Introduce TypeSignatureClass::Any (#19485)

## Which issue does this PR close?

<!--
We generally require a GitHub issue to be filed for all bug fixes and
enhancements and this helps us generate change logs for our releases.
You can link an issue to this PR using the GitHub syntax. For example
`Closes #123` indicates that this PR will close issue #123.
-->

- Closes #19438

## Rationale for this change

<!--
Why are you proposing this change? If this is already explained clearly
in the issue then this section is not needed.
Explaining clearly why changes are proposed helps reviewers understand
your changes and offer better suggestions for fixes.
-->

See issue.

## What changes are included in this PR?

<!--
There is no need to duplicate the description in the issue here but it
is sometimes worth providing a summary of the individual changes in this
PR.
-->

Add new `TypeSignatureClass` variant `Any` and refactor `arrow_typeof`
and `arrow_metadata` function signatures to use this.

## Are these changes tested?

<!--
We typically require tests for all PRs in order to:
1. Prevent the code from being accidentally broken by subsequent changes
2. Serve as another way to document the expected behavior of the code

If tests are not included in your PR, please explain why (for example,
are they covered by existing tests)?
-->

Existing tests.

## Are there any user-facing changes?

<!--
If there are user-facing changes then we may require documentation to be
updated before approving the PR.
-->

No.

<!--
If there are any breaking changes to public APIs, please add the `api
change` label.
-->

Author: Jefffrey

datafusion/expr-common/src/signature.rs

@@ -332,6 +332,8 @@ impl TypeSignature {
 /// arguments that can be coerced to a particular class of types.
 #[derive(Debug, Clone, Eq, PartialEq, PartialOrd, Hash)]
 pub enum TypeSignatureClass {
+    /// Allows an arbitrary type argument without coercing the argument.
+    Any,
     /// Timestamps, allowing arbitrary (or no) timezones
     Timestamp,
     /// All time types
@@ -367,6 +369,9 @@ impl TypeSignatureClass {
     /// documentation or error messages.
     fn get_example_types(&self) -> Vec<DataType> {
         match self {
+            // TODO: might be too much info to return every single type here
+            //       maybe https://github.com/apache/datafusion/issues/14761 will help here?
+            TypeSignatureClass::Any => vec![],
             TypeSignatureClass::Native(l) => get_data_types(l.native()),
             TypeSignatureClass::Timestamp => {
                 vec![
@@ -409,6 +414,7 @@ impl TypeSignatureClass {
         }
 
         match self {
+            TypeSignatureClass::Any => true,
             TypeSignatureClass::Native(t) if t.native() == logical_type => true,
             TypeSignatureClass::Timestamp if logical_type.is_timestamp() => true,
             TypeSignatureClass::Time if logical_type.is_time() => true,
@@ -430,6 +436,7 @@ impl TypeSignatureClass {
         origin_type: &DataType,
     ) -> Result<DataType> {
         match self {
+            TypeSignatureClass::Any => Ok(origin_type.to_owned()),
             TypeSignatureClass::Native(logical_type) => {
                 logical_type.native().default_cast_for(origin_type)
             }

datafusion/functions/src/core/arrow_cast.rs

@@ -19,6 +19,7 @@
 
 use arrow::datatypes::{DataType, Field, FieldRef};
 use arrow::error::ArrowError;
+use datafusion_common::types::logical_string;
 use datafusion_common::{
     Result, ScalarValue, arrow_datafusion_err, exec_err, internal_err,
 };
@@ -27,8 +28,8 @@ use std::any::Any;
 
 use datafusion_expr::simplify::{ExprSimplifyResult, SimplifyInfo};
 use datafusion_expr::{
-    ColumnarValue, Documentation, Expr, ReturnFieldArgs, ScalarFunctionArgs,
-    ScalarUDFImpl, Signature, Volatility,
+    Coercion, ColumnarValue, Documentation, Expr, ReturnFieldArgs, ScalarFunctionArgs,
+    ScalarUDFImpl, Signature, TypeSignatureClass, Volatility,
 };
 use datafusion_macros::user_doc;
 
