use std::any::Any;
use std::cmp::{max, Ordering};
use std::sync::Arc;
use arrow::array::{ArrayRef, GenericStringArray, OffsetSizeTrait};
use arrow::datatypes::DataType;
use datafusion_common::cast::{as_generic_string_array, as_int64_array};
use datafusion_common::exec_err;
use datafusion_common::Result;
use datafusion_expr::TypeSignature::Exact;
use datafusion_expr::{ColumnarValue, ScalarUDFImpl, Signature, Volatility};
use crate::utils::{make_scalar_function, utf8_to_str_type};
#[derive(Debug)]
pub struct RightFunc {
signature: Signature,
}
impl Default for RightFunc {
fn default() -> Self {
Self::new()
}
}
impl RightFunc {
pub fn new() -> Self {
use DataType::*;
Self {
signature: Signature::one_of(
vec![Exact(vec![Utf8, Int64]), Exact(vec![LargeUtf8, Int64])],
Volatility::Immutable,
),
}
}
}
impl ScalarUDFImpl for RightFunc {
fn as_any(&self) -> &dyn Any {
self
}
fn name(&self) -> &str {
"right"
}
fn signature(&self) -> &Signature {
&self.signature
}
fn return_type(&self, arg_types: &[DataType]) -> Result<DataType> {
utf8_to_str_type(&arg_types[0], "right")
}
fn invoke(&self, args: &[ColumnarValue]) -> Result<ColumnarValue> {
match args[0].data_type() {
DataType::Utf8 => make_scalar_function(right::<i32>, vec![])(args),
DataType::LargeUtf8 => make_scalar_function(right::<i64>, vec![])(args),
other => exec_err!("Unsupported data type {other:?} for function right"),
}
}
}
pub fn right<T: OffsetSizeTrait>(args: &[ArrayRef]) -> Result<ArrayRef> {
let string_array = as_generic_string_array::<T>(&args[0])?;
let n_array = as_int64_array(&args[1])?;
let result = string_array
.iter()
.zip(n_array.iter())
.map(|(string, n)| match (string, n) {
(Some(string), Some(n)) => match n.cmp(&0) {
Ordering::Less => Some(
string
.chars()
.skip(n.unsigned_abs() as usize)
.collect::<String>(),
),
Ordering::Equal => Some("".to_string()),
Ordering::Greater => Some(
string
.chars()
.skip(max(string.chars().count() as i64 - n, 0) as usize)
.collect::<String>(),
),
},
_ => None,
})
.collect::<GenericStringArray<T>>();
Ok(Arc::new(result) as ArrayRef)
}
#[cfg(test)]
mod tests {
use arrow::array::{Array, StringArray};
use arrow::datatypes::DataType::Utf8;
use datafusion_common::{Result, ScalarValue};
use datafusion_expr::{ColumnarValue, ScalarUDFImpl};
use crate::unicode::right::RightFunc;
use crate::utils::test::test_function;
#[test]
fn test_functions() -> Result<()> {
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("abcde")),
ColumnarValue::Scalar(ScalarValue::from(2i64)),
],
Ok(Some("de")),
&str,
Utf8,
StringArray
);
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("abcde")),
ColumnarValue::Scalar(ScalarValue::from(200i64)),
],
Ok(Some("abcde")),
&str,
Utf8,
StringArray
);
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("abcde")),
ColumnarValue::Scalar(ScalarValue::from(-2i64)),
],
Ok(Some("cde")),
&str,
Utf8,
StringArray
);
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("abcde")),
ColumnarValue::Scalar(ScalarValue::from(-200i64)),
],
Ok(Some("")),
&str,
Utf8,
StringArray
);
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("abcde")),
ColumnarValue::Scalar(ScalarValue::from(0i64)),
],
Ok(Some("")),
&str,
Utf8,
StringArray
);
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::Utf8(None)),
ColumnarValue::Scalar(ScalarValue::from(2i64)),
],
Ok(None),
&str,
Utf8,
StringArray
);
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("abcde")),
ColumnarValue::Scalar(ScalarValue::Int64(None)),
],
Ok(None),
&str,
Utf8,
StringArray
);
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("joséésoj")),
ColumnarValue::Scalar(ScalarValue::from(5i64)),
],
Ok(Some("éésoj")),
&str,
Utf8,
StringArray
);
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("joséésoj")),
ColumnarValue::Scalar(ScalarValue::from(-3i64)),
],
Ok(Some("éésoj")),
&str,
Utf8,
StringArray
);
#[cfg(not(feature = "unicode_expressions"))]
test_function!(
RightFunc::new(),
&[
ColumnarValue::Scalar(ScalarValue::from("abcde")),
ColumnarValue::Scalar(ScalarValue::from(2i64)),
],
internal_err!(
"function right requires compilation with feature flag: unicode_expressions."
),
&str,
Utf8,
StringArray
);
Ok(())
}
}