feat: Update sqlparser-rs, enabling "LEFT" keyword to be optional f…

…or anti/semi joins
pola-rs · Jan 6, 2025 · 804944f · 804944f
1 parent 11dd4b3
commit 804944f
Show file tree

Hide file tree

Showing 6 changed files with 25 additions and 10 deletions.
diff --git a/Cargo.lock b/Cargo.lock
diff --git a/Cargo.toml b/Cargo.toml
@@ -77,7 +77,7 @@ serde_json = "1"
 simd-json = { version = "0.14", features = ["known-key"] }
 simdutf8 = "0.1.4"
 slotmap = "1"
-sqlparser = "0.52"
+sqlparser = "0.53"
 stacker = "0.1"
 streaming-iterator = "0.1.9"
 strength_reduce = "0.2"

diff --git a/crates/polars-sql/src/context.rs b/crates/polars-sql/src/context.rs
@@ -566,6 +566,8 @@ impl SQLContext {
                     | JoinOperator::LeftOuter(constraint)
                     | JoinOperator::RightOuter(constraint)
                     | JoinOperator::Inner(constraint)
+                    | JoinOperator::Anti(constraint)
+                    | JoinOperator::Semi(constraint)
                     | JoinOperator::LeftAnti(constraint)
                     | JoinOperator::LeftSemi(constraint)
                     | JoinOperator::RightAnti(constraint)
@@ -592,9 +594,9 @@ impl SQLContext {
                                 JoinOperator::RightOuter(_) => JoinType::Right,
                                 JoinOperator::Inner(_) => JoinType::Inner,
                                 #[cfg(feature = "semi_anti_join")]
-                                JoinOperator::LeftAnti(_) | JoinOperator::RightAnti(_) => JoinType::Anti,
+                                JoinOperator::Anti(_) | JoinOperator::LeftAnti(_) | JoinOperator::RightAnti(_) => JoinType::Anti,
                                 #[cfg(feature = "semi_anti_join")]
-                                JoinOperator::LeftSemi(_) | JoinOperator::RightSemi(_) => JoinType::Semi,
+                                JoinOperator::Semi(_) | JoinOperator::LeftSemi(_) | JoinOperator::RightSemi(_) => JoinType::Semi,
                                 join_type => polars_bail!(SQLInterface: "join type '{:?}' not currently supported", join_type),
                             },
                         )?
@@ -1028,10 +1030,10 @@ impl SQLContext {
                         .columns
                         .iter()
                         .map(|c| {
-                            if c.value.is_empty() {
+                            if c.name.value.is_empty() {
                                 None
                             } else {
-                                Some(PlSmallStr::from_str(c.value.as_str()))
+                                Some(PlSmallStr::from_str(c.name.value.as_str()))
                             }
                         })
                         .collect();
@@ -1399,7 +1401,8 @@ impl SQLContext {
                 )
             } else {
                 let existing_columns: Vec<_> = schema.iter_names().collect();
-                let new_columns: Vec<_> = alias.columns.iter().map(|c| c.value.clone()).collect();
+                let new_columns: Vec<_> =
+                    alias.columns.iter().map(|c| c.name.value.clone()).collect();
                 Ok(lf.rename(existing_columns, new_columns, true))
             }
         }

diff --git a/crates/polars-sql/src/functions.rs b/crates/polars-sql/src/functions.rs
@@ -13,11 +13,13 @@ use polars_plan::plans::{typed_lit, LiteralValue};
 use polars_plan::prelude::LiteralValue::Null;
 use polars_plan::prelude::{col, cols, lit, StrptimeOptions};
 use polars_utils::pl_str::PlSmallStr;
+use sqlparser::ast::helpers::attached_token::AttachedToken;
 use sqlparser::ast::{
     DateTimeField, DuplicateTreatment, Expr as SQLExpr, Function as SQLFunction, FunctionArg,
     FunctionArgExpr, FunctionArgumentClause, FunctionArgumentList, FunctionArguments, Ident,
     OrderByExpr, Value as SQLValue, WindowSpec, WindowType,
 };
+use sqlparser::tokenizer::Span;
 
 use crate::sql_expr::{adjust_one_indexed_param, parse_extract_date_part, parse_sql_expr};
 use crate::SQLContext;
@@ -1058,6 +1060,7 @@ impl SQLFunctionVisitor<'_> {
                             &DateTimeField::Custom(Ident {
                                 value: p.to_string(),
                                 quote_style: None,
+                                span: Span::empty(),
                             }),
                         )
                     },
@@ -1511,7 +1514,7 @@ impl SQLFunctionVisitor<'_> {
                 f(parse_sql_expr(sql_expr, self.ctx, self.active_schema)?)
             },
             [FunctionArgExpr::Wildcard] => f(parse_sql_expr(
-                &SQLExpr::Wildcard,
+                &SQLExpr::Wildcard(AttachedToken::empty()),
                 self.ctx,
                 self.active_schema,
             )?),
@@ -1813,6 +1816,7 @@ fn _extract_func_args(
                     .iter()
                     .map(|arg| match arg {
                         FunctionArg::Named { arg, .. } => arg,
+                        FunctionArg::ExprNamed { arg, .. } => arg,
                         FunctionArg::Unnamed(arg) => arg,
                     })
                     .collect();

diff --git a/crates/polars-sql/src/sql_expr.rs b/crates/polars-sql/src/sql_expr.rs
@@ -246,7 +246,7 @@ impl SQLExprVisitor<'_> {
             },
             SQLExpr::UnaryOp { op, expr } => self.visit_unary_op(op, expr),
             SQLExpr::Value(value) => self.visit_literal(value),
-            SQLExpr::Wildcard => Ok(Expr::Wildcard),
+            SQLExpr::Wildcard(_) => Ok(Expr::Wildcard),
             e @ SQLExpr::Case { .. } => self.visit_case_when_then(e),
             other => {
                 polars_bail!(SQLInterface: "expression {:?} is not currently supported", other)

diff --git a/py-polars/tests/unit/sql/test_joins.py b/py-polars/tests/unit/sql/test_joins.py
@@ -23,6 +23,10 @@ def foods_ipc_path() -> Path:
             "SELECT * FROM tbl_a LEFT SEMI JOIN tbl_b USING (a,c)",
             pl.DataFrame({"a": [2], "b": [0], "c": ["y"]}),
         ),
+        (
+            "SELECT * FROM tbl_a SEMI JOIN tbl_b USING (a,c)",
+            pl.DataFrame({"a": [2], "b": [0], "c": ["y"]}),
+        ),
         (
             "SELECT * FROM tbl_a LEFT SEMI JOIN tbl_b USING (a)",
             pl.DataFrame({"a": [1, 2, 3], "b": [4, 0, 6], "c": ["w", "y", "z"]}),
@@ -31,6 +35,10 @@ def foods_ipc_path() -> Path:
             "SELECT * FROM tbl_a LEFT ANTI JOIN tbl_b USING (a)",
             pl.DataFrame(schema={"a": pl.Int64, "b": pl.Int64, "c": pl.String}),
         ),
+        (
+            "SELECT * FROM tbl_a ANTI JOIN tbl_b USING (a)",
+            pl.DataFrame(schema={"a": pl.Int64, "b": pl.Int64, "c": pl.String}),
+        ),
         (
             "SELECT * FROM tbl_a LEFT SEMI JOIN tbl_b USING (b) LEFT SEMI JOIN tbl_c USING (c)",
             pl.DataFrame({"a": [1, 3], "b": [4, 6], "c": ["w", "z"]}),