mirrors/datafusion-sqlparse
1
0
Fork 0
mirror of https://github.com/apache/datafusion-sqlparser-rs.git synced 2025-08-04 06:18:17 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

Named window frames (#881)

Browse source 
* after over clause, named window can be parsed with window ... as after having clause

* Lint errors are fixed

* Support for multiple windows

* fix lint errors

* simplifications

* rename function

* Rewrite named window search in functional style

* Test added and some minor changes

* Minor changes on tests and namings, and semantic check is removed

---------

Co-authored-by: Mustafa Akur <mustafa.akur@synnada.ai>
Co-authored-by: Mehmet Ozan Kabak <ozankabak@gmail.com>
This commit is contained in:
Berkay Şahin 2023-05-18 22:00:24 +03:00 • committed by GitHub
parent 1b86abebe2
commit ef46cd3752
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
8 changed files with 225 additions and 37 deletions
Download patch file Download diff file Expand all files Collapse all files

31
src/ast/mod.rs
View file

@ -17,7 +17,7 @@ use alloc::{
string::{String, ToString},
vec::Vec,
};
use core::fmt;
use core::fmt::{self, Display};
#[cfg(feature = "serde")]
use serde::{Deserialize, Serialize};
@ -35,10 +35,10 @@ pub use self::ddl::{
pub use self::operator::{BinaryOperator, UnaryOperator};
pub use self::query::{
Cte, Distinct, ExceptSelectItem, ExcludeSelectItem, Fetch, IdentWithAlias, Join,
JoinConstraint, JoinOperator, LateralView, LockClause, LockType, NonBlock, Offset, OffsetRows,
OrderByExpr, Query, RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select,
SelectInto, SelectItem, SetExpr, SetOperator, SetQuantifier, Table, TableAlias, TableFactor,
TableWithJoins, Top, Values, WildcardAdditionalOptions, With,
JoinConstraint, JoinOperator, LateralView, LockClause, LockType, NamedWindowDefinition,
NonBlock, Offset, OffsetRows, OrderByExpr, Query, RenameSelectItem, ReplaceSelectElement,
ReplaceSelectItem, Select, SelectInto, SelectItem, SetExpr, SetOperator, SetQuantifier, Table,
TableAlias, TableFactor, TableWithJoins, Top, Values, WildcardAdditionalOptions, With,
};
pub use self::value::{
escape_quoted_string, DateTimeField, DollarQuotedString, TrimWhereField, Value,
@ -930,6 +930,23 @@ impl fmt::Display for Expr {
}
}
#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
pub enum WindowType {
WindowSpec(WindowSpec),
NamedWindow(Ident),
}
impl Display for WindowType {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match self {
WindowType::WindowSpec(spec) => write!(f, "({})", spec),
WindowType::NamedWindow(name) => write!(f, "{}", name),
}
}
}
/// A window specification (i.e. `OVER (PARTITION BY .. ORDER BY .. etc.)`)
#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
@ -3360,7 +3377,7 @@ impl fmt::Display for CloseCursor {
pub struct Function {
pub name: ObjectName,
pub args: Vec<FunctionArg>,
pub over: Option<WindowSpec>,
pub over: Option<WindowType>,
// aggregate functions may specify eg `COUNT(DISTINCT x)`
pub distinct: bool,
// Some functions must be called without trailing parentheses, for example Postgres
@ -3409,7 +3426,7 @@ impl fmt::Display for Function {
)?;
if let Some(o) = &self.over {
write!(f, " OVER ({o})")?;
write!(f, " OVER {o}")?;
}
}

16
src/ast/query.rs
View file

@ -216,6 +216,8 @@ pub struct Select {
pub sort_by: Vec<Expr>,
/// HAVING
pub having: Option<Expr>,
/// WINDOW AS
pub named_window: Vec<NamedWindowDefinition>,
/// QUALIFY (Snowflake)
pub qualify: Option<Expr>,
}
@ -269,6 +271,9 @@ impl fmt::Display for Select {
if let Some(ref having) = self.having {
write!(f, " HAVING {having}")?;
}
if !self.named_window.is_empty() {
write!(f, " WINDOW {}", display_comma_separated(&self.named_window))?;
}
if let Some(ref qualify) = self.qualify {
write!(f, " QUALIFY {qualify}")?;
}
@ -311,6 +316,17 @@ impl fmt::Display for LateralView {
}
}
#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
pub struct NamedWindowDefinition(pub Ident, pub WindowSpec);
impl fmt::Display for NamedWindowDefinition {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "{} AS ({})", self.0, self.1)
}
}
#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]

