Faster BinaryArray to StringArray conversion (#3168)

* Faster ByteArray to StringArray conversion

* Add benchmark

* Fix logical conflict

arