CALL namespace.func() AS my_alias"); + + // multiple call pipe operators in sequence + dialects.verified_stmt("SELECT * FROM input_table |> CALL tvf1(arg1) |> CALL tvf2(arg2, arg3)"); + dialects.verified_stmt( + "SELECT * FROM data |> CALL transform(col1) |> CALL validate() |> CALL process(param)", + ); + + // multiple call pipe operators with aliases + dialects.verified_stmt( + "SELECT * FROM input_table |> CALL tvf1(arg1) AS step1 |> CALL tvf2(arg2) AS step2", + ); + dialects.verified_stmt( + "SELECT * FROM data |> CALL preprocess() AS clean_data |> CALL analyze(mode) AS results", + ); + + // call pipe operators mixed with other pipe operators + dialects.verified_stmt( + "SELECT * FROM users |> CALL transform() |> WHERE status = 'active' |> CALL process(param)", + ); + dialects.verified_stmt( + "SELECT * FROM data |> CALL preprocess() AS clean |> SELECT col1, col2 |> CALL validate()", + ); + + // pivot pipe operator + dialects.verified_stmt( + "SELECT * FROM monthly_sales |> PIVOT(SUM(amount) FOR quarter IN ('Q1', 'Q2', 'Q3', 'Q4'))", + ); + dialects.verified_stmt("SELECT * FROM sales_data |> PIVOT(AVG(revenue) FOR region IN ('North', 'South', 'East', 'West'))"); + + // pivot pipe operator with multiple aggregate functions + dialects.verified_stmt("SELECT * FROM data |> PIVOT(SUM(sales) AS total_sales, COUNT(*) AS num_transactions FOR month IN ('Jan', 'Feb', 'Mar'))"); + + // pivot pipe operator with compound column names + dialects.verified_stmt("SELECT * FROM sales |> PIVOT(SUM(amount) FOR product.category IN ('Electronics', 'Clothing'))"); + + // pivot pipe operator mixed with other pipe operators + dialects.verified_stmt("SELECT * FROM sales_data |> WHERE year = 2023 |> PIVOT(SUM(revenue) FOR quarter IN ('Q1', 'Q2', 'Q3', 'Q4'))"); + + // pivot pipe operator with aliases + dialects.verified_stmt("SELECT * FROM monthly_sales |> PIVOT(SUM(sales) FOR quarter IN ('Q1', 'Q2')) AS quarterly_sales"); + dialects.verified_stmt("SELECT * FROM data |> PIVOT(AVG(price) FOR category IN ('A', 'B', 'C')) AS avg_by_category"); + dialects.verified_stmt("SELECT * FROM sales |> PIVOT(COUNT(*) AS transactions, SUM(amount) AS total FOR region IN ('North', 'South')) AS regional_summary"); + + // pivot pipe operator with implicit aliases (without AS keyword) + dialects.verified_query_with_canonical( + "SELECT * FROM monthly_sales |> PIVOT(SUM(sales) FOR quarter IN ('Q1', 'Q2')) quarterly_sales", + "SELECT * FROM monthly_sales |> PIVOT(SUM(sales) FOR quarter IN ('Q1', 'Q2')) AS quarterly_sales", + ); + dialects.verified_query_with_canonical( + "SELECT * FROM data |> PIVOT(AVG(price) FOR category IN ('A', 'B', 'C')) avg_by_category", + "SELECT * FROM data |> PIVOT(AVG(price) FOR category IN ('A', 'B', 'C')) AS avg_by_category", + ); + + // unpivot pipe operator basic usage + dialects + .verified_stmt("SELECT * FROM sales |> UNPIVOT(revenue FOR quarter IN (Q1, Q2, Q3, Q4))"); + dialects.verified_stmt("SELECT * FROM data |> UNPIVOT(value FOR category IN (A, B, C))"); + dialects.verified_stmt( + "SELECT * FROM metrics |> UNPIVOT(measurement FOR metric_type IN (cpu, memory, disk))", + ); + + // unpivot pipe operator with multiple columns + dialects.verified_stmt("SELECT * FROM quarterly_sales |> UNPIVOT(amount FOR period IN (jan, feb, mar, apr, may, jun))"); + dialects.verified_stmt( + "SELECT * FROM report |> UNPIVOT(score FOR subject IN (math, science, english, history))", + ); + + // unpivot pipe operator mixed with other pipe operators + dialects.verified_stmt("SELECT * FROM sales_data |> WHERE year = 2023 |> UNPIVOT(revenue FOR quarter IN (Q1, Q2, Q3, Q4))"); + + // unpivot pipe operator with aliases + dialects.verified_stmt("SELECT * FROM quarterly_sales |> UNPIVOT(amount FOR period IN (Q1, Q2)) AS unpivoted_sales"); + dialects.verified_stmt( + "SELECT * FROM data |> UNPIVOT(value FOR category IN (A, B, C)) AS transformed_data", + ); + dialects.verified_stmt("SELECT * FROM metrics |> UNPIVOT(measurement FOR metric_type IN (cpu, memory)) AS metric_measurements"); + + // unpivot pipe operator with implicit aliases (without AS keyword) + dialects.verified_query_with_canonical( + "SELECT * FROM quarterly_sales |> UNPIVOT(amount FOR period IN (Q1, Q2)) unpivoted_sales", + "SELECT * FROM quarterly_sales |> UNPIVOT(amount FOR period IN (Q1, Q2)) AS unpivoted_sales", + ); + dialects.verified_query_with_canonical( + "SELECT * FROM data |> UNPIVOT(value FOR category IN (A, B, C)) transformed_data", + "SELECT * FROM data |> UNPIVOT(value FOR category IN (A, B, C)) AS transformed_data", + ); + // many pipes dialects.verified_stmt( "SELECT * FROM CustomerOrders |> AGGREGATE SUM(cost) AS total_cost GROUP BY customer_id, state, item_type |> EXTEND COUNT(*) OVER (PARTITION BY customer_id) AS num_orders |> WHERE num_orders > 1 |> AGGREGATE AVG(total_cost) AS average GROUP BY state DESC, item_type ASC", ); + + // join pipe operator - INNER JOIN + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id"); + dialects.verified_stmt("SELECT * FROM users |> INNER JOIN orders ON users.id = orders.user_id"); + + // join pipe operator - LEFT JOIN + dialects.verified_stmt("SELECT * FROM users |> LEFT JOIN orders ON users.id = orders.user_id"); + dialects.verified_stmt( + "SELECT * FROM users |> LEFT OUTER JOIN orders ON users.id = orders.user_id", + ); + + // join pipe operator - RIGHT JOIN + dialects.verified_stmt("SELECT * FROM users |> RIGHT JOIN orders ON users.id = orders.user_id"); + dialects.verified_stmt( + "SELECT * FROM users |> RIGHT OUTER JOIN orders ON users.id = orders.user_id", + ); + + // join pipe operator - FULL JOIN + dialects.verified_stmt("SELECT * FROM users |> FULL JOIN orders ON users.id = orders.user_id"); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> FULL OUTER JOIN orders ON users.id = orders.user_id", + "SELECT * FROM users |> FULL JOIN orders ON users.id = orders.user_id", + ); + + // join pipe operator - CROSS JOIN + dialects.verified_stmt("SELECT * FROM users |> CROSS JOIN orders"); + + // join pipe operator with USING + dialects.verified_query_with_canonical( + "SELECT * FROM users |> JOIN orders USING (user_id)", + "SELECT * FROM users |> JOIN orders USING(user_id)", + ); + dialects.verified_query_with_canonical( + "SELECT * FROM users |> LEFT JOIN orders USING (user_id, order_date)", + "SELECT * FROM users |> LEFT JOIN orders USING(user_id, order_date)", + ); + + // join pipe operator with alias + dialects.verified_query_with_canonical( + "SELECT * FROM users |> JOIN orders o ON users.id = o.user_id", + "SELECT * FROM users |> JOIN orders AS o ON users.id = o.user_id", + ); + dialects.verified_stmt("SELECT * FROM users |> LEFT JOIN orders AS o ON users.id = o.user_id"); + + // join pipe operator with complex ON condition + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id AND orders.status = 'active'"); + dialects.verified_stmt("SELECT * FROM users |> LEFT JOIN orders ON users.id = orders.user_id AND orders.amount > 100"); + + // multiple join pipe operators + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id |> JOIN products ON orders.product_id = products.id"); + dialects.verified_stmt("SELECT * FROM users |> LEFT JOIN orders ON users.id = orders.user_id |> RIGHT JOIN products ON orders.product_id = products.id"); + + // join pipe operator with other pipe operators + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id |> WHERE orders.amount > 100"); + dialects.verified_stmt("SELECT * FROM users |> WHERE users.active = true |> LEFT JOIN orders ON users.id = orders.user_id"); + dialects.verified_stmt("SELECT * FROM users |> JOIN orders ON users.id = orders.user_id |> SELECT users.name, orders.amount"); +} + +#[test] +fn parse_pipeline_operator_negative_tests() { + let dialects = all_dialects_where(|d| d.supports_pipe_operator()); + + // Test that plain EXCEPT without DISTINCT fails + assert_eq!( + ParserError::ParserError("EXCEPT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> EXCEPT (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that EXCEPT ALL fails + assert_eq!( + ParserError::ParserError("EXCEPT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> EXCEPT ALL (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that EXCEPT BY NAME without DISTINCT fails + assert_eq!( + ParserError::ParserError("EXCEPT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> EXCEPT BY NAME (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that EXCEPT ALL BY NAME fails + assert_eq!( + ParserError::ParserError("EXCEPT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements( + "SELECT * FROM users |> EXCEPT ALL BY NAME (SELECT * FROM admins)" + ) + .unwrap_err() + ); + + // Test that plain INTERSECT without DISTINCT fails + assert_eq!( + ParserError::ParserError("INTERSECT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> INTERSECT (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that INTERSECT ALL fails + assert_eq!