1
src/keywords.rs
View file

@ -688,6 +688,7 @@ pub const RESERVED_FOR_TABLE_ALIAS: &[Keyword] = &[
Keyword::OUTER,
Keyword::SET,
Keyword::QUALIFY,
Keyword::WINDOW,
];
/// Can't be used as a column alias, so that `SELECT <expr> alias`

70
src/parser.rs
View file

@ -884,32 +884,12 @@ impl<'a> Parser<'a> {
let distinct = self.parse_all_or_distinct()?.is_some();
let (args, order_by) = self.parse_optional_args_with_orderby()?;
let over = if self.parse_keyword(Keyword::OVER) {
// TBD: support window names (`OVER mywin`) in place of inline specification
self.expect_token(&Token::LParen)?;
let partition_by = if self.parse_keywords(&[Keyword::PARTITION, Keyword::BY]) {
// a list of possibly-qualified column names
self.parse_comma_separated(Parser::parse_expr)?
if self.consume_token(&Token::LParen) {
let window_spec = self.parse_window_spec()?;
Some(WindowType::WindowSpec(window_spec))
} else {
vec![]
};
let order_by = if self.parse_keywords(&[Keyword::ORDER, Keyword::BY]) {
self.parse_comma_separated(Parser::parse_order_by_expr)?
} else {
vec![]
};
let window_frame = if !self.consume_token(&Token::RParen) {
let window_frame = self.parse_window_frame()?;
self.expect_token(&Token::RParen)?;
Some(window_frame)
} else {
None
};
Some(WindowSpec {
partition_by,
order_by,
window_frame,
})
Some(WindowType::NamedWindow(self.parse_identifier()?))
}
} else {
None
};
@ -5367,6 +5347,12 @@ impl<'a> Parser<'a> {
None
};
let named_windows = if self.parse_keyword(Keyword::WINDOW) {
self.parse_comma_separated(Parser::parse_named_window)?
} else {
vec![]
};
let qualify = if self.parse_keyword(Keyword::QUALIFY) {
Some(self.parse_expr()?)
} else {
@ -5386,6 +5372,7 @@ impl<'a> Parser<'a> {
distribute_by,
sort_by,
having,
named_window: named_windows,
qualify,
})
}
@ -7033,6 +7020,39 @@ impl<'a> Parser<'a> {
pub fn index(&self) -> usize {
self.index
}
pub fn parse_named_window(&mut self) -> Result<NamedWindowDefinition, ParserError> {
let ident = self.parse_identifier()?;
self.expect_keyword(Keyword::AS)?;
self.expect_token(&Token::LParen)?;
let window_spec = self.parse_window_spec()?;
Ok(NamedWindowDefinition(ident, window_spec))
}
pub fn parse_window_spec(&mut self) -> Result<WindowSpec, ParserError> {
let partition_by = if self.parse_keywords(&[Keyword::PARTITION, Keyword::BY]) {
self.parse_comma_separated(Parser::parse_expr)?
} else {
vec![]
};
let order_by = if self.parse_keywords(&[Keyword::ORDER, Keyword::BY]) {
self.parse_comma_separated(Parser::parse_order_by_expr)?
} else {
vec![]
};
let window_frame = if !self.consume_token(&Token::RParen) {
let window_frame = self.parse_window_frame()?;
self.expect_token(&Token::RParen)?;
Some(window_frame)
} else {
None
};
Ok(WindowSpec {
partition_by,
order_by,
window_frame,
})
}
}
impl Word {

1
tests/sqlparser_clickhouse.rs
View file

@ -101,6 +101,7 @@ fn parse_map_access_expr() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None
},
select

