alamb commented on code in PR #12097:
URL: https://github.com/apache/datafusion/pull/12097#discussion_r1729194704
##########
datafusion/functions-nested/src/array_has.rs:
##########
@@ -251,75 +237,176 @@ impl ScalarUDFImpl for ArrayHasAny {
}
/// Represents the type of comparison for array_has.
-#[derive(Debug, PartialEq)]
+#[derive(Debug, PartialEq, Clone, Copy)]
enum ComparisonType {
// array_has_all
All,
// array_has_any
Any,
- // array_has
- Single,
}
-fn general_array_has_dispatch<O: OffsetSizeTrait>(
+fn array_has_dispatch<O: OffsetSizeTrait>(
+ haystack: &ArrayRef,
+ needle: &ArrayRef,
+) -> Result<ArrayRef> {
+ let haystack = as_generic_list_array::<O>(haystack)?;
+ match needle.data_type() {
+ DataType::Utf8 | DataType::LargeUtf8 | DataType::Utf8View => {
+ array_has_string_internal::<O>(haystack, needle)
+ }
+ _ => general_array_has::<O>(haystack, needle),
+ }
+}
+
+fn array_has_all_and_any_dispatch<O: OffsetSizeTrait>(
haystack: &ArrayRef,
needle: &ArrayRef,
comparison_type: ComparisonType,
) -> Result<ArrayRef> {
- let array = if comparison_type == ComparisonType::Single {
- let arr = as_generic_list_array::<O>(haystack)?;
- check_datatypes("array_has", &[arr.values(), needle])?;
- arr
- } else {
- check_datatypes("array_has", &[haystack, needle])?;
- as_generic_list_array::<O>(haystack)?
- };
+ let haystack = as_generic_list_array::<O>(haystack)?;
+ let needle = as_generic_list_array::<O>(needle)?;
+ match needle.data_type() {
+ DataType::Utf8 | DataType::LargeUtf8 | DataType::Utf8View => {
+ array_has_all_and_any_string_internal::<O>(haystack, needle,
comparison_type)
+ }
+ _ => general_array_has_for_all_and_any::<O>(haystack, needle,
comparison_type),
+ }
+}
- let mut boolean_builder = BooleanArray::builder(array.len());
+fn array_has_string_internal<O: OffsetSizeTrait>(
+ haystack: &GenericListArray<O>,
+ needle: &ArrayRef,
+) -> Result<ArrayRef> {
+ let mut boolean_builder = BooleanArray::builder(haystack.len());
+ for (arr, element) in
haystack.iter().zip(string_array_to_vec(needle).into_iter()) {
+ match (arr, element) {
+ (Some(arr), Some(element)) => {
+ boolean_builder.append_value(
+ string_array_to_vec(&arr)
Review Comment:
Thanks @jayzhan211 -- when reviewing
https://github.com/apache/datafusion/pull/12125 and re-reading
https://github.com/apache/datafusion/issues/12062, I wonder if what would
really be useful is making specialized implementations when the second argument
is a constant?
Spefically, what I think what will happen now is `array_has(col, 'foo')`
will make a `StringArray` that has the string `'foo'` repeated many times
I wonder how much better implementing a special implementation for
`array_has(array: ArrayRef, val: ScalarValue)` would be? I bet *much* faster
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