( + ParserError::ParserError("INTERSECT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> INTERSECT ALL (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that INTERSECT BY NAME without DISTINCT fails + assert_eq!( + ParserError::ParserError("INTERSECT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements("SELECT * FROM users |> INTERSECT BY NAME (SELECT * FROM admins)") + .unwrap_err() + ); + + // Test that INTERSECT ALL BY NAME fails + assert_eq!( + ParserError::ParserError("INTERSECT pipe operator requires DISTINCT modifier".to_string()), + dialects + .parse_sql_statements( + "SELECT * FROM users |> INTERSECT ALL BY NAME (SELECT * FROM admins)" + ) + .unwrap_err() + ); + + // Test that CALL without function name fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL") + .is_err()); + + // Test that CALL without parentheses fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL my_function") + .is_err()); + + // Test that CALL with invalid function syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL 123invalid") + .is_err()); + + // Test that CALL with malformed arguments fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL my_function(,)") + .is_err()); + + // Test that CALL with invalid alias syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CALL my_function() AS") + .is_err()); + + // Test that PIVOT without parentheses fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT SUM(amount) FOR month IN ('Jan')") + .is_err()); + + // Test that PIVOT without FOR keyword fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT(SUM(amount) month IN ('Jan'))") + .is_err()); + + // Test that PIVOT without IN keyword fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT(SUM(amount) FOR month ('Jan'))") + .is_err()); + + // Test that PIVOT with empty IN list fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT(SUM(amount) FOR month IN ())") + .is_err()); + + // Test that PIVOT with invalid alias syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> PIVOT(SUM(amount) FOR month IN ('Jan')) AS") + .is_err()); + + // Test UNPIVOT negative cases + + // Test that UNPIVOT without parentheses fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT value FOR name IN col1, col2") + .is_err()); + + // Test that UNPIVOT without FOR keyword fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value name IN (col1, col2))") + .is_err()); + + // Test that UNPIVOT without IN keyword fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR name (col1, col2))") + .is_err()); + + // Test that UNPIVOT with missing value column fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(FOR name IN (col1, col2))") + .is_err()); + + // Test that UNPIVOT with missing name column fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR IN (col1, col2))") + .is_err()); + + // Test that UNPIVOT with empty IN list fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR name IN ())") + .is_err()); + + // Test that UNPIVOT with invalid alias syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR name IN (col1, col2)) AS") + .is_err()); + + // Test that UNPIVOT with missing closing parenthesis fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> UNPIVOT(value FOR name IN (col1, col2)") + .is_err()); + + // Test that JOIN without table name fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> JOIN ON users.id = orders.user_id") + .is_err()); + + // Test that CROSS JOIN with ON condition fails + assert!(dialects + .parse_sql_statements( + "SELECT * FROM users |> CROSS JOIN orders ON users.id = orders.user_id" + ) + .is_err()); + + // Test that CROSS JOIN with USING condition fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> CROSS JOIN orders USING (user_id)") + .is_err()); + + // Test that JOIN with empty USING list fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> JOIN orders USING ()") + .is_err()); + + // Test that JOIN with malformed ON condition fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> JOIN orders ON") + .is_err()); + + // Test that JOIN with invalid USING syntax fails + assert!(dialects + .parse_sql_statements("SELECT * FROM users |> JOIN orders USING user_id") + .is_err()); } #[test] @@ -15300,6 +15873,11 @@ fn parse_return() { let _ = all_dialects().verified_stmt("RETURN 1"); } +#[test] +fn parse_subquery_limit() { + let _ = all_dialects().verified_stmt("SELECT t1_id, t1_name FROM t1 WHERE t1_id IN (SELECT t2_id FROM t2 WHERE t1_name = t2_name LIMIT 10)"); +} + #[test] fn test_open() { let open_cursor = "OPEN Employee_Cursor"; @@ -15346,3 +15924,262 @@ fn check_enforced() { "CREATE TABLE t (a INT, b INT, c INT, CHECK (a > 0) NOT ENFORCED, CHECK (b > 0) ENFORCED, CHECK (c > 0))", ); } + +#[test] +fn join_precedence() { + all_dialects_except(|d| !d.supports_left_associative_joins_without_parens()) + .verified_query_with_canonical( + "SELECT * + FROM t1 + NATURAL JOIN t5 + INNER JOIN t0 ON (t0.v1 + t5.v0) > 0 + WHERE t0.v1 = t1.v0", + // canonical string without parentheses + "SELECT * FROM t1 NATURAL JOIN t5 INNER JOIN t0 ON (t0.v1 + t5.v0) > 0 WHERE t0.v1 = t1.v0", + ); + all_dialects_except(|d| d.supports_left_associative_joins_without_parens()).verified_query_with_canonical( + "SELECT * + FROM t1 + NATURAL JOIN t5 + INNER JOIN t0 ON (t0.v1 + t5.v0) > 0 + WHERE t0.v1 = t1.v0", + // canonical string with parentheses + "SELECT * FROM t1 NATURAL JOIN (t5 INNER JOIN t0 ON (t0.v1 + t5.v0) > 0) WHERE t0.v1 = t1.v0", + ); +} + +#[test] +fn parse_create_procedure_with_language() { + let sql = r#"CREATE PROCEDURE test_proc LANGUAGE sql AS BEGIN SELECT 1; END"#; + match verified_stmt(sql) { + Statement::CreateProcedure { + or_alter, + name, + params, + language, + .. + } => { + assert_eq!(or_alter, false); + assert_eq!(name.to_string(), "test_proc"); + assert_eq!(params, Some(vec![])); + assert_eq!( + language, + Some(Ident { + value: "sql".into(), + quote_style: None, + span: Span { + start: Location::empty(), + end: Location::empty() + } + }) + ); + } + _ => unreachable!(), + } +} + +#[test] +fn parse_create_procedure_with_parameter_modes() { + let sql = r#"CREATE PROCEDURE test_proc (IN a INTEGER, OUT b TEXT, INOUT c TIMESTAMP, d BOOL) AS BEGIN SELECT 1; END"#; + match verified_stmt(sql) { + Statement::CreateProcedure { + or_alter, + name, + params, + .. + } => { + assert_eq!(or_alter, false); + assert_eq!(name.to_string(), "test_proc"); + let fake_span = Span { + start: Location { line: 0, column: 0 }, + end: Location { line: 0, column: 0 }, + }; + assert_eq!( + params, + Some(vec![ + ProcedureParam { + name: Ident { + value: "a".into(), + quote_style: None, + span: fake_span, + }, + data_type: DataType::Integer(None), + mode: Some(ArgMode::In) + }, + ProcedureParam { + name: Ident { + value: "b".into(), + quote_style: None, + span: fake_span, + }, + data_type: DataType::Text, + mode: Some(ArgMode::Out) + }, + ProcedureParam { + name: Ident { + value: "c".into(), + quote_style: None, + span: fake_span, + }, + data_type: DataType::Timestamp(None, TimezoneInfo::None), + mode: Some(ArgMode::InOut) + }, + ProcedureParam { + name: Ident { + value: "d".into(), + quote_style: None, + span: fake_span, + }, + data_type: DataType::Bool, + mode: None + }, + ]) + ); + } + _ => unreachable!(), + } +} + +#[test] +fn test_select_exclude() { + let dialects = all_dialects_where(|d| d.supports_select_wildcard_exclude()); + match &dialects + .verified_only_select("SELECT * EXCLUDE c1 FROM test") + .projection[0] + { + SelectItem::Wildcard(WildcardAdditionalOptions { opt_exclude, .. }) => { + assert_eq!( + *opt_exclude, + Some(ExcludeSelectItem::Single(Ident::new("c1"))) + ); + } + _ => unreachable!(), + } + match &dialects + .verified_only_select("SELECT * EXCLUDE (c1, c2) FROM test") + .projection[0] + { + SelectItem::Wildcard(WildcardAdditionalOptions { opt_exclude, .. }) => { + assert_eq!( + *opt_exclude, + Some(ExcludeSelectItem::Multiple(vec![ + Ident::new("c1"), + Ident::new("c2") + ])) + ); + } + _ => unreachable!(), + } + let select = dialects.verified_only_select("SELECT * EXCLUDE c1, c2 FROM test"); + match &select.projection[0] { + SelectItem::Wildcard(WildcardAdditionalOptions { opt_exclude, .. }) => { + assert_eq!( + *opt_exclude, + Some(ExcludeSelectItem::Single(Ident::new("c1"))) + ); + } + _ => unreachable!(), + } + match &select.projection[1] { + SelectItem::UnnamedExpr(Expr::Identifier(ident)) => { + assert_eq!(*ident, Ident::new("c2")); + } + _ => unreachable!(), + } + + let dialects = all_dialects_where(|d| d.supports_select_exclude()); + let select = dialects.verified_only_select("SELECT *, c1 EXCLUDE c1 FROM test"); + match &select.projection[0] { + SelectItem::Wildcard(additional_options) => { + assert_eq!(*additional_options, WildcardAdditionalOptions::default()); + } + _ => unreachable!(), + } + assert_eq!( + select.exclude, + Some(ExcludeSelectItem::Single(Ident::new("c1"))) + ); + + let dialects = all_dialects_where(|d| { + d.supports_select_wildcard_exclude() && !d.supports_select_exclude() + }); + let select = dialects.verified_only_select("SELECT * EXCLUDE c1 FROM test"); + match &select.projection[0] { + SelectItem::Wildcard(WildcardAdditionalOptions { opt_exclude, .. }) => { + assert_eq!