mirrors/datafusion-sqlparse
1
0
Fork 0
mirror of https://github.com/apache/datafusion-sqlparser-rs.git synced 2025-07-07 17:04:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 3

Support FILTER in over clause (#1007)

Browse source 
Co-authored-by: Andrew Lamb <andrew@nerdnetworks.org>
This commit is contained in:
Ophir LOJKINE 2023-10-24 00:06:39 +02:00 • committed by GitHub
parent e857a45201
commit ce62fe6d27
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
15 changed files with 102 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

2
README.md
View file

@ -34,7 +34,7 @@ println!("AST: {:?}", ast);
This outputs
```rust
AST: [Query(Query { ctes: [], body: Select(Select { distinct: false, projection: [UnnamedExpr(Identifier("a")), UnnamedExpr(Identifier("b")), UnnamedExpr(Value(Long(123))), UnnamedExpr(Function(Function { name: ObjectName(["myfunc"]), args: [Identifier("b")], over: None, distinct: false }))], from: [TableWithJoins { relation: Table { name: ObjectName(["table_1"]), alias: None, args: [], with_hints: [] }, joins: [] }], selection: Some(BinaryOp { left: BinaryOp { left: Identifier("a"), op: Gt, right: Identifier("b") }, op: And, right: BinaryOp { left: Identifier("b"), op: Lt, right: Value(Long(100)) } }), group_by: [], having: None }), order_by: [OrderByExpr { expr: Identifier("a"), asc: Some(false) }, OrderByExpr { expr: Identifier("b"), asc: None }], limit: None, offset: None, fetch: None })]
AST: [Query(Query { ctes: [], body: Select(Select { distinct: false, projection: [UnnamedExpr(Identifier("a")), UnnamedExpr(Identifier("b")), UnnamedExpr(Value(Long(123))), UnnamedExpr(Function(Function { name: ObjectName(["myfunc"]), args: [Identifier("b")], filter: None, over: None, distinct: false }))], from: [TableWithJoins { relation: Table { name: ObjectName(["table_1"]), alias: None, args: [], with_hints: [] }, joins: [] }], selection: Some(BinaryOp { left: BinaryOp { left: Identifier("a"), op: Gt, right: Identifier("b") }, op: And, right: BinaryOp { left: Identifier("b"), op: Lt, right: Value(Long(100)) } }), group_by: [], having: None }), order_by: [OrderByExpr { expr: Identifier("a"), asc: Some(false) }, OrderByExpr { expr: Identifier("b"), asc: None }], limit: None, offset: None, fetch: None })]
```

9
src/ast/mod.rs
View file

@ -1070,8 +1070,11 @@ impl Display for WindowType {
#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
pub struct WindowSpec {
/// `OVER (PARTITION BY ...)`
pub partition_by: Vec<Expr>,
/// `OVER (ORDER BY ...)`
pub order_by: Vec<OrderByExpr>,
/// `OVER (window frame)`
pub window_frame: Option<WindowFrame>,
}
@ -3729,6 +3732,8 @@ impl fmt::Display for CloseCursor {
pub struct Function {
pub name: ObjectName,
pub args: Vec<FunctionArg>,
/// e.g. `x > 5` in `COUNT(x) FILTER (WHERE x > 5)`
pub filter: Option<Box<Expr>>,
pub over: Option<WindowType>,
// aggregate functions may specify eg `COUNT(DISTINCT x)`
pub distinct: bool,
@ -3777,6 +3782,10 @@ impl fmt::Display for Function {
display_comma_separated(&self.order_by),
)?;
if let Some(filter_cond) = &self.filter {
write!(f, " FILTER (WHERE {filter_cond})")?;
}
if let Some(o) = &self.over {
write!(f, " OVER {o}")?;
}

2
src/ast/visitor.rs
View file

@ -506,7 +506,7 @@ where
/// *expr = Expr::Function(Function {
/// name: ObjectName(vec![Ident::new("f")]),
/// args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(old_expr))],
/// over: None, distinct: false, special: false, order_by: vec![],
/// filter: None, over: None, distinct: false, special: false, order_by: vec![],
/// });
/// }
/// ControlFlow::<()>::Continue(())

4
src/dialect/sqlite.rs
View file

@ -35,6 +35,10 @@ impl Dialect for SQLiteDialect {
|| ('\u{007f}'..='\u{ffff}').contains(&ch)
}
fn supports_filter_during_aggregation(&self) -> bool {
true
}
fn is_identifier_part(&self, ch: char) -> bool {
self.is_identifier_start(ch) || ch.is_ascii_digit()
}

