fix: allow utf8 and largeutf8 to be cast into utf8view

apache · Aug 2, 2024 · f38b363 · f38b363
1 parent 65ba700
commit f38b363
Show file tree

Hide file tree

Showing 3 changed files with 16 additions and 1 deletion.
diff --git a/datafusion/expr/src/expr_schema.rs b/datafusion/expr/src/expr_schema.rs
@@ -148,6 +148,7 @@ impl ExprSchemable for Expr {
                     .iter()
                     .map(|e| e.get_type(schema))
                     .collect::<Result<Vec<_>>>()?;
+
                 // verify that function is invoked with correct number and type of arguments as defined in `TypeSignature`
                 data_types_with_scalar_udf(&arg_data_types, func).map_err(|err| {
                     plan_datafusion_err!(

diff --git a/datafusion/expr/src/type_coercion/functions.rs b/datafusion/expr/src/type_coercion/functions.rs
@@ -573,6 +573,8 @@ fn coerced_from<'a>(
         (Interval(_), _) if matches!(type_from, Utf8 | LargeUtf8) => {
             Some(type_into.clone())
         }
+        // We can go into a Utf8View from a Utf8 or LargeUtf8
+        (Utf8View, _) if matches!(type_from, Utf8 | LargeUtf8) => Some(type_into.clone()),
         // Any type can be coerced into strings
         (Utf8 | LargeUtf8, _) => Some(type_into.clone()),
         (Null, _) if can_cast_types(type_from, type_into) => Some(type_into.clone()),
@@ -636,6 +638,18 @@ mod tests {
     use super::*;
     use arrow::datatypes::Field;
 
+    #[test]
+    fn test_string_conversion() {
+        let cases = vec![
+            (DataType::Utf8View, DataType::Utf8, true),
+            (DataType::Utf8View, DataType::LargeUtf8, true),
+        ];
+
+        for case in cases {
+            assert_eq!(can_coerce_from(&case.0, &case.1), case.2);
+        }
+    }
+
     #[test]
     fn test_maybe_data_types() {
         // this vec contains: arg1, arg2, expected result

diff --git a/datafusion/sqllogictest/test_files/string_view.slt b/datafusion/sqllogictest/test_files/string_view.slt
@@ -364,7 +364,7 @@ EXPLAIN SELECT
 FROM test;
 ----
 logical_plan
-01)Projection: starts_with(CAST(test.column1_utf8view AS Utf8), Utf8("foo")) AS c, starts_with(test.column1_utf8view, test.column2_utf8view) AS c2
+01)Projection: starts_with(test.column1_utf8view, Utf8View("foo")) AS c, starts_with(test.column1_utf8view, test.column2_utf8view) AS c2
 02)--TableScan: test projection=[column1_utf8view, column2_utf8view]
 >>>>>>> 222263c15 (simplify string view handling)