( + *opt_exclude, + Some(ExcludeSelectItem::Single(Ident::new("c1"))) + ); + } + _ => unreachable!(), + } + + // Dialects that only support the wildcard form and do not accept EXCLUDE as an implicity alias + // will fail when encountered with the `c2` ident + let dialects = all_dialects_where(|d| { + d.supports_select_wildcard_exclude() + && !d.supports_select_exclude() + && d.is_column_alias(&Keyword::EXCLUDE, &mut Parser::new(d)) + }); + assert_eq!( + dialects + .parse_sql_statements("SELECT *, c1 EXCLUDE c2 FROM test") + .err() + .unwrap(), + ParserError::ParserError("Expected: end of statement, found: c2".to_string()) + ); + + // Dialects that only support the wildcard form and accept EXCLUDE as an implicity alias + // will fail when encountered with the `EXCLUDE` keyword + let dialects = all_dialects_where(|d| { + d.supports_select_wildcard_exclude() + && !d.supports_select_exclude() + && !d.is_column_alias(&Keyword::EXCLUDE, &mut Parser::new(d)) + }); + assert_eq!( + dialects + .parse_sql_statements("SELECT *, c1 EXCLUDE c2 FROM test") + .err() + .unwrap(), + ParserError::ParserError("Expected: end of statement, found: EXCLUDE".to_string()) + ); +} + +#[test] +fn test_no_semicolon_required_between_statements() { + let sql = r#" +SELECT * FROM tbl1 +SELECT * FROM tbl2 + "#; + + let dialects = all_dialects_with_options(ParserOptions { + trailing_commas: false, + unescape: true, + require_semicolon_stmt_delimiter: false, + }); + let stmts = dialects.parse_sql_statements(sql).unwrap(); + assert_eq!(stmts.len(), 2); + assert!(stmts.iter().all(|s| matches!(s, Statement::Query { .. }))); +} + +#[test] +fn test_identifier_unicode_support() { + let sql = r#"SELECT phoneǤЖשचᎯ⻩☯♜🦄⚛🀄ᚠ⌛🌀 AS tbl FROM customers"#; + let dialects = TestedDialects::new(vec![ + Box::new(MySqlDialect {}), + Box::new(RedshiftSqlDialect {}), + Box::new(PostgreSqlDialect {}), + ]); + let _ = dialects.verified_stmt(sql); +} + +#[test] +fn test_identifier_unicode_start() { + let sql = r#"SELECT 💝phone AS 💝 FROM customers"#; + let dialects = TestedDialects::new(vec![ + Box::new(MySqlDialect {}), + Box::new(RedshiftSqlDialect {}), + Box::new(PostgreSqlDialect {}), + ]); + let _ = dialects.verified_stmt(sql); +} diff --git a/tests/sqlparser_databricks.rs b/tests/sqlparser_databricks.rs index 99b7eecd..a27e0699 100644 --- a/tests/sqlparser_databricks.rs +++ b/tests/sqlparser_databricks.rs @@ -19,6 +19,7 @@ use sqlparser::ast::helpers::attached_token::AttachedToken; use sqlparser::ast::*; use sqlparser::dialect::{DatabricksDialect, GenericDialect}; use sqlparser::parser::ParserError; +use sqlparser::tokenizer::Span; use test_utils::*; #[macro_use] @@ -213,7 +214,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - databricks().verified_stmt(&format!("USE {}", object_name)), + databricks().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -221,7 +222,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - databricks().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), + databricks().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -233,21 +234,21 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with keyword and different type of quotes assert_eq!( - databricks().verified_stmt(&format!("USE CATALOG {0}my_catalog{0}", quote)), + databricks().verified_stmt(&format!("USE CATALOG {quote}my_catalog{quote}")), Statement::Use(Use::Catalog(ObjectName::from(vec![Ident::with_quote( quote, "my_catalog".to_string(), )]))) ); assert_eq!( - databricks().verified_stmt(&format!("USE DATABASE {0}my_database{0}", quote)), + databricks().verified_stmt(&format!("USE DATABASE {quote}my_database{quote}")), Statement::Use(Use::Database(ObjectName::from(vec![Ident::with_quote( quote, "my_database".to_string(), )]))) ); assert_eq!( - databricks().verified_stmt(&format!("USE SCHEMA {0}my_schema{0}", quote)), + databricks().verified_stmt(&format!("USE SCHEMA {quote}my_schema{quote}")), Statement::Use(Use::Schema(ObjectName::from(vec![Ident::with_quote( quote, "my_schema".to_string(), @@ -328,7 +329,10 @@ fn data_type_timestamp_ntz() { databricks().verified_expr("TIMESTAMP_NTZ '2025-03-29T18:52:00'"), Expr::TypedString { data_type: DataType::TimestampNtz, - value: Value::SingleQuotedString("2025-03-29T18:52:00".to_owned()) + value: ValueWithSpan { + value: Value::SingleQuotedString("2025-03-29T18:52:00".to_owned()), + span: Span::empty(), + } } ); @@ -357,6 +361,6 @@ fn data_type_timestamp_ntz() { }] ); } - s => panic!("Unexpected statement: {:?}", s), + s => panic!("Unexpected statement: {s:?}"), } } diff --git a/tests/sqlparser_duckdb.rs b/tests/sqlparser_duckdb.rs index 8e498365..fe14b7ba 100644 --- a/tests/sqlparser_duckdb.rs +++ b/tests/sqlparser_duckdb.rs @@ -24,6 +24,7 @@ use test_utils::*; use sqlparser::ast::*; use sqlparser::dialect::{DuckDbDialect, GenericDialect}; +use sqlparser::parser::ParserError; fn duckdb() -> TestedDialects { TestedDialects::new(vec![Box::new(DuckDbDialect {})]) @@ -44,10 +45,12 @@ fn test_struct() { StructField { field_name: Some(Ident::new("v")), field_type: DataType::Varchar(None), + options: None, }, StructField { field_name: Some(Ident::new("i")), field_type: DataType::Integer(None), + options: None, }, ], StructBracketKind::Parentheses, @@ -84,6 +87,7 @@ fn test_struct() { StructField { field_name: Some(Ident::new("v")), field_type: DataType::Varchar(None), + options: None, }, StructField { field_name: Some(Ident::new("s")), @@ -92,14 +96,17 @@ fn test_struct() { StructField { field_name: Some(Ident::new("a1")), field_type: DataType::Integer(None), + options: None, }, StructField { field_name: Some(Ident::new("a2")), field_type: DataType::Varchar(None), + options: None, }, ], StructBracketKind::Parentheses, ), + options: None, }, ], StructBracketKind::Parentheses, @@ -262,6 +269,7 @@ fn test_select_union_by_name() { distinct: None, top: None, projection: vec![SelectItem::Wildcard(WildcardAdditionalOptions::default())], + exclude: None, top_before_distinct: false, into: None, from: vec![TableWithJoins { @@ -292,6 +300,7 @@ fn test_select_union_by_name() { distinct: None, top: None, projection: vec![SelectItem::Wildcard(WildcardAdditionalOptions::default())], + exclude: None, top_before_distinct: false, into: None, from: vec![TableWithJoins { @@ -362,7 +371,7 @@ fn test_duckdb_specific_int_types() { ("HUGEINT", DataType::HugeInt), ]; for (dtype_string, data_type) in duckdb_dtypes { - let sql = format!("SELECT 123::{}", dtype_string); + let sql = format!("SELECT 123::{dtype_string}"); let select = duckdb().verified_only_select(&sql); assert_eq!( &Expr::Cast { @@ -786,7 +795,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - duckdb().verified_stmt(&format!("USE {}", object_name)), + duckdb().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -794,7 +803,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - duckdb().verified_stmt(&format!("USE {0}{1}{0}", quote, object_name)), + duckdb().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -806,7 +815,9 @@ fn parse_use() { for "e in "e_styles { // Test double identifier with different type of quotes assert_eq!( - duckdb().verified_stmt(&format!("USE {0}CATALOG{0}.{0}my_schema{0}", quote)), + duckdb().verified_stmt(&format!( + "USE {quote}CATALOG{quote}.{quote}my_schema{quote}" + )), Statement::Use(Use::Object(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") @@ -822,3 +833,32 @@ fn parse_use() { ]))) ); } + +#[test] +fn test_duckdb_trim() { + let real_sql = r#"SELECT customer_id, TRIM(item_price_id, '"', "a") AS item_price_id FROM models_staging.subscriptions"#; + assert_eq!(duckdb().verified_stmt(real_sql).to_string(), real_sql); + + let sql_only_select = "SELECT TRIM('xyz', 'a')"; + let select = duckdb().verified_only_select(sql_only_select); + assert_eq!( + &Expr::Trim { + expr: Box::new(Expr::Value( + Value::SingleQuotedString("xyz".to_owned()).with_empty_span() + )), + trim_where: None, + trim_what: None, + trim_characters: Some(vec![Expr::Value( + Value::SingleQuotedString("a".to_owned()).with_empty_span() + )]), + }, + expr_from_projection(only(&select.projection)) + ); + + // missing comma separation + let error_sql = "SELECT TRIM('xyz' 'a')"; + assert_eq!( + ParserError::ParserError("Expected: ), found: 'a'".to_owned()), + duckdb().parse_sql_statements(error_sql).unwrap_err() + ); +} diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs index fd52b773..56a72ec8 100644 --- a/tests/sqlparser_hive.rs +++ b/tests/sqlparser_hive.rs @@ -524,7 +524,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - hive().verified_stmt(&format!("USE {}", object_name)), + hive().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -532,7 +532,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - hive().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), + hive().