14
src/parser/mod.rs
View file

@ -772,6 +772,7 @@ impl<'a> Parser<'a> {
Ok(Expr::Function(Function {
name: ObjectName(vec![w.to_ident()]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: true,
@ -957,6 +958,17 @@ impl<'a> Parser<'a> {
self.expect_token(&Token::LParen)?;
let distinct = self.parse_all_or_distinct()?.is_some();
let (args, order_by) = self.parse_optional_args_with_orderby()?;
let filter = if self.dialect.supports_filter_during_aggregation()
&& self.parse_keyword(Keyword::FILTER)
&& self.consume_token(&Token::LParen)
&& self.parse_keyword(Keyword::WHERE)
{
let filter = Some(Box::new(self.parse_expr()?));
self.expect_token(&Token::RParen)?;
filter
} else {
None
};
let over = if self.parse_keyword(Keyword::OVER) {
if self.consume_token(&Token::LParen) {
let window_spec = self.parse_window_spec()?;
@ -970,6 +982,7 @@ impl<'a> Parser<'a> {
Ok(Expr::Function(Function {
name,
args,
filter,
over,
distinct,
special: false,
@ -987,6 +1000,7 @@ impl<'a> Parser<'a> {
Ok(Expr::Function(Function {
name,
args,
filter: None,
over: None,
distinct: false,
special,

1
tests/sqlparser_bigquery.rs
View file

@ -564,6 +564,7 @@ fn parse_map_access_offset() {
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(
number("0")
))),],
filter: None,
over: None,
distinct: false,
special: false,

4
tests/sqlparser_clickhouse.rs
View file

@ -50,6 +50,7 @@ fn parse_map_access_expr() {
Value::SingleQuotedString("endpoint".to_string())
))),
],
filter: None,
over: None,
distinct: false,
special: false,
@ -89,6 +90,7 @@ fn parse_map_access_expr() {
Value::SingleQuotedString("app".to_string())
))),
],
filter: None,
over: None,
distinct: false,
special: false,
@ -138,6 +140,7 @@ fn parse_array_fn() {
FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Identifier(Ident::new("x1")))),
FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Identifier(Ident::new("x2")))),
],
filter: None,
over: None,
distinct: false,
special: false,
@ -196,6 +199,7 @@ fn parse_delimited_identifiers() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::with_quote('"', "myfun")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: false,

19
tests/sqlparser_common.rs
View file

@ -875,6 +875,7 @@ fn parse_select_count_wildcard() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::new("COUNT")]),
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Wildcard)],
filter: None,
over: None,
distinct: false,
special: false,
@ -895,6 +896,7 @@ fn parse_select_count_distinct() {
op: UnaryOperator::Plus,
expr: Box::new(Expr::Identifier(Ident::new("x"))),
}))],
filter: None,
over: None,
distinct: true,
special: false,
@ -1862,6 +1864,7 @@ fn parse_select_having() {
left: Box::new(Expr::Function(Function {
name: ObjectName(vec![Ident::new("COUNT")]),
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Wildcard)],
filter: None,
over: None,
distinct: false,
special: false,
@ -1887,6 +1890,7 @@ fn parse_select_qualify() {
left: Box::new(Expr::Function(Function {
name: ObjectName(vec![Ident::new("ROW_NUMBER")]),
args: vec![],
filter: None,
over: Some(WindowType::WindowSpec(WindowSpec {
partition_by: vec![Expr::Identifier(Ident::new("p"))],
order_by: vec![OrderByExpr {
@ -3342,6 +3346,7 @@ fn parse_scalar_function_in_projection() {
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident::new("id"))
))],
filter: None,
over: None,
distinct: false,
special: false,
@ -3461,6 +3466,7 @@ fn parse_named_argument_function() {
))),
},
],
filter: None,
over: None,
distinct: false,
special: false,
@ -3492,6 +3498,7 @@ fn parse_window_functions() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::new("row_number")]),
args: vec![],
filter: None,
over: Some(WindowType::WindowSpec(WindowSpec {
partition_by: vec![],
order_by: vec![OrderByExpr {
@ -3535,6 +3542,7 @@ fn test_parse_named_window() {
quote_style: None,
}),
))],
filter: None,
over: Some(WindowType::NamedWindow(Ident {
value: "window1".to_string(),
quote_style: None,
@ -3560,6 +3568,7 @@ fn test_parse_named_window() {
quote_style: None,
}),
))],
filter: None,
over: Some(WindowType::NamedWindow(Ident {
value: "window2".to_string(),
quote_style: None,
@ -4029,6 +4038,7 @@ fn parse_at_timezone() {
quote_style: None,
}]),
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(zero.clone()))],
filter: None,
over: None,
distinct: false,
special: false,
@ -4056,6 +4066,7 @@ fn parse_at_timezone() {
quote_style: None,
},],),
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(zero))],
filter: None,
over: None,
distinct: false,
special: false,
@ -4067,6 +4078,7 @@ fn parse_at_timezone() {
Value::SingleQuotedString("%Y-%m-%dT%H".to_string()),
),),),
],
filter: None,
over: None,
distinct: false,
special: false,
@ -4225,6 +4237,7 @@ fn parse_table_function() {
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(
Value::SingleQuotedString("1".to_owned()),
)))],
filter: None,
over: None,
distinct: false,
special: false,
@ -4376,6 +4389,7 @@ fn parse_unnest_in_from_clause() {
FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("2")))),
FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("3")))),
],
filter: None,
over: None,
distinct: false,
special: false,
@ -4405,6 +4419,7 @@ fn parse_unnest_in_from_clause() {
FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("2")))),
FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("3")))),
],
filter: None,
over: None,
distinct: false,
special: false,
@ -4416,6 +4431,7 @@ fn parse_unnest_in_from_clause() {
FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("5")))),
FunctionArg::Unnamed(FunctionArgExpr::Expr(Expr::Value(number("6")))),
],
filter: None,
over: None,
distinct: false,
special: false,
@ -6888,6 +6904,7 @@ fn parse_time_functions() {
let select_localtime_func_call_ast = Function {
name: ObjectName(vec![Ident::new(func_name)]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: false,
@ -7374,6 +7391,7 @@ fn parse_pivot_table() {
args: (vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::CompoundIdentifier(vec![Ident::new("a"), Ident::new("amount"),])
))]),
filter: None,
over: None,
distinct: false,
special: false,
@ -7523,6 +7541,7 @@ fn parse_pivot_unpivot_table() {
args: (vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident::new("population"))
))]),
filter: None,
over: None,
distinct: false,
special: false,

