jayzhan211 commented on code in PR #12097:
URL: https://github.com/apache/datafusion/pull/12097#discussion_r1728402666
##########
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:
https://github.com/apache/datafusion/pull/12125
I unnest the array and compare with eq kernel but the performance is worse.
I guess although the comparison is vectorized (eq kernel) but we need to copy
the value. However, it is faster than Row approach
```
array_has eq kernel approach
time: [913.21 µs 919.23 µs 926.38 µs]
Found 4 outliers among 100 measurements (4.00%)
2 (2.00%) high mild
2 (2.00%) high severe
array_has simple iterate approach
time: [640.01 µs 656.86 µs 687.51 µs]
Found 6 outliers among 100 measurements (6.00%)
3 (3.00%) high mild
3 (3.00%) high severe
array_has row approach time: [2.6768 ms 2.6913 ms 2.7062 ms]
Found 1 outliers among 100 measurements (1.00%)
1 (1.00%) high mild
```
--
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: github-unsubscr...@datafusion.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscr...@datafusion.apache.org
For additional commands, e-mail: github-h...@datafusion.apache.org