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs index 2a314502..50c6448d 100644 --- a/tests/sqlparser_mssql.rs +++ b/tests/sqlparser_mssql.rs @@ -32,7 +32,7 @@ use sqlparser::ast::DeclareAssignment::MsSqlAssignment; use sqlparser::ast::Value::SingleQuotedString; use sqlparser::ast::*; use sqlparser::dialect::{GenericDialect, MsSqlDialect}; -use sqlparser::parser::{Parser, ParserError}; +use sqlparser::parser::{Parser, ParserError, ParserOptions}; #[test] fn parse_mssql_identifiers() { @@ -126,6 +126,7 @@ fn parse_create_procedure() { projection: vec![SelectItem::UnnamedExpr(Expr::Value( (number("1")).with_empty_span() ))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -153,7 +154,8 @@ fn parse_create_procedure() { quote_style: None, span: Span::empty(), }, - data_type: DataType::Int(None) + data_type: DataType::Int(None), + mode: None, }, ProcedureParam { name: Ident { @@ -164,14 +166,16 @@ fn parse_create_procedure() { data_type: DataType::Varchar(Some(CharacterLength::IntegerLength { length: 256, unit: None - })) + })), + mode: None, } ]), name: ObjectName::from(vec![Ident { value: "test".into(), quote_style: None, span: Span::empty(), - }]) + }]), + language: None, } ) } @@ -1365,6 +1369,7 @@ fn parse_substring_in_select() { special: true, shorthand: false, })], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident { @@ -1513,6 +1518,7 @@ fn parse_mssql_declare() { (Value::Number("4".parse().unwrap(), false)).with_empty_span() )), })], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1670,7 +1676,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - ms().verified_stmt(&format!("USE {}", object_name)), + ms().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -1678,7 +1684,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - ms().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), + ms().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -2184,7 +2190,7 @@ fn parse_mssql_if_else() { "IF 1 = 1 BEGIN SET @A = 1; END ELSE SET @A = 2;" ); } - _ => panic!("Unexpected statements: {:?}", stmts), + _ => panic!("Unexpected statements: {stmts:?}"), } } @@ -2234,7 +2240,7 @@ fn test_mssql_if_statements_span() { Span::new(Location::new(1, 21), Location::new(1, 36)) ); } - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Blocks @@ -2255,7 +2261,7 @@ fn test_mssql_if_statements_span() { Span::new(Location::new(1, 32), Location::new(1, 57)) ); } - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } } @@ -2321,6 +2327,18 @@ fn ms() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {})]) } +// MS SQL dialect with support for optional semi-colon statement delimiters +fn tsql() -> TestedDialects { + TestedDialects::new_with_options( + vec![Box::new(MsSqlDialect {})], + ParserOptions { + trailing_commas: false, + unescape: true, + require_semicolon_stmt_delimiter: false, + }, + ) +} + fn ms_and_generic() -> TestedDialects { TestedDialects::new(vec![Box::new(MsSqlDialect {}), Box::new(GenericDialect {})]) } @@ -2477,3 +2495,15 @@ fn parse_mssql_grant() { fn parse_mssql_deny() { ms().verified_stmt("DENY SELECT ON my_table TO public, db_admin"); } + +#[test] +fn test_tsql_no_semicolon_delimiter() { + let sql = r#" +DECLARE @X AS NVARCHAR(MAX)='x' +DECLARE @Y AS NVARCHAR(MAX)='y' + "#; + + let stmts = tsql().parse_sql_statements(sql).unwrap(); + assert_eq!(stmts.len(), 2); + assert!(stmts.iter().all(|s| matches!(s, Statement::Declare { .. }))); +} diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs index 540348ff..9068ed9c 100644 --- a/tests/sqlparser_mysql.rs +++ b/tests/sqlparser_mysql.rs @@ -593,7 +593,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - mysql_and_generic().verified_stmt(&format!("USE {}", object_name)), + mysql_and_generic().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -601,8 +601,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - mysql_and_generic() - .verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), + mysql_and_generic().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -670,6 +669,20 @@ fn table_constraint_unique_primary_ctor( characteristics: Option, unique_index_type_display: Option, ) -> TableConstraint { + let columns = columns + .into_iter() + .map(|ident| IndexColumn { + column: OrderByExpr { + expr: Expr::Identifier(ident), + options: OrderByOptions { + asc: None, + nulls_first: None, + }, + with_fill: None, + }, + operator_class: None, + }) + .collect(); match unique_index_type_display { Some(index_type_display) => TableConstraint::Unique { name, @@ -795,6 +808,67 @@ fn parse_create_table_primary_and_unique_key_with_index_options() { } } +#[test] +fn parse_prefix_key_part() { + let expected = vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::value( + number("10"), + )))]; + for sql in [ + "CREATE INDEX idx_index ON t(textcol(10))", + "ALTER TABLE tab ADD INDEX idx_index (textcol(10))", + "ALTER TABLE tab ADD PRIMARY KEY (textcol(10))", + "ALTER TABLE tab ADD UNIQUE KEY (textcol(10))", + "ALTER TABLE tab ADD UNIQUE KEY (textcol(10))", + "ALTER TABLE tab ADD FULLTEXT INDEX (textcol(10))", + "CREATE TABLE t (textcol TEXT, INDEX idx_index (textcol(10)))", + ] { + match index_column(mysql_and_generic().verified_stmt(sql)) { + Expr::Function(Function { + name, + args: FunctionArguments::List(FunctionArgumentList { args, .. }), + .. + }) => { + assert_eq!(name.to_string(), "textcol"); + assert_eq!(args, expected); + } + expr => panic!("unexpected expression {expr} for {sql}"), + } + } +} + +#[test] +fn test_functional_key_part() { + assert_eq!( + index_column( + mysql_and_generic() + .verified_stmt("CREATE INDEX idx_index ON t((col COLLATE utf8mb4_bin) DESC)") + ), + Expr::Nested(Box::new(Expr::Collate { + expr: Box::new(Expr::Identifier("col".into())), + collation: ObjectName(vec![sqlparser::ast::ObjectNamePart::Identifier( + Ident::new("utf8mb4_bin") + )]), + })) + ); + assert_eq!( + index_column(mysql_and_generic().verified_stmt( + r#"CREATE TABLE t (jsoncol JSON, PRIMARY KEY ((CAST(col ->> '$.id' AS UNSIGNED)) ASC))"# + )), + Expr::Nested(Box::new(Expr::Cast { + kind: CastKind::Cast, + expr: Box::new(Expr::BinaryOp { + left: Box::new(Expr::Identifier(Ident::new("col"))), + op: BinaryOperator::LongArrow, + right: Box::new(Expr::Value( + Value::SingleQuotedString("$.id".to_string()).with_empty_span() + )), + }), + data_type: DataType::Unsigned, + format: None, + })), + ); +} + #[test] fn parse_create_table_primary_and_unique_key_with_index_type() { let sqls = ["UNIQUE", "PRIMARY KEY"].map(|key_ty| { @@ -1329,6 +1403,7 @@ fn parse_escaped_quote_identifiers_with_escape() { quote_style: Some('`'), span: Span::empty(), }))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1367,6 +1442,7 @@ fn parse_escaped_quote_identifiers_with_no_escape() { ParserOptions { trailing_commas: false, unescape: false, + require_semicolon_stmt_delimiter: true, } ) .verified_stmt(sql), @@ -1382,6 +1458,7 @@ fn parse_escaped_quote_identifiers_with_no_escape() { quote_style: Some('`'), span: Span::empty(), }))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1429,6 +1506,7 @@ fn parse_escaped_backticks_with_escape() { quote_style: Some('`'), span: Span::empty(), }))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1480,6 +1558,7 @@ fn parse_escaped_backticks_with_no_escape() { quote_style: Some('`'), span: Span::empty(), }))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -2151,6 +2230,7 @@ fn parse_select_with_numeric_prefix_column_name() { projection: vec![SelectItem::UnnamedExpr(Expr::Identifier(Ident::new( "123col_$@123abc" )))], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::with_quote( @@ -2188,11 +2268,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { assert_eq!(&[Ident::new("t"), Ident::new("15to29")], &parts[..]); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 2: Qualified column name that starts with digits and on its own represents a number. @@ -2202,11 +2282,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { Some(SelectItem::UnnamedExpr(Expr::CompoundIdentifier(parts))) => { assert_eq!(&[Ident::new("t"), Ident::new("15e29")], &parts[..]); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 3: Unqualified, the same token is parsed as a number. @@ -2220,11 +2300,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { Some(SelectItem::UnnamedExpr(Expr::Value(ValueWithSpan { value, .. }))) => { assert_eq!(&number("15e29"), value); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 4: Quoted simple identifier. @@ -2234,11 +2314,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { Some(SelectItem::UnnamedExpr(Expr::Identifier(name))) => { assert_eq!(&Ident::with_quote('`', "15e29"), name); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 5: Quoted compound identifier. @@ -2251,11 +2331,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { &parts[..] ); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 6: Multi-level compound identifiers. @@ -2272,11 +2352,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { &parts[..] ); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } // Case 7: Multi-level compound quoted identifiers. @@ -2293,11 +2373,11 @@ fn parse_qualified_identifiers_with_numeric_prefix() { &parts[..] ); } - proj => panic!