1
tests/sqlparser_hive.rs
View file

@ -346,6 +346,7 @@ fn parse_delimited_identifiers() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::with_quote('"', "myfun")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: false,

1
tests/sqlparser_mssql.rs
View file

@ -334,6 +334,7 @@ fn parse_delimited_identifiers() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::with_quote('"', "myfun")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: false,

6
tests/sqlparser_mysql.rs
View file

@ -1071,6 +1071,7 @@ fn parse_insert_with_on_duplicate_update() {
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident::new("description"))
))],
filter: None,
over: None,
distinct: false,
special: false,
@ -1084,6 +1085,7 @@ fn parse_insert_with_on_duplicate_update() {
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident::new("perm_create"))
))],
filter: None,
over: None,
distinct: false,
special: false,
@ -1097,6 +1099,7 @@ fn parse_insert_with_on_duplicate_update() {
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident::new("perm_read"))
))],
filter: None,
over: None,
distinct: false,
special: false,
@ -1110,6 +1113,7 @@ fn parse_insert_with_on_duplicate_update() {
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident::new("perm_update"))
))],
filter: None,
over: None,
distinct: false,
special: false,
@ -1123,6 +1127,7 @@ fn parse_insert_with_on_duplicate_update() {
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident::new("perm_delete"))
))],
filter: None,
over: None,
distinct: false,
special: false,
@ -1512,6 +1517,7 @@ fn parse_table_colum_option_on_update() {
option: ColumnOption::OnUpdate(Expr::Function(Function {
name: ObjectName(vec![Ident::new("CURRENT_TIMESTAMP")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: false,

6
tests/sqlparser_postgres.rs
View file

@ -2275,6 +2275,7 @@ fn test_composite_value() {
named: true
}
)))],
filter: None,
over: None,
distinct: false,
special: false,
@ -2436,6 +2437,7 @@ fn parse_current_functions() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::new("CURRENT_CATALOG")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: true,
@ -2447,6 +2449,7 @@ fn parse_current_functions() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::new("CURRENT_USER")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: true,
@ -2458,6 +2461,7 @@ fn parse_current_functions() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::new("SESSION_USER")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: true,
@ -2469,6 +2473,7 @@ fn parse_current_functions() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::new("USER")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: true,
@ -2919,6 +2924,7 @@ fn parse_delimited_identifiers() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::with_quote('"', "myfun")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: false,

1
tests/sqlparser_redshift.rs
View file

@ -137,6 +137,7 @@ fn parse_delimited_identifiers() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::with_quote('"', "myfun")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: false,

1
tests/sqlparser_snowflake.rs
View file

@ -248,6 +248,7 @@ fn parse_delimited_identifiers() {
&Expr::Function(Function {
name: ObjectName(vec![Ident::with_quote('"', "myfun")]),
args: vec![],
filter: None,
over: None,
distinct: false,
special: false,

33
tests/sqlparser_sqlite.rs
View file

@ -290,6 +290,39 @@ fn parse_create_table_with_strict() {
}
}
#[test]
fn parse_window_function_with_filter() {
for func_name in [
"row_number",
"rank",
"max",
"count",
"user_defined_function",
] {
let sql = format!("SELECT {}(x) FILTER (WHERE y) OVER () FROM t", func_name);
let select = sqlite().verified_only_select(&sql);
assert_eq!(select.to_string(), sql);
assert_eq!(
select.projection,
vec![SelectItem::UnnamedExpr(Expr::Function(Function {
name: ObjectName(vec![Ident::new(func_name)]),
args: vec![FunctionArg::Unnamed(FunctionArgExpr::Expr(
Expr::Identifier(Ident::new("x"))
))],
over: Some(WindowType::WindowSpec(WindowSpec {
partition_by: vec![],
order_by: vec![],
window_frame: None,
})),
filter: Some(Box::new(Expr::Identifier(Ident::new("y")))),
distinct: false,
special: false,
order_by: vec![]
}))]
);
}
}
#[test]
fn parse_attach_database() {
let sql = "ATTACH DATABASE 'test.db' AS test";