@@ -102,7 +103,13 @@ impl Default for ArrowCastFunc {
 impl ArrowCastFunc {
     pub fn new() -> Self {
         Self {
-            signature: Signature::any(2, Volatility::Immutable),
+            signature: Signature::coercible(
+                vec![
+                    Coercion::new_exact(TypeSignatureClass::Any),
+                    Coercion::new_exact(TypeSignatureClass::Native(logical_string())),
+                ],
+                Volatility::Immutable,
+            ),
         }
     }
 }

datafusion/functions/src/core/arrow_metadata.rs

@@ -17,10 +17,11 @@
 
 use arrow::array::{MapBuilder, StringBuilder};
 use arrow::datatypes::{DataType, Field, Fields};
-use datafusion_common::{Result, ScalarValue, exec_err};
+use datafusion_common::types::logical_string;
+use datafusion_common::{Result, ScalarValue, exec_err, internal_err};
 use datafusion_expr::{
-    ColumnarValue, Documentation, ScalarFunctionArgs, ScalarUDFImpl, Signature,
-    Volatility,
+    Coercion, ColumnarValue, Documentation, ScalarFunctionArgs, ScalarUDFImpl, Signature,
+    TypeSignature, TypeSignatureClass, Volatility,
 };
 use datafusion_macros::user_doc;
 use std::any::Any;
@@ -29,7 +30,7 @@ use std::sync::Arc;
 #[user_doc(
     doc_section(label = "Other Functions"),
     description = "Returns the metadata of the input expression. If a key is provided, returns the value for that key. If no key is provided, returns a Map of all metadata.",
-    syntax_example = "arrow_metadata(expression, [key])",
+    syntax_example = "arrow_metadata(expression[, key])",
     sql_example = r#"```sql
 > select arrow_metadata(col) from table;
 +----------------------------+
@@ -61,7 +62,18 @@ pub struct ArrowMetadataFunc {
 impl ArrowMetadataFunc {
     pub fn new() -> Self {
         Self {
-            signature: Signature::variadic_any(Volatility::Immutable),
+            signature: Signature::one_of(
+                vec![
+                    TypeSignature::Coercible(vec![Coercion::new_exact(
+                        TypeSignatureClass::Any,
+                    )]),
+                    TypeSignature::Coercible(vec![
+                        Coercion::new_exact(TypeSignatureClass::Any),
+                        Coercion::new_exact(TypeSignatureClass::Native(logical_string())),
+                    ]),
+                ],
+                Volatility::Immutable,
+            ),
         }
     }
 }
@@ -105,7 +117,7 @@ impl ScalarUDFImpl for ArrowMetadataFunc {
                 false,
             ))
         } else {
-            exec_err!("arrow_metadata requires 1 or 2 arguments")
+            internal_err!("arrow_metadata requires 1 or 2 arguments")
         }
     }
 
@@ -114,7 +126,7 @@ impl ScalarUDFImpl for ArrowMetadataFunc {
 
         if args.args.len() == 2 {
             let key = match &args.args[1] {
-                ColumnarValue::Scalar(ScalarValue::Utf8(Some(k))) => k,
+                ColumnarValue::Scalar(ScalarValue::Utf8(Some(key))) => key,
                 _ => {
                     return exec_err!(
                         "Second argument to arrow_metadata must be a string literal key"
@@ -142,7 +154,7 @@ impl ScalarUDFImpl for ArrowMetadataFunc {
                 &map_array, 0,
             )?))
         } else {
-            exec_err!("arrow_metadata requires 1 or 2 arguments")
+            internal_err!("arrow_metadata requires 1 or 2 arguments")
         }
     }
 }

datafusion/sqllogictest/test_files/arrow_typeof.slt

@@ -95,9 +95,12 @@ SELECT arrow_cast('1', 'Int16')
 query error
 SELECT arrow_cast('1')
 
-query error DataFusion error: Execution error: arrow_cast requires its second argument to be a non\-empty constant string
+query error Expect TypeSignatureClass::Native\(LogicalType\(Native\(String\), String\)\) but received NativeType::Int64, DataType: Int64
 SELECT arrow_cast('1', 43)
 
+query error DataFusion error: Execution error: arrow_cast requires its second argument to be a non\-empty constant string
+SELECT arrow_cast('1', arrow_cast('Utf8', 'Utf8'))
+
 query error DataFusion error: Execution error: Unsupported type 'unknown'\. Must be a supported arrow type name such as 'Int32' or 'Timestamp\(ns\)'\. Error unknown token: unknown
 SELECT arrow_cast('1', 'unknown')
 

docs/source/user-guide/sql/scalar_functions.md

@@ -5047,7 +5047,7 @@ arrow_cast(expression, datatype)
 Returns the metadata of the input expression. If a key is provided, returns the value for that key. If no key is provided, returns a Map of all metadata.
 
 ```sql
-arrow_metadata(expression, [key])
+arrow_metadata(expression[, key])
 ```
 
 #### Arguments