("Unexpected projection: {:?}", proj), + proj => panic!("Unexpected projection: {proj:?}"), }, - body => panic!("Unexpected statement body: {:?}", body), + body => panic!("Unexpected statement body: {body:?}"), }, - stmt => panic!("Unexpected statement: {:?}", stmt), + stmt => panic!("Unexpected statement: {stmt:?}"), } } @@ -2318,7 +2398,6 @@ fn parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column() { q.body, Box::new(SetExpr::Select(Box::new(Select { select_token: AttachedToken::empty(), - distinct: None, top: None, top_before_distinct: false, @@ -2326,6 +2405,7 @@ fn parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column() { SelectItem::UnnamedExpr(Expr::value(number("123e4"))), SelectItem::UnnamedExpr(Expr::Identifier(Ident::new("123col_$@123abc"))) ], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident::with_quote( @@ -2802,7 +2882,7 @@ fn parse_alter_table_with_algorithm() { vec![ AlterTableOperation::DropColumn { has_column_keyword: true, - column_name: Ident::new("password_digest"), + column_names: vec![Ident::new("password_digest")], if_exists: false, drop_behavior: None, }, @@ -2850,7 +2930,7 @@ fn parse_alter_table_with_lock() { vec![ AlterTableOperation::DropColumn { has_column_keyword: true, - column_name: Ident::new("password_digest"), + column_names: vec![Ident::new("password_digest")], if_exists: false, drop_behavior: None, }, @@ -2969,6 +3049,7 @@ fn parse_substring_in_select() { special: true, shorthand: false, })], + exclude: None, into: None, from: vec![TableWithJoins { relation: table_from_name(ObjectName::from(vec![Ident { @@ -3283,6 +3364,7 @@ fn parse_hex_string_introducer() { ) .into(), })], + exclude: None, from: vec![], lateral_views: vec![], prewhere: None, @@ -3542,6 +3624,7 @@ fn parse_grant() { with_grant_option, as_grantor: _, granted_by, + current_grants: _, } = stmt { assert_eq!( @@ -4035,3 +4118,28 @@ fn parse_alter_table_drop_index() { AlterTableOperation::DropIndex { name } if name.value == "idx_index" ); } + +#[test] +fn parse_json_member_of() { + mysql().verified_stmt(r#"SELECT 17 MEMBER OF('[23, "abc", 17, "ab", 10]')"#); + let sql = r#"SELECT 'ab' MEMBER OF('[23, "abc", 17, "ab", 10]')"#; + let stmt = mysql().verified_stmt(sql); + match stmt { + Statement::Query(query) => { + let select = query.body.as_select().unwrap(); + assert_eq!( + select.projection, + vec![SelectItem::UnnamedExpr(Expr::MemberOf(MemberOf { + value: Box::new(Expr::Value( + Value::SingleQuotedString("ab".to_string()).into() + )), + array: Box::new(Expr::Value( + Value::SingleQuotedString(r#"[23, "abc", 17, "ab", 10]"#.to_string()) + .into() + )), + }))] + ); + } + _ => panic!("Unexpected statement {stmt}"), + } +} diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs index 6f0ba9c6..d163096c 100644 --- a/tests/sqlparser_postgres.rs +++ b/tests/sqlparser_postgres.rs @@ -606,9 +606,10 @@ fn parse_alter_table_constraints_unique_nulls_distinct() { .verified_stmt("ALTER TABLE t ADD CONSTRAINT b UNIQUE NULLS NOT DISTINCT (c)") { Statement::AlterTable { operations, .. } => match &operations[0] { - AlterTableOperation::AddConstraint(TableConstraint::Unique { - nulls_distinct, .. - }) => { + AlterTableOperation::AddConstraint { + constraint: TableConstraint::Unique { nulls_distinct, .. }, + .. + } => { assert_eq!(nulls_distinct, &NullsDistinctOption::NotDistinct) } _ => unreachable!(), @@ -764,10 +765,7 @@ fn parse_drop_extension() { #[test] fn parse_alter_table_alter_column() { - pg().one_statement_parses_to( - "ALTER TABLE tab ALTER COLUMN is_active TYPE TEXT USING 'text'", - "ALTER TABLE tab ALTER COLUMN is_active SET DATA TYPE TEXT USING 'text'", - ); + pg().verified_stmt("ALTER TABLE tab ALTER COLUMN is_active TYPE TEXT USING 'text'"); match alter_table_op( pg().verified_stmt( @@ -783,6 +781,7 @@ fn parse_alter_table_alter_column() { AlterColumnOperation::SetDataType { data_type: DataType::Text, using: Some(using_expr), + had_set: true, } ); } @@ -1306,6 +1305,7 @@ fn parse_copy_to() { }, } ], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -1666,7 +1666,9 @@ fn parse_execute() { has_parentheses: false, using: vec![], immediate: false, - into: vec![] + into: vec![], + output: false, + default: false, } ); @@ -1682,7 +1684,9 @@ fn parse_execute() { has_parentheses: true, using: vec![], immediate: false, - into: vec![] + into: vec![], + output: false, + default: false, } ); @@ -1719,7 +1723,9 @@ fn parse_execute() { }, ], immediate: false, - into: vec![] + into: vec![], + output: false, + default: false, } ); } @@ -2181,21 +2187,39 @@ fn parse_pg_regex_match_ops() { ("!~*", BinaryOperator::PGRegexNotIMatch), ]; + // Match against a single value for (str_op, op) in pg_regex_match_ops { - let select = pg().verified_only_select(&format!("SELECT 'abc' {} '^a'", &str_op)); + let select = pg().verified_only_select(&format!("SELECT 'abc' {str_op} '^a'")); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value( - (Value::SingleQuotedString("abc".into())).with_empty_span() - )), + left: Box::new(Expr::Value(single_quoted_string("abc").with_empty_span(),)), op: op.clone(), - right: Box::new(Expr::Value( - (Value::SingleQuotedString("^a".into())).with_empty_span() - )), + right: Box::new(Expr::Value(single_quoted_string("^a").with_empty_span(),)), }), select.projection[0] ); } + + // Match against any value from an array + for (str_op, op) in pg_regex_match_ops { + let select = + pg().verified_only_select(&format!("SELECT 'abc' {str_op} ANY(ARRAY['^a', 'x'])")); + assert_eq!( + SelectItem::UnnamedExpr(Expr::AnyOp { + left: Box::new(Expr::Value(single_quoted_string("abc").with_empty_span(),)), + compare_op: op.clone(), + right: Box::new(Expr::Array(Array { + elem: vec![ + Expr::Value(single_quoted_string("^a").with_empty_span()), + Expr::Value(single_quoted_string("x").with_empty_span()), + ], + named: true, + })), + is_some: false, + }), + select.projection[0] + ) + } } #[test] @@ -2207,21 +2231,35 @@ fn parse_pg_like_match_ops() { ("!~~*", BinaryOperator::PGNotILikeMatch), ]; + // Match against a single value for (str_op, op) in pg_like_match_ops { - let select = pg().verified_only_select(&format!("SELECT 'abc' {} 'a_c%'", &str_op)); + let select = pg().verified_only_select(&format!("SELECT 'abc' {str_op} 'a_c%'")); assert_eq!( SelectItem::UnnamedExpr(Expr::BinaryOp { - left: Box::new(Expr::Value( - (Value::SingleQuotedString("abc".into())).with_empty_span() - )), + left: Box::new(Expr::Value(single_quoted_string("abc").with_empty_span(),)), op: op.clone(), - right: Box::new(Expr::Value( - (Value::SingleQuotedString("a_c%".into())).with_empty_span() - )), + right: Box::new(Expr::Value(single_quoted_string("a_c%").with_empty_span(),)), }), select.projection[0] ); } + + // Match against all values from an array + for (str_op, op) in pg_like_match_ops { + let select = + pg().verified_only_select(&format!("SELECT 'abc' {str_op} ALL(ARRAY['a_c%'])")); + assert_eq!( + SelectItem::UnnamedExpr(Expr::AllOp { + left: Box::new(Expr::Value(single_quoted_string("abc").with_empty_span(),)), + compare_op: op.clone(), + right: Box::new(Expr::Array(Array { + elem: vec![Expr::Value(single_quoted_string("a_c%").with_empty_span())], + named: true, + })), + }), + select.projection[0] + ) + } } #[test] @@ -2535,12 +2573,12 @@ fn parse_create_indices_with_operator_classes() { for expected_operator_class in &operator_classes { let single_column_sql_statement = format!( "CREATE INDEX the_index_name ON users USING {expected_index_type} (concat_users_name(first_name, last_name){})", - expected_operator_class.as_ref().map(|oc| format!(" {}", oc)) + expected_operator_class.as_ref().map(|oc| format!(" {oc}")) .unwrap_or_default() ); let multi_column_sql_statement = format!( "CREATE INDEX the_index_name ON users USING {expected_index_type} (column_name,concat_users_name(first_name, last_name){})", - expected_operator_class.as_ref().map(|oc| format!(" {}", oc)) + expected_operator_class.as_ref().map(|oc| format!(" {oc}")) .unwrap_or_default() ); @@ -2949,6 +2987,7 @@ fn parse_array_subquery_expr() { projection: vec![SelectItem::UnnamedExpr(Expr::Value( (number("1")).with_empty_span() ))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -2974,6 +3013,7 @@ fn parse_array_subquery_expr() { projection: vec![SelectItem::UnnamedExpr(Expr::Value( (number("2")).with_empty_span() ))], + exclude: None, into: None, from: vec![], lateral_views: vec![], @@ -3273,7 +3313,7 @@ fn test_fn_arg_with_value_operator() { assert!(matches!( &args[..], &[FunctionArg::ExprNamed { operator: FunctionArgOperator::Value, .. }] - ), "Invalid function argument: {:?}", args); + ), "Invalid function argument: {args:?}"); } other => panic!("Expected: JSON_OBJECT('name' VALUE 'value') to be parsed as a function, but got {other:?