[arrow-cast] Improve support for casting string view to numeric

Signed-off-by: Tai Le Manh <[email protected]>

arrow-array/src/array/mod.rs

@@ -76,6 +76,8 @@ mod list_view_array;
 
 pub use list_view_array::*;
 
+use crate::iterator::ArrayIter;
+
 /// An array in the [arrow columnar format](https://arrow.apache.org/docs/format/Columnar.html)
 pub trait Array: std::fmt::Debug + Send + Sync {
     /// Returns the array as [`Any`] so that it can be
@@ -570,6 +572,40 @@ pub trait ArrayAccessor: Array {
     unsafe fn value_unchecked(&self, index: usize) -> Self::Item;
 }
 
+/// A trait for Arrow String Arrays, currently three types are supported:
+/// - `StringArray`
+/// - `LargeStringArray`
+/// - `StringViewArray`
+///
+/// This trait helps to abstract over the different types of string arrays
+/// so that we don't need to duplicate the implementation for each type.
+pub trait StringArrayType<'a>: ArrayAccessor<Item = &'a str> + Sized {
+    /// Returns true if all data within this string array is ASCII
+    fn is_ascii(&self) -> bool;
+
+    /// Constructs a new iterator
+    fn iter(&self) -> ArrayIter<Self>;
+}
+
+impl<'a, O: OffsetSizeTrait> StringArrayType<'a> for &'a GenericStringArray<O> {
+    fn is_ascii(&self) -> bool {
+        GenericStringArray::<O>::is_ascii(self)
+    }
+
+    fn iter(&self) -> ArrayIter<Self> {
+        GenericStringArray::<O>::iter(self)
+    }
+}
+impl<'a> StringArrayType<'a> for &'a StringViewArray {
+    fn is_ascii(&self) -> bool {
+        StringViewArray::is_ascii(self)
+    }
+
+    fn iter(&self) -> ArrayIter<Self> {
+        StringViewArray::iter(self)
+    }
+}
+
 impl PartialEq for dyn Array + '_ {
     fn eq(&self, other: &Self) -> bool {
         self.to_data().eq(&other.to_data())

arrow-cast/src/cast/mod.rs

@@ -59,7 +59,6 @@ use crate::parse::{
     parse_interval_day_time, parse_interval_month_day_nano, parse_interval_year_month,
     string_to_datetime, Parser,
 };
-use arrow_array::iterator::ArrayIter;
 use arrow_array::{builder::*, cast::*, temporal_conversions::*, timezone::Tz, types::*, *};
 use arrow_buffer::{i256, ArrowNativeType, OffsetBuffer};
 use arrow_data::transform::MutableArrayData;
@@ -2484,21 +2483,6 @@ where
     Ok(Arc::new(byte_array_builder.finish()))
 }
 
-trait StringArrayType<'a>: ArrayAccessor<Item = &'a str> + Sized {
-    /// Constructs a new iterator
-    fn iter(&self) -> ArrayIter<Self>;
-}
-impl<'a, O: OffsetSizeTrait> StringArrayType<'a> for &'a GenericStringArray<O> {
-    fn iter(&self) -> ArrayIter<Self> {
-        GenericStringArray::<O>::iter(self)
-    }
-}
-impl<'a> StringArrayType<'a> for &'a StringViewArray {
-    fn iter(&self) -> ArrayIter<Self> {
-        StringViewArray::iter(self)
-    }
-}
-
 #[cfg(test)]
 mod tests {
     use super::*;