134
tests/sqlparser_common.rs
View file

@ -19,7 +19,6 @@
//! dialect-specific parsing rules).
use matches::assert_matches;
use sqlparser::ast::SelectItem::UnnamedExpr;
use sqlparser::ast::TableFactor::Pivot;
use sqlparser::ast::*;
@ -251,6 +250,7 @@ fn parse_update_set_from() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None
}))),
order_by: vec![],
@ -1721,7 +1721,7 @@ fn parse_select_qualify() {
left: Box::new(Expr::Function(Function {
name: ObjectName(vec![Ident::new("ROW_NUMBER")]),
args: vec![],
over: Some(WindowSpec {
over: Some(WindowType::WindowSpec(WindowSpec {
partition_by: vec![Expr::Identifier(Ident::new("p"))],
order_by: vec![OrderByExpr {
expr: Expr::Identifier(Ident::new("o")),
@ -1729,7 +1729,7 @@ fn parse_select_qualify() {
nulls_first: None,
}],
window_frame: None,
}),
})),
distinct: false,
special: false,
order_by: vec![],
@ -3295,7 +3295,7 @@ fn parse_window_functions() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::new("row_number")]),
args: vec![],
over: Some(WindowSpec {
over: Some(WindowType::WindowSpec(WindowSpec {
partition_by: vec![],
order_by: vec![OrderByExpr {
expr: Expr::Identifier(Ident::new("dt")),
@ -3303,7 +3303,7 @@ fn parse_window_functions() {
nulls_first: None,
}],
window_frame: None,
}),
})),
distinct: false,
special: false,
order_by: vec![],
@ -3312,6 +3312,128 @@ fn parse_window_functions() {
);
}
#[test]
fn test_parse_named_window() {
let sql = "SELECT \
MIN(c12) OVER window1 AS min1, \
MAX(c12) OVER window2 AS max1 \
FROM aggregate_test_100 \
WINDOW window1 AS (ORDER BY C12), \
window2 AS (PARTITION BY C11) \
ORDER BY C3";
let actual_select_only = verified_only_select(sql);
let expected = Select {
distinct: None,
top: None,
projection: vec![
SelectItem::ExprWithAlias {
expr: Expr::Function(Function {
name: ObjectName(vec![Ident {
value: "MIN".to_string(),
quote_style: None,
}]),
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident {
value: "c12".to_string(),
quote_style: None,
}),
))],
over: Some(WindowType::NamedWindow(Ident {
value: "window1".to_string(),
quote_style: None,
})),
distinct: false,
special: false,
}),
alias: Ident {
value: "min1".to_string(),
quote_style: None,
},
},
SelectItem::ExprWithAlias {
expr: Expr::Function(Function {
name: ObjectName(vec![Ident {
value: "MAX".to_string(),
quote_style: None,
}]),
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident {
value: "c12".to_string(),
quote_style: None,
}),
))],
over: Some(WindowType::NamedWindow(Ident {
value: "window2".to_string(),
quote_style: None,
})),
distinct: false,
special: false,
}),
alias: Ident {
value: "max1".to_string(),
quote_style: None,
},
},
],
into: None,
from: vec![TableWithJoins {
relation: TableFactor::Table {
name: ObjectName(vec![Ident {
value: "aggregate_test_100".to_string(),
quote_style: None,
}]),
alias: None,
args: None,
with_hints: vec![],
},
joins: vec![],
}],
lateral_views: vec![],
selection: None,
group_by: vec![],
cluster_by: vec![],
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![
NamedWindowDefinition(
Ident {
value: "window1".to_string(),
quote_style: None,
},
WindowSpec {
partition_by: vec![],
order_by: vec![OrderByExpr {
expr: Expr::Identifier(Ident {
value: "C12".to_string(),
quote_style: None,
}),
asc: None,
nulls_first: None,
}],
window_frame: None,
},
),
NamedWindowDefinition(
Ident {
value: "window2".to_string(),
quote_style: None,
},
WindowSpec {
partition_by: vec![Expr::Identifier(Ident {
value: "C11".to_string(),
quote_style: None,
})],
order_by: vec![],
window_frame: None,
},
),
],
qualify: None,
};
assert_eq!(actual_select_only, expected);
}
#[test]
fn parse_aggregate_with_group_by() {
let sql = "SELECT a, COUNT(1), MIN(b), MAX(b) FROM foo GROUP BY a";
@ -3659,6 +3781,7 @@ fn parse_interval_and_or_xor() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None,
}))),
order_by: vec![],
@ -5929,6 +6052,7 @@ fn parse_merge() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None,
}))),
order_by: vec![],

6
tests/sqlparser_mysql.rs
View file

@ -460,6 +460,7 @@ fn parse_quote_identifiers_2() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None
}))),
order_by: vec![],
@ -494,6 +495,7 @@ fn parse_quote_identifiers_3() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None
}))),
order_by: vec![],
@ -884,6 +886,7 @@ fn parse_select_with_numeric_prefix_column_name() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None,
})))
);
@ -927,6 +930,7 @@ fn parse_select_with_concatenation_of_exp_number_and_numeric_prefix_column() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None,
})))
);
@ -1116,6 +1120,7 @@ fn parse_substring_in_select() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None
}))),
order_by: vec![],
@ -1394,6 +1399,7 @@ fn parse_hex_string_introducer() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None,
into: None
}))),

3
tests/sqlparser_postgres.rs
View file

@ -954,6 +954,7 @@ fn parse_copy_to() {
selection: None,
group_by: vec![],
having: None,
named_window: vec![],
cluster_by: vec![],
distribute_by: vec![],
sort_by: vec![],
@ -1799,6 +1800,7 @@ fn parse_array_subquery_expr() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None,
}))),
right: Box::new(SetExpr::Select(Box::new(Select {
@ -1820,6 +1822,7 @@ fn parse_array_subquery_expr() {
distribute_by: vec![],
sort_by: vec![],
having: None,
named_window: vec![],
qualify: None,
}))),
}),