}"), } @@ -5258,7 +5298,10 @@ fn parse_at_time_zone() { left: Box::new(Expr::AtTimeZone { timestamp: Box::new(Expr::TypedString { data_type: DataType::Timestamp(None, TimezoneInfo::None), - value: Value::SingleQuotedString("2001-09-28 01:00".to_string()), + value: ValueWithSpan { + value: Value::SingleQuotedString("2001-09-28 01:00".to_string()), + span: Span::empty(), + }, }), time_zone: Box::new(Expr::Cast { kind: CastKind::DoubleColon, @@ -5679,7 +5722,7 @@ fn parse_drop_trigger() { "DROP TRIGGER{} check_update ON table_name{}", if if_exists { " IF EXISTS" } else { "" }, option - .map(|o| format!(" {}", o)) + .map(|o| format!(" {o}")) .unwrap_or_else(|| "".to_string()) ); assert_eq!( @@ -5773,8 +5816,7 @@ fn parse_trigger_related_functions() { // Now we parse the statements and check if they are parsed correctly. let mut statements = pg() .parse_sql_statements(&format!( - "{}{}{}{}", - sql_table_creation, sql_create_function, sql_create_trigger, sql_drop_trigger + "{sql_table_creation}{sql_create_function}{sql_create_trigger}{sql_drop_trigger}" )) .unwrap(); @@ -6201,3 +6243,153 @@ fn parse_alter_table_replica_identity() { _ => unreachable!(), } } + +#[test] +fn parse_ts_datatypes() { + match pg_and_generic().verified_stmt("CREATE TABLE foo (x TSVECTOR)") { + Statement::CreateTable(CreateTable { columns, .. }) => { + assert_eq!( + columns, + vec![ColumnDef { + name: "x".into(), + data_type: DataType::TsVector, + options: vec![], + }] + ); + } + _ => unreachable!(), + } + + match pg_and_generic().verified_stmt("CREATE TABLE foo (x TSQUERY)") { + Statement::CreateTable(CreateTable { columns, .. }) => { + assert_eq!( + columns, + vec![ColumnDef { + name: "x".into(), + data_type: DataType::TsQuery, + options: vec![], + }] + ); + } + _ => unreachable!(), + } +} + +#[test] +fn parse_alter_table_constraint_not_valid() { + match pg_and_generic().verified_stmt( + "ALTER TABLE foo ADD CONSTRAINT bar FOREIGN KEY (baz) REFERENCES other(ref) NOT VALID", + ) { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![AlterTableOperation::AddConstraint { + constraint: TableConstraint::ForeignKey { + name: Some("bar".into()), + index_name: None, + columns: vec!["baz".into()], + foreign_table: ObjectName::from(vec!["other".into()]), + referred_columns: vec!["ref".into()], + on_delete: None, + on_update: None, + characteristics: None, + }, + not_valid: true, + }] + ); + } + _ => unreachable!(), + } +} + +#[test] +fn parse_alter_table_validate_constraint() { + match pg_and_generic().verified_stmt("ALTER TABLE foo VALIDATE CONSTRAINT bar") { + Statement::AlterTable { operations, .. } => { + assert_eq!( + operations, + vec![AlterTableOperation::ValidateConstraint { name: "bar".into() }] + ); + } + _ => unreachable!(), + } +} + +#[test] +fn parse_create_server() { + let test_cases = vec![ + ( + "CREATE SERVER myserver FOREIGN DATA WRAPPER postgres_fdw", + CreateServerStatement { + name: ObjectName::from(vec!["myserver".into()]), + if_not_exists: false, + server_type: None, + version: None, + foreign_data_wrapper: ObjectName::from(vec!["postgres_fdw".into()]), + options: None, + }, + ), + ( + "CREATE SERVER IF NOT EXISTS myserver TYPE 'server_type' VERSION 'server_version' FOREIGN DATA WRAPPER postgres_fdw", + CreateServerStatement { + name: ObjectName::from(vec!["myserver".into()]), + if_not_exists: true, + server_type: Some(Ident { + value: "server_type".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }), + version: Some(Ident { + value: "server_version".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }), + foreign_data_wrapper: ObjectName::from(vec!["postgres_fdw".into()]), + options: None, + } + ), + ( + "CREATE SERVER myserver2 FOREIGN DATA WRAPPER postgres_fdw OPTIONS (host 'foo', dbname 'foodb', port '5432')", + CreateServerStatement { + name: ObjectName::from(vec!["myserver2".into()]), + if_not_exists: false, + server_type: None, + version: None, + foreign_data_wrapper: ObjectName::from(vec!["postgres_fdw".into()]), + options: Some(vec![ + CreateServerOption { + key: "host".into(), + value: Ident { + value: "foo".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }, + }, + CreateServerOption { + key: "dbname".into(), + value: Ident { + value: "foodb".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }, + }, + CreateServerOption { + key: "port".into(), + value: Ident { + value: "5432".to_string(), + quote_style: Some('\''), + span: Span::empty(), + }, + }, + ]), + } + ) + ]; + + for (sql, expected) in test_cases { + let Statement::CreateServer(stmt) = pg_and_generic().verified_stmt(sql) else { + unreachable!() + }; + assert_eq!(stmt, expected); + } +} diff --git a/tests/sqlparser_redshift.rs b/tests/sqlparser_redshift.rs index be2b6722..d539adf6 100644 --- a/tests/sqlparser_redshift.rs +++ b/tests/sqlparser_redshift.rs @@ -402,3 +402,8 @@ fn parse_extract_single_quotes() { fn parse_string_literal_backslash_escape() { redshift().one_statement_parses_to(r#"SELECT 'l\'auto'"#, "SELECT 'l''auto'"); } + +#[test] +fn parse_utf8_multibyte_idents() { + redshift().verified_stmt("SELECT 🚀.city AS 🎸 FROM customers AS 🚀"); +} diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs index 7d734ca2..a7a63315 100644 --- a/tests/sqlparser_snowflake.rs +++ b/tests/sqlparser_snowflake.rs @@ -270,8 +270,8 @@ fn test_snowflake_create_table_with_tag() { assert_eq!("my_table", name.to_string()); assert_eq!( Some(vec![ - Tag::new("A".into(), "TAG A".to_string()), - Tag::new("B".into(), "TAG B".to_string()) + Tag::new(ObjectName::from(vec![Ident::new("A")]), "TAG A".to_string()), + Tag::new(ObjectName::from(vec![Ident::new("B")]), "TAG B".to_string()) ]), with_tags ); @@ -291,8 +291,8 @@ fn test_snowflake_create_table_with_tag() { assert_eq!("my_table", name.to_string()); assert_eq!( Some(vec![ - Tag::new("A".into(), "TAG A".to_string()), - Tag::new("B".into(), "TAG B".to_string()) + Tag::new(ObjectName::from(vec![Ident::new("A")]), "TAG A".to_string()), + Tag::new(ObjectName::from(vec![Ident::new("B")]), "TAG B".to_string()) ]), with_tags ); @@ -731,7 +731,7 @@ fn test_snowflake_create_table_with_columns_masking_policy() { option: ColumnOption::Policy(ColumnPolicy::MaskingPolicy( ColumnPolicyProperty { with, - policy_name: "p".into(), + policy_name: ObjectName::from(vec![Ident::new("p")]), using_columns, } )) @@ -765,7 +765,7 @@ fn test_snowflake_create_table_with_columns_projection_policy() { option: ColumnOption::Policy(ColumnPolicy::ProjectionPolicy( ColumnPolicyProperty { with, - policy_name: "p".into(), + policy_name: ObjectName::from(vec![Ident::new("p")]), using_columns: None, } )) @@ -802,8 +802,14 @@ fn test_snowflake_create_table_with_columns_tags() { option: ColumnOption::Tags(TagsColumnOption { with, tags: vec![ - Tag::new("A".into(), "TAG A".into()), - Tag::new("B".into(), "TAG B".into()), + Tag::new( + ObjectName::from(vec![Ident::new("A")]), + "TAG A".into() + ), + Tag::new( + ObjectName::from(vec![Ident::new("B")]), + "TAG B".into() + ), ] }), }], @@ -846,7 +852,7 @@ fn test_snowflake_create_table_with_several_column_options() { option: ColumnOption::Policy(ColumnPolicy::MaskingPolicy( ColumnPolicyProperty { with: true, - policy_name: "p1".into(), + policy_name: ObjectName::from(vec![Ident::new("p1")]), using_columns: Some(vec!["a".into(), "b".into()]), } )), @@ -856,8 +862,14 @@ fn test_snowflake_create_table_with_several_column_options() { option: ColumnOption::Tags(TagsColumnOption { with: true, tags: vec![ - Tag::new("A".into(), "TAG A".into()), - Tag::new("B".into(), "TAG B".into()), + Tag::new( + ObjectName::from(vec![Ident::new("A")]), + "TAG A".into() + ), + Tag::new( + ObjectName::from(vec![Ident::new("B")]), + "TAG B".into() + ), ] }), } @@ -878,7 +890,7 @@ fn test_snowflake_create_table_with_several_column_options() { option: ColumnOption::Policy(ColumnPolicy::ProjectionPolicy( ColumnPolicyProperty { with: false, - policy_name: "p2".into(), + policy_name: ObjectName::from(vec![Ident::new("p2")]), using_columns: None, } )), @@ -888,8 +900,14 @@ fn test_snowflake_create_table_with_several_column_options() { option: ColumnOption::Tags(TagsColumnOption { with: false, tags: vec![ - Tag::new("C".into(), "TAG C".into()), - Tag::new("D".into(), "TAG D".into()), + Tag::new( + ObjectName::from(vec![Ident::new("C")]), + "TAG C".into() + ), + Tag::new( + ObjectName::from(vec![Ident::new("D")]), + "TAG D".into() + ), ] }), } @@ -942,8 +960,8 @@ fn test_snowflake_create_iceberg_table_all_options() { with_aggregation_policy.map(|name| name.to_string()) ); assert_eq!