feat: change Expr OuterReferenceColumn and Alias to Box type for reducing expr struct size #16771
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -279,7 +279,7 @@ use sqlparser::ast::{ | |
| #[derive(Clone, PartialEq, PartialOrd, Eq, Debug, Hash)] | ||
| pub enum Expr { | ||
| /// An expression with a specific name. | ||
| Alias(Box<Alias>), | ||
| /// A named reference to a qualified field in a schema. | ||
| Column(Column), | ||
| /// A named reference to a variable in a registry. | ||
|
|
@@ -362,7 +362,7 @@ pub enum Expr { | |
| Placeholder(Placeholder), | ||
| /// A placeholder which holds a reference to a qualified field | ||
| /// in the outer query, used for correlated sub queries. | ||
| OuterReferenceColumn(Box<(DataType, Column)>), | ||
|
There was a problem hiding this comment. If we are going to make this change anyways, can we also pull this into a named struct rather than a unnamed tuple like struct OuterReference {
// fields here
}
enum Expr {
...
OuterReferenceColumn(OuterReference),
...
}There was a problem hiding this comment. Thank you @alamb for this good suggestion, addressed in latest PR. |
||
| /// Unnest expression | ||
| Unnest(Unnest), | ||
| } | ||
|
|
@@ -1424,7 +1424,9 @@ impl Expr { | |
| name, | ||
| spans: _, | ||
| }) => (relation.clone(), name.clone()), | ||
| Expr::Alias(boxed_alias) => { | ||
| (boxed_alias.relation.clone(), boxed_alias.name.clone()) | ||
| } | ||
| _ => (None, self.schema_name().to_string()), | ||
| } | ||
| } | ||
|
|
@@ -1446,7 +1448,7 @@ impl Expr { | |
| Expr::Case { .. } => "Case", | ||
| Expr::Cast { .. } => "Cast", | ||
| Expr::Column(..) => "Column", | ||
| Expr::OuterReferenceColumn(_) => "Outer", | ||
| Expr::Exists { .. } => "Exists", | ||
| Expr::GroupingSet(..) => "GroupingSet", | ||
| Expr::InList { .. } => "InList", | ||
|
|
@@ -1572,7 +1574,7 @@ impl Expr { | |
|
|
||
| /// Return `self AS name` alias expression | ||
| pub fn alias(self, name: impl Into<String>) -> Expr { | ||
| Expr::Alias(Box::new(Alias::new(self, None::<&str>, name.into()))) | ||
| } | ||
|
|
||
| /// Return `self AS name` alias expression with metadata | ||
|
|
@@ -1595,7 +1597,9 @@ impl Expr { | |
| name: impl Into<String>, | ||
| metadata: Option<FieldMetadata>, | ||
| ) -> Expr { | ||
| Expr::Alias(Box::new( | ||
| Alias::new(self, None::<&str>, name.into()).with_metadata(metadata), | ||
| )) | ||
| } | ||
|
|
||
| /// Return `self AS name` alias expression with a specific qualifier | ||
|
|
@@ -1604,7 +1608,7 @@ impl Expr { | |
| relation: Option<impl Into<TableReference>>, | ||
| name: impl Into<String>, | ||
| ) -> Expr { | ||
| Expr::Alias(Box::new(Alias::new(self, relation, name.into()))) | ||
| } | ||
|
|
||
| /// Return `self AS name` alias expression with a specific qualifier and metadata | ||
|
|
@@ -1628,7 +1632,9 @@ impl Expr { | |
| name: impl Into<String>, | ||
| metadata: Option<FieldMetadata>, | ||
| ) -> Expr { | ||
| Expr::Alias(Box::new( | ||
| Alias::new(self, relation, name.into()).with_metadata(metadata), | ||
| )) | ||
| } | ||
|
|
||
| /// Remove an alias from an expression if one exists. | ||
|
|
@@ -2021,7 +2027,7 @@ impl Expr { | |
| | Expr::SimilarTo(..) | ||
| | Expr::Not(..) | ||
| | Expr::Negative(..) | ||
| | Expr::OuterReferenceColumn(_) | ||
| | Expr::TryCast(..) | ||
| | Expr::Unnest(..) | ||
| | Expr::Wildcard { .. } | ||
|
|
@@ -2109,23 +2115,10 @@ impl NormalizeEq for Expr { | |
