Implement Spanned to retrieve source locations on AST nodes

src/ast/mod.rs

@@ -28,6 +28,8 @@ use serde::{Deserialize, Serialize};
 #[cfg(feature = "visitor")]
 use sqlparser_derive::{Visit, VisitMut};
 
+use crate::tokenizer::{Span, TokenWithLocation};
+
 pub use self::data_type::{
     ArrayElemTypeDef, CharLengthUnits, CharacterLength, DataType, ExactNumberInfo,
     StructBracketKind, TimezoneInfo,
@@ -79,6 +81,9 @@ mod dml;
 pub mod helpers;
 mod operator;
 mod query;
+mod spans;
+pub use spans::{union_spans, Spanned};
+
 mod trigger;
 mod value;
 
@@ -123,7 +128,7 @@ where
 }
 
 /// An identifier, decomposed into its value or character data and the quote style.
-#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
+#[derive(Debug, Clone, PartialOrd, Ord)]
 #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
 #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
 pub struct Ident {
@@ -132,8 +137,25 @@ pub struct Ident {
     /// The starting quote if any. Valid quote characters are the single quote,
     /// double quote, backtick, and opening square bracket.
     pub quote_style: Option<char>,
+    /// The span of the identifier in the original SQL string.
+    pub span: Span,
+}
+
+impl PartialEq for Ident {
+    fn eq(&self, other: &Self) -> bool {
+        self.value == other.value && self.quote_style == other.quote_style
+    }
+}
+
+impl core::hash::Hash for Ident {
+    fn hash<H: std::hash::Hasher>(&self, state: &mut H) {
+        self.value.hash(state);
+        self.quote_style.hash(state);
+    }
 }
 
+impl Eq for Ident {}
+
 impl Ident {
     /// Create a new identifier with the given value and no quotes.
     pub fn new<S>(value: S) -> Self
@@ -143,6 +165,7 @@ impl Ident {
         Ident {
             value: value.into(),
             quote_style: None,
+            span: Span::empty(),
         }
     }
 
@@ -156,6 +179,30 @@ impl Ident {
         Ident {
             value: value.into(),
             quote_style: Some(quote),
+            span: Span::empty(),
+        }
+    }
+
+    pub fn with_span<S>(span: Span, value: S) -> Self
+    where
+        S: Into<String>,
+    {
+        Ident {
+            value: value.into(),
+            quote_style: None,
+            span,
+        }
+    }
+
+    pub fn with_quote_and_span<S>(quote: char, span: Span, value: S) -> Self
+    where
+        S: Into<String>,
+    {
+        assert!(quote == '\'' || quote == '"' || quote == '`' || quote == '[');
+        Ident {
+            value: value.into(),
+            quote_style: Some(quote),
+            span,
         }
     }
 }
@@ -165,6 +212,7 @@ impl From<&str> for Ident {
         Ident {
             value: value.to_string(),
             quote_style: None,
+            span: Span::empty(),
         }
     }
 }
@@ -881,10 +929,10 @@ pub enum Expr {
         /// `<search modifier>`
         opt_search_modifier: Option<SearchModifier>,
     },
-    Wildcard,
+    Wildcard(TokenWithLocation),
     /// Qualified wildcard, e.g. `alias.*` or `schema.table.*`.
     /// (Same caveats apply to `QualifiedWildcard` as to `Wildcard`.)
-    QualifiedWildcard(ObjectName),
+    QualifiedWildcard(ObjectName, TokenWithLocation),
     /// Some dialects support an older syntax for outer joins where columns are
     /// marked with the `(+)` operator in the WHERE clause, for example:
     ///
@@ -1173,8 +1221,8 @@ impl fmt::Display for Expr {
             Expr::MapAccess { column, keys } => {
                 write!(f, "{column}{}", display_separated(keys, ""))
             }
-            Expr::Wildcard => f.write_str("*"),
-            Expr::QualifiedWildcard(prefix) => write!(f, "{}.*", prefix),
+            Expr::Wildcard(_) => f.write_str("*"),
+            Expr::QualifiedWildcard(prefix, _) => write!(f, "{}.*", prefix),
             Expr::CompoundIdentifier(s) => write!(f, "{}", display_separated(s, ".")),
             Expr::IsTrue(ast) => write!(f, "{ast} IS TRUE"),
             Expr::IsNotTrue(ast) => write!(f, "{ast} IS NOT TRUE"),
@@ -5169,8 +5217,8 @@ pub enum FunctionArgExpr {
 impl From<Expr> for FunctionArgExpr {
     fn from(wildcard_expr: Expr) -> Self {
         match wildcard_expr {
-            Expr::QualifiedWildcard(prefix) => Self::QualifiedWildcard(prefix),
-            Expr::Wildcard => Self::Wildcard,
+            Expr::QualifiedWildcard(prefix, _) => Self::QualifiedWildcard(prefix),
+            Expr::Wildcard(_) => Self::Wildcard,
             expr => Self::Expr(expr),
         }
     }

src/ast/query.rs

@@ -19,7 +19,10 @@ use serde::{Deserialize, Serialize};
 #[cfg(feature = "visitor")]
 use sqlparser_derive::{Visit, VisitMut};
 
-use crate::ast::*;
+use crate::{
+    ast::*,
+    tokenizer::{Token, TokenWithLocation},
+};
 
 /// The most complete variant of a `SELECT` query expression, optionally
 /// including `WITH`, `UNION` / other set operations, and `ORDER BY`.
@@ -271,6 +274,8 @@ impl fmt::Display for Table {
 #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
 #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
 pub struct Select {
+    /// SELECT
+    pub select_token: TokenWithLocation,
     pub distinct: Option<Distinct>,
     /// MSSQL syntax: `TOP (<N>) [ PERCENT ] [ WITH TIES ]`
     pub top: Option<Top>,
@@ -490,6 +495,7 @@ impl fmt::Display for NamedWindowDefinition {
 #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
 #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
 pub struct With {
+    pub with_token: TokenWithLocation,
     pub recursive: bool,
     pub cte_tables: Vec<Cte>,
 }
@@ -541,6 +547,8 @@ pub struct Cte {
     pub query: Box<Query>,
     pub from: Option<Ident>,
     pub materialized: Option<CteAsMaterialized>,
+    // needed for accurate span reporting
+    pub closing_paren_token: TokenWithLocation,
 }
 
 impl fmt::Display for Cte {
@@ -592,10 +600,11 @@ impl fmt::Display for IdentWithAlias {
 }
 
 /// Additional options for wildcards, e.g. Snowflake `EXCLUDE`/`RENAME` and Bigquery `EXCEPT`.
-#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash, Default)]
+#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
 #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
 #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
 pub struct WildcardAdditionalOptions {
+    pub wildcard_token: TokenWithLocation,
     /// `[ILIKE...]`.
     ///  Snowflake syntax: <https://docs.snowflake.com/en/sql-reference/sql/select#parameters>
     pub opt_ilike: Option<IlikeSelectItem>,
@@ -613,6 +622,19 @@ pub struct WildcardAdditionalOptions {
     pub opt_rename: Option<RenameSelectItem>,
 }
 
+impl Default for WildcardAdditionalOptions {
+    fn default() -> Self {
+        Self {
+            wildcard_token: TokenWithLocation::wrap(Token::Mul),
+            opt_ilike: None,
+            opt_exclude: None,
+            opt_except: None,
+            opt_replace: None,
+            opt_rename: None,
+        }
+    }
+}
+
 impl fmt::Display for WildcardAdditionalOptions {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
         if let Some(ilike) = &self.opt_ilike {