(Some(vec![ - Tag::new("A".into(), "TAG A".into()), - Tag::new("B".into(), "TAG B".into()), + Tag::new(ObjectName::from(vec![Ident::new("A")]), "TAG A".into()), + Tag::new(ObjectName::from(vec![Ident::new("B")]), "TAG B".into()), ]), with_tags); } @@ -977,6 +995,51 @@ fn test_snowflake_create_iceberg_table_without_location() { ); } +#[test] +fn test_snowflake_create_table_trailing_options() { + // Serialization to SQL assume that in `CREATE TABLE AS` the options come before the `AS ()` + // but Snowflake supports also the other way around + snowflake() + .verified_stmt("CREATE TEMPORARY TABLE dst ON COMMIT PRESERVE ROWS AS (SELECT * FROM src)"); + snowflake() + .parse_sql_statements( + "CREATE TEMPORARY TABLE dst AS (SELECT * FROM src) ON COMMIT PRESERVE ROWS", + ) + .unwrap(); + + // Same for `CREATE TABLE LIKE|CLONE`: + snowflake().verified_stmt("CREATE TEMPORARY TABLE dst LIKE src ON COMMIT PRESERVE ROWS"); + snowflake() + .parse_sql_statements("CREATE TEMPORARY TABLE dst ON COMMIT PRESERVE ROWS LIKE src") + .unwrap(); + + snowflake().verified_stmt("CREATE TEMPORARY TABLE dst CLONE src ON COMMIT PRESERVE ROWS"); + snowflake() + .parse_sql_statements("CREATE TEMPORARY TABLE dst ON COMMIT PRESERVE ROWS CLONE src") + .unwrap(); +} + +#[test] +fn test_snowflake_create_table_valid_schema_info() { + // Validate there's exactly one source of information on the schema of the new table + assert_eq!( + snowflake() + .parse_sql_statements("CREATE TABLE dst") + .is_err(), + true + ); + assert_eq!( + snowflake().parse_sql_statements("CREATE OR REPLACE TEMP TABLE dst LIKE src AS (SELECT * FROM CUSTOMERS) ON COMMIT PRESERVE ROWS").is_err(), + true + ); + assert_eq!( + snowflake() + .parse_sql_statements("CREATE OR REPLACE TEMP TABLE dst CLONE customers LIKE customer2") + .is_err(), + true + ); +} + #[test] fn parse_sf_create_or_replace_view_with_comment_missing_equal() { assert!(snowflake_and_generic() @@ -2510,10 +2573,7 @@ fn test_snowflake_stage_object_names_into_location() { .zip(allowed_object_names.iter_mut()) { let (formatted_name, object_name) = it; - let sql = format!( - "COPY INTO {} FROM 'gcs://mybucket/./../a.csv'", - formatted_name - ); + let sql = format!("COPY INTO {formatted_name} FROM 'gcs://mybucket/./../a.csv'"); match snowflake().verified_stmt(&sql) { Statement::CopyIntoSnowflake { into, .. } => { assert_eq!(into.0, object_name.0) @@ -2536,10 +2596,7 @@ fn test_snowflake_stage_object_names_into_table() { .zip(allowed_object_names.iter_mut()) { let (formatted_name, object_name) = it; - let sql = format!( - "COPY INTO {} FROM 'gcs://mybucket/./../a.csv'", - formatted_name - ); + let sql = format!("COPY INTO {formatted_name} FROM 'gcs://mybucket/./../a.csv'"); match snowflake().verified_stmt(&sql) { Statement::CopyIntoSnowflake { into, .. } => { assert_eq!(into.0, object_name.0) @@ -2569,6 +2626,26 @@ fn test_snowflake_copy_into() { } _ => unreachable!(), } + + // Test for non-ident characters in stage names + let sql = "COPY INTO a.b FROM @namespace.stage_name/x@x~x%x+"; + assert_eq!(snowflake().verified_stmt(sql).to_string(), sql); + match snowflake().verified_stmt(sql) { + Statement::CopyIntoSnowflake { into, from_obj, .. } => { + assert_eq!( + into, + ObjectName::from(vec![Ident::new("a"), Ident::new("b")]) + ); + assert_eq!( + from_obj, + Some(ObjectName::from(vec![ + Ident::new("@namespace"), + Ident::new("stage_name/x@x~x%x+") + ])) + ) + } + _ => unreachable!(), + } } #[test] @@ -3020,7 +3097,7 @@ fn parse_use() { for object_name in &valid_object_names { // Test single identifier without quotes assert_eq!( - snowflake().verified_stmt(&format!("USE {}", object_name)), + snowflake().verified_stmt(&format!("USE {object_name}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::new( object_name.to_string() )]))) @@ -3028,7 +3105,7 @@ fn parse_use() { for "e in "e_styles { // Test single identifier with different type of quotes assert_eq!( - snowflake().verified_stmt(&format!("USE {}{}{}", quote, object_name, quote)), + snowflake().verified_stmt(&format!("USE {quote}{object_name}{quote}")), Statement::Use(Use::Object(ObjectName::from(vec![Ident::with_quote( quote, object_name.to_string(), @@ -3040,7 +3117,9 @@ fn parse_use() { for "e in "e_styles { // Test double identifier with different type of quotes assert_eq!( - snowflake().verified_stmt(&format!("USE {0}CATALOG{0}.{0}my_schema{0}", quote)), + snowflake().verified_stmt(&format!( + "USE {quote}CATALOG{quote}.{quote}my_schema{quote}" + )), Statement::Use(Use::Object(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") @@ -3059,35 +3138,37 @@ fn parse_use() { for "e in "e_styles { // Test single and double identifier with keyword and different type of quotes assert_eq!( - snowflake().verified_stmt(&format!("USE DATABASE {0}my_database{0}", quote)), + snowflake().verified_stmt(&format!("USE DATABASE {quote}my_database{quote}")), Statement::Use(Use::Database(ObjectName::from(vec![Ident::with_quote( quote, "my_database".to_string(), )]))) ); assert_eq!( - snowflake().verified_stmt(&format!("USE SCHEMA {0}my_schema{0}", quote)), + snowflake().verified_stmt(&format!("USE SCHEMA {quote}my_schema{quote}")), Statement::Use(Use::Schema(ObjectName::from(vec![Ident::with_quote( quote, "my_schema".to_string(), )]))) ); assert_eq!( - snowflake().verified_stmt(&format!("USE SCHEMA {0}CATALOG{0}.{0}my_schema{0}", quote)), + snowflake().verified_stmt(&format!( + "USE SCHEMA {quote}CATALOG{quote}.{quote}my_schema{quote}" + )), Statement::Use(Use::Schema(ObjectName::from(vec![ Ident::with_quote(quote, "CATALOG"), Ident::with_quote(quote, "my_schema") ]))) ); assert_eq!( - snowflake().verified_stmt(&format!("USE ROLE {0}my_role{0}", quote)), + snowflake().verified_stmt(&format!("USE ROLE {quote}my_role{quote}")), Statement::Use(Use::Role(ObjectName::from(vec![Ident::with_quote( quote, "my_role".to_string(), )]))) ); assert_eq!( - snowflake().verified_stmt(&format!("USE WAREHOUSE {0}my_wh{0}", quote)), + snowflake().verified_stmt(&format!("USE WAREHOUSE {quote}my_wh{quote}")), Statement::Use(Use::Warehouse(ObjectName::from(vec![Ident::with_quote( quote, "my_wh".to_string(), @@ -3124,7 +3205,7 @@ fn view_comment_option_should_be_after_column_list() { "CREATE OR REPLACE VIEW v (a COMMENT 'a comment', b, c COMMENT 'c comment') COMMENT = 'Comment' AS SELECT a FROM t", "CREATE OR REPLACE VIEW v (a COMMENT 'a comment', b, c COMMENT 'c comment') WITH (foo = bar) COMMENT = 'Comment' AS SELECT a FROM t", ] { - snowflake_and_generic() + snowflake() .verified_stmt(sql); } } @@ -3133,7 +3214,7 @@ fn view_comment_option_should_be_after_column_list() { fn parse_view_column_descriptions() { let sql = "CREATE OR REPLACE VIEW v (a COMMENT 'Comment', b) AS SELECT a, b FROM table1"; - match snowflake_and_generic().verified_stmt(sql) { + match snowflake().verified_stmt(sql) { Statement::CreateView { name, columns, .. } => { assert_eq!(name.to_string(), "v"); assert_eq!( @@ -3142,7 +3223,9 @@ fn parse_view_column_descriptions() { ViewColumnDef { name: Ident::new("a"), data_type: None, - options: Some(vec![ColumnOption::Comment("Comment".to_string())]), + options: Some(ColumnOptions::SpaceSeparated(vec![ColumnOption::Comment( + "Comment".to_string() + )])), }, ViewColumnDef { name: Ident::new("b"), @@ -3397,10 +3480,38 @@ fn parse_ls_and_rm() { .unwrap(); } +#[test] +fn test_sql_keywords_as_select_item_ident() { + // Some keywords that should be parsed as an alias + let unreserved_kws = vec!["CLUSTER", "FETCH", "RETURNING", "LIMIT", "EXCEPT", "SORT"]; + for kw in unreserved_kws { + snowflake().verified_stmt(&format!("SELECT 1, {kw}")); + } + + // Some keywords that should not be parsed as an alias + let reserved_kws = vec![ + "FROM", + "GROUP", + "HAVING", + "INTERSECT", + "INTO", + "ORDER", + "SELECT", + "UNION", + "WHERE", + "WITH", + ]; + for kw in reserved_kws { + assert!(snowflake() + .parse_sql_statements(&format!("SELECT 1, {kw}")) + .is_err()); + } +} + #[test] fn test_sql_keywords_as_select_item_aliases() { // Some keywords that should be parsed as an alias - let unreserved_kws = vec!["CLUSTER", "FETCH", "RETURNING", "LIMIT", "EXCEPT"]; + let unreserved_kws = vec!["CLUSTER", "FETCH", "RETURNING", "LIMIT", "EXCEPT", "SORT"]; for kw in unreserved_kws { snowflake() .one_statement_parses_to(&format!("SELECT 1 {kw}"), &format!("SELECT 1 AS {kw}")); @@ -3424,6 +3535,75 @@ fn test_sql_keywords_as_select_item_aliases() { .parse_sql_statements(&format!("SELECT 1 {kw}")) .is_err()); } + + // LIMIT is alias + snowflake().one_statement_parses_to("SELECT 1 LIMIT", "SELECT 1 AS LIMIT"); + // LIMIT is not an alias + snowflake().verified_stmt("SELECT 1 LIMIT 1"); + snowflake().verified_stmt("SELECT 1 LIMIT $1"); + snowflake().verified_stmt("SELECT 1 LIMIT ''"); + snowflake().verified_stmt("SELECT 1 LIMIT NULL"); + snowflake().verified_stmt("SELECT 1 LIMIT $$$$"); +} + +#[test] +fn test_sql_keywords_as_table_aliases() { + // Some keywords that should be parsed as an alias implicitly + let unreserved_kws = vec![ + "VIEW", + "EXPLAIN", + "ANALYZE", + "SORT", + "PIVOT", + "UNPIVOT", + "TOP", + "LIMIT", + "OFFSET", + "FETCH", + "EXCEPT", + "CLUSTER", + "DISTRIBUTE", + "GLOBAL", + "ANTI", + "SEMI", + "RETURNING", + "OUTER", + "WINDOW", + "END", + "PARTITION", + "PREWHERE", + "SETTINGS", + "FORMAT", + "MATCH_RECOGNIZE", + "OPEN", + ]; + + for kw in unreserved_kws { + snowflake().verified_stmt(&format!("SELECT * FROM tbl AS {kw}")); + snowflake().one_statement_parses_to( + &format!("SELECT * FROM tbl {kw}"), + &format!("SELECT * FROM tbl AS {kw}"), + ); + } + + // Some keywords that should not be parsed as an alias implicitly + let reserved_kws = vec![ + "FROM", "GROUP", "HAVING", "ORDER", "SELECT", "UNION", "WHERE", "WITH", + ]; + for kw in reserved_kws { + assert!(snowflake() + .parse_sql_statements(&format!("SELECT * FROM tbl {kw}")) + .is_err()); + } + + // LIMIT is alias + snowflake().one_statement_parses_to("SELECT * FROM tbl LIMIT", "SELECT * FROM tbl AS LIMIT"); + // LIMIT is not an alias + snowflake().verified_stmt("SELECT * FROM tbl LIMIT 1"); + snowflake().verified_stmt("SELECT * FROM tbl LIMIT $1"); + snowflake().verified_stmt("SELECT * FROM tbl LIMIT ''"); + snowflake().verified_stmt("SELECT * FROM tbl LIMIT NULL"); + snowflake().verified_stmt("SELECT * FROM tbl LIMIT $$$$"); } #[test] @@ -3627,7 +3807,7 @@ fn test_alter_session_followed_by_statement() { .unwrap(); match stmts[..] { [Statement::AlterSession { .. }, Statement::Query { .. }] => {} - _ => panic!("Unexpected statements: {:?}", stmts), + _ => panic!("Unexpected statements: {stmts:?}"), } } @@ -4082,3 +4262,242 @@ fn parse_connect_by_root_operator() { "sql parser error: Expected an expression, found: FROM" ); } + +#[test] +fn test_begin_exception_end() { + for sql in [ + "BEGIN SELECT 1; EXCEPTION WHEN OTHER THEN SELECT 2; RAISE; END", + "BEGIN SELECT 1; EXCEPTION WHEN OTHER THEN SELECT 2; RAISE EX_1; END", + "BEGIN SELECT 1; EXCEPTION WHEN FOO THEN SELECT 2; WHEN OTHER THEN SELECT 3; RAISE; END", + "BEGIN BEGIN SELECT 1; EXCEPTION WHEN OTHER THEN SELECT 2; RAISE; END; END", + ] { + snowflake().verified_stmt(sql); + } + + let sql = r#" +DECLARE + EXCEPTION_1 EXCEPTION (-20001, 'I caught the expected exception.'); + EXCEPTION_2 EXCEPTION (-20002, 'Not the expected exception!'); + EXCEPTION_3 EXCEPTION (-20003, 'The worst exception...'); +BEGIN + BEGIN + SELECT 1; + EXCEPTION + WHEN EXCEPTION_1 THEN + SELECT 1; + WHEN EXCEPTION_2 OR EXCEPTION_3 THEN + SELECT 2; + SELECT 3; + WHEN OTHER THEN + SELECT 4; + RAISE; + END; +END +"#; + + // Outer `BEGIN` of the two nested `BEGIN` statements. + let Statement::StartTransaction { mut statements, .. } = snowflake() + .parse_sql_statements(sql) + .unwrap() + .pop() + .unwrap() + else { + unreachable!(); + }; + + // Inner `BEGIN` of the two nested `BEGIN` statements. + let Statement::StartTransaction { + statements, + exception, + has_end_keyword, + .. + } = statements.pop().unwrap() + else { + unreachable!(); + }; + + assert_eq!(1, statements.len()); + assert!(has_end_keyword); + + let exception = exception.unwrap(); + assert_eq!(3, exception.len()); + assert_eq!(1, exception[0].idents.len()); + assert_eq!(1, exception[0].statements.len()); + assert_eq!(2, exception[1].idents.len()); + assert_eq!(2, exception[1].statements.len()); +} + +#[test] +fn test_snowflake_fetch_clause_syntax() { + let canonical = "SELECT c1 FROM fetch_test FETCH FIRST 2 ROWS ONLY"; + snowflake().verified_only_select_with_canonical("SELECT c1 FROM fetch_test FETCH 2", canonical); + + snowflake() + .verified_only_select_with_canonical("SELECT c1 FROM fetch_test FETCH FIRST 2", canonical); + snowflake() + .verified_only_select_with_canonical("SELECT c1 FROM fetch_test FETCH NEXT 2", canonical); + + snowflake() + .verified_only_select_with_canonical("SELECT c1 FROM fetch_test FETCH 2 ROW", canonical); + + snowflake().verified_only_select_with_canonical( + "SELECT c1 FROM fetch_test FETCH FIRST 2 ROWS", + canonical, + ); +} + +#[test] +fn test_snowflake_create_view_with_multiple_column_options() { + let create_view_with_tag = + r#"CREATE VIEW X (COL WITH TAG (pii='email') COMMENT 'foobar') AS SELECT * FROM Y"#; + snowflake().verified_stmt(create_view_with_tag); +} + +#[test] +fn test_snowflake_create_view_with_composite_tag() { + let create_view_with_tag = + r#"CREATE VIEW X (COL WITH TAG (foo.bar.baz.pii='email')) AS SELECT * FROM Y"#; + snowflake().verified_stmt(create_view_with_tag); +} + +#[test] +fn test_snowflake_create_view_with_composite_policy_name() { + let create_view_with_tag = + r#"CREATE VIEW X (COL WITH MASKING POLICY foo.bar.baz) AS SELECT * FROM Y"#; + snowflake().verified_stmt(create_view_with_tag); +} + +#[test] +fn test_snowflake_identifier_function() { + // Using IDENTIFIER to reference a column + match &snowflake() + .verified_only_select("SELECT identifier('email') FROM customers") + .projection[0] + { + SelectItem::UnnamedExpr(Expr::Function(Function { name, args, .. })) => { + assert_eq!(*name, ObjectName::from(vec![Ident::new("identifier")])); + assert_eq!( + *args, + FunctionArguments::List(FunctionArgumentList { + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("email".to_string()).into() + )))], + clauses: vec![], + duplicate_treatment: None + }) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference a case-sensitive column + match &snowflake() + .verified_only_select(r#"SELECT identifier('"Email"') FROM customers"#) + .projection[0] + { + SelectItem::UnnamedExpr(Expr::Function(Function { name, args, .. })) => { + assert_eq!(*name, ObjectName::from(vec![Ident::new("identifier")])); + assert_eq!( + *args, + FunctionArguments::List(FunctionArgumentList { + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("\"Email\"".to_string()).into() + )))], + clauses: vec![], + duplicate_treatment: None + }) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference an alias of a table + match &snowflake() + .verified_only_select("SELECT identifier('alias1').* FROM tbl AS alias1") + .projection[0] + { + SelectItem::QualifiedWildcard( + SelectItemQualifiedWildcardKind::Expr(Expr::Function(Function { name, args, .. })), + _, + ) => { + assert_eq!(*name, ObjectName::from(vec![Ident::new("identifier")])); + assert_eq!( + *args, + FunctionArguments::List(FunctionArgumentList { + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("alias1".to_string()).into() + )))], + clauses: vec![], + duplicate_treatment: None + }) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference a database + match snowflake().verified_stmt("CREATE DATABASE IDENTIFIER('tbl')") { + Statement::CreateDatabase { db_name, .. } => { + assert_eq!( + db_name, + ObjectName(vec![ObjectNamePart::Function(ObjectNamePartFunction { + name: Ident::new("IDENTIFIER"), + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("tbl".to_string()).into() + )))] + })]) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference a schema + match snowflake().verified_stmt("CREATE SCHEMA IDENTIFIER('db1.sc1')") { + Statement::CreateSchema { schema_name, .. } => { + assert_eq!( + schema_name, + SchemaName::Simple(ObjectName(vec![ObjectNamePart::Function( + ObjectNamePartFunction { + name: Ident::new("IDENTIFIER"), + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("db1.sc1".to_string()).into() + )))] + } + )])) + ); + } + _ => unreachable!(), + } + + // Using IDENTIFIER to reference a table + match snowflake().verified_stmt("CREATE TABLE IDENTIFIER('tbl') (id INT)") { + Statement::CreateTable(CreateTable { name, .. }) => { + assert_eq!( + name, + ObjectName(vec![ObjectNamePart::Function(ObjectNamePartFunction { + name: Ident::new("IDENTIFIER"), + args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value( + Value::SingleQuotedString("tbl".to_string()).into() + )))] + })]) + ); + } + _ => unreachable!(), + } + + // Cannot have more than one IDENTIFIER part in an object name + assert_eq!( + snowflake() + .parse_sql_statements( + "CREATE TABLE IDENTIFIER('db1').IDENTIFIER('sc1').IDENTIFIER('tbl') (id INT)" + ) + .is_err(), + true + ); + assert_eq!( + snowflake() + .parse_sql_statements("CREATE TABLE IDENTIFIER('db1')..IDENTIFIER('tbl') (id INT)") + .is_err(), + true + ); +} diff --git a/tests/sqlparser_sqlite.rs b/tests/sqlparser_sqlite.rs index b759065f..06496f0c 100644 --- a/tests/sqlparser_sqlite.rs +++ b/tests/sqlparser_sqlite.rs @@ -324,7 +324,7 @@ fn parse_create_table_on_conflict_col() { Keyword::IGNORE, Keyword::REPLACE, ] { - let sql = format!("CREATE TABLE t1 (a INT, b INT ON CONFLICT {:?})", keyword); + let sql = format!("CREATE TABLE t1 (a INT, b INT ON CONFLICT {keyword:?})"); match sqlite_and_generic().verified_stmt(&sql) { Statement::CreateTable(CreateTable { columns, .. }) => { assert_eq!( @@ -410,7 +410,7 @@ fn parse_window_function_with_filter() { "count", "user_defined_function", ] { - let sql = format!("SELECT {}(x) FILTER (WHERE y) OVER () FROM t", func_name); + let sql = format!("SELECT {func_name}(x) FILTER (WHERE y) OVER () FROM t"); let select = sqlite().verified_only_select(&sql); assert_eq!(select.to_string(), sql); assert_eq!( @@ -444,7 +444,7 @@ fn parse_window_function_with_filter() { fn parse_attach_database() { let sql = "ATTACH DATABASE 'test.db' AS test"; let verified_stmt = sqlite().verified_stmt(sql); - assert_eq!(sql, format!("{}", verified_stmt)); + assert_eq!(sql, format!("{verified_stmt}")); match verified_stmt { Statement::AttachDatabase { schema_name,