JasonLi-cn commented on code in PR #9971:
URL: https://github.com/apache/arrow-datafusion/pull/9971#discussion_r1555930261
##########
datafusion/functions/src/string/common.rs:
##########
@@ -97,80 +101,141 @@ pub(crate) fn general_trim<T: OffsetSizeTrait>(
}
}
-/// applies a unary expression to `args[0]` that is expected to be
downcastable to
-/// a `GenericStringArray` and returns a `GenericStringArray` (which may have
a different offset)
-/// # Errors
-/// This function errors when:
-/// * the number of arguments is not 1
-/// * the first argument is not castable to a `GenericStringArray`
-pub(crate) fn unary_string_function<'a, T, O, F, R>(
- args: &[&'a dyn Array],
- op: F,
- name: &str,
-) -> Result<GenericStringArray<O>>
-where
- R: AsRef<str>,
- O: OffsetSizeTrait,
- T: OffsetSizeTrait,
- F: Fn(&'a str) -> R,
-{
- if args.len() != 1 {
- return exec_err!(
- "{:?} args were supplied but {} takes exactly one argument",
- args.len(),
- name
- );
- }
-
- let string_array = as_generic_string_array::<T>(args[0])?;
+pub(crate) fn to_lower(args: &[ColumnarValue], name: &str) ->
Result<ColumnarValue> {
+ case_conversion(args, |string| string.to_lowercase(), name)
+}
- // first map is the iterator, second is for the `Option<_>`
- Ok(string_array.iter().map(|string| string.map(&op)).collect())
+pub(crate) fn to_upper(args: &[ColumnarValue], name: &str) ->
Result<ColumnarValue> {
+ case_conversion(args, |string| string.to_uppercase(), name)
}
-pub(crate) fn handle<'a, F, R>(
+fn case_conversion<'a, F>(
args: &'a [ColumnarValue],
op: F,
name: &str,
) -> Result<ColumnarValue>
where
- R: AsRef<str>,
- F: Fn(&'a str) -> R,
+ F: Fn(&'a str) -> String,
{
match &args[0] {
- ColumnarValue::Array(a) => match a.data_type() {
+ ColumnarValue::Array(array) => match array.data_type() {
DataType::Utf8 => {
- Ok(ColumnarValue::Array(Arc::new(unary_string_function::<
- i32,
- i32,
- _,
- _,
- >(
- &[a.as_ref()], op, name
- )?)))
+ Ok(ColumnarValue::Array(convert_array::<i32, _>(array, op)?))
}
DataType::LargeUtf8 => {
- Ok(ColumnarValue::Array(Arc::new(unary_string_function::<
- i64,
- i64,
- _,
- _,
- >(
- &[a.as_ref()], op, name
- )?)))
+ Ok(ColumnarValue::Array(convert_array::<i64, _>(array, op)?))
}
other => exec_err!("Unsupported data type {other:?} for function
{name}"),
},
ColumnarValue::Scalar(scalar) => match scalar {
ScalarValue::Utf8(a) => {
- let result = a.as_ref().map(|x| (op)(x).as_ref().to_string());
+ let result = a.as_ref().map(|x| op(x));
Ok(ColumnarValue::Scalar(ScalarValue::Utf8(result)))
}
ScalarValue::LargeUtf8(a) => {
- let result = a.as_ref().map(|x| (op)(x).as_ref().to_string());
+ let result = a.as_ref().map(|x| op(x));
Ok(ColumnarValue::Scalar(ScalarValue::LargeUtf8(result)))
}
other => exec_err!("Unsupported data type {other:?} for function
{name}"),
},
}
}
+
+fn convert_array<'a, O, F>(array: &'a ArrayRef, op: F) -> Result<ArrayRef>
+where
+ O: OffsetSizeTrait,
+ F: Fn(&'a str) -> String,
+{
+ let string_array = as_generic_string_array::<O>(array)?;
+ let item_len = string_array.len();
+
+ // Find the ASCII string at the beginning.
+ let mut i = 0;
+ while i < item_len {
+ let item = unsafe { string_array.value_unchecked(i) };
Review Comment:
`iter()` always checks for bounds and null, but we don't need these checks
here, so theoretically, using iter() would result in a performance loss.
```rust
impl<T: ArrayAccessor> Iterator for ArrayIter<T> {
type Item = Option<T::Item>;
#[inline]
fn next(&mut self) -> Option<Self::Item> {
if self.current == self.current_end {
None
} else if self.is_null(self.current) {
self.current += 1;
Some(None)
} else {
let old = self.current;
self.current += 1;
// Safety:
// we just checked bounds in `self.current_end == self.current`
// this is safe on the premise that this struct is initialized
with
// current = array.len()
// and that current_end is ever only decremented
unsafe { Some(Some(self.array.value_unchecked(old))) }
}
}
```
--
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]