| && self_right.normalize_eq(other_right) | ||
| } | ||
| } | ||
| (Expr::Alias(boxed_alias), Expr::Alias(boxed_other_alias)) => { | ||
| boxed_alias.name == boxed_other_alias.name | ||
| && boxed_alias.relation == boxed_other_alias.relation | ||
| && boxed_alias.expr.normalize_eq(&*boxed_other_alias.expr) | ||
| } | ||
| ( | ||
| Expr::Like(Like { | ||
|
|
@@ -2459,14 +2452,9 @@ impl HashNode for Expr { | |
| fn hash_node<H: Hasher>(&self, state: &mut H) { | ||
| mem::discriminant(self).hash(state); | ||
| match self { | ||
| Expr::Alias(boxed_alias) => { | ||
| boxed_alias.relation.hash(state); | ||
| boxed_alias.name.hash(state); | ||
| } | ||
| Expr::Column(column) => { | ||
| column.hash(state); | ||
|
|
@@ -2609,7 +2597,8 @@ impl HashNode for Expr { | |
| Expr::Placeholder(place_holder) => { | ||
| place_holder.hash(state); | ||
| } | ||
| Expr::OuterReferenceColumn(boxed_orc) => { | ||
| let (data_type, column) = boxed_orc.as_ref(); | ||
| data_type.hash(state); | ||
| column.hash(state); | ||
| } | ||
|
|
@@ -2673,12 +2662,14 @@ impl Display for SchemaDisplay<'_> { | |
| } | ||
| } | ||
| // Expr is not shown since it is aliased | ||
| Expr::Alias(boxed_alias) => { | ||
| let alias = boxed_alias.as_ref(); | ||
| if let Some(relation) = &alias.relation { | ||
| write!(f, "{relation}.{}", alias.name) | ||
| } else { | ||
| write!(f, "{}", alias.name) | ||
| } | ||
| } | ||
| Expr::Between(Between { | ||
| expr, | ||
| negated, | ||
|
|
@@ -2929,7 +2920,10 @@ impl Display for SqlDisplay<'_> { | |
| fn fmt(&self, f: &mut Formatter<'_>) -> fmt::Result { | ||
| match self.0 { | ||
| Expr::Literal(scalar, _) => scalar.fmt(f), | ||
| Expr::Alias(boxed_alias) => { | ||
| let name = &boxed_alias.as_ref().name; | ||
| write!(f, "{name}") | ||
| } | ||
| Expr::Between(Between { | ||
| expr, | ||
| negated, | ||
|
|
@@ -3189,9 +3183,13 @@ pub const UNNEST_COLUMN_PREFIX: &str = "UNNEST"; | |
| impl Display for Expr { | ||
| fn fmt(&self, f: &mut Formatter) -> fmt::Result { | ||
| match self { | ||
| Expr::Alias(boxed_alias) => { | ||
| let Alias { expr, name, .. } = boxed_alias.as_ref(); | ||
| write!(f, "{expr} AS {name}") | ||
| } | ||
| Expr::Column(c) => write!(f, "{c}"), | ||
| Expr::OuterReferenceColumn(boxed_orc) => { | ||
| let (_, c) = boxed_orc.as_ref(); | ||
| write!(f, "{OUTER_REFERENCE_COLUMN_PREFIX}({c})") | ||
| } | ||
| Expr::ScalarVariable(_, var_names) => write!(f, "{}", var_names.join(".")), | ||
|
|
@@ -3839,7 +3837,7 @@ mod test { | |
| // If this test fails when you change `Expr`, please try | ||
| // `Box`ing the fields to make `Expr` smaller | ||
| // See https://github.com/apache/datafusion/issues/16199 for details | ||
| assert_eq!(size_of::<Expr>(), 112); | ||
|
There was a problem hiding this comment. This change saves 16 bytes per Expr. Nice. I will run some planning benchmarks and see if we can see any difference There was a problem hiding this comment. Thank you @alamb , i am doing more amazing experiment, try to reduce from 128 to 80, so we can save 48 bytes per Expr! |
||
| assert_eq!(size_of::<ScalarValue>(), 64); | ||
| assert_eq!(size_of::<DataType>(), 24); // 3 ptrs | ||
| assert_eq!(size_of::<Vec<Expr>>(), 24); | ||
|
|
||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Another change that might be a bit less impactful would be to box the fields of Alias instead
So Alias {
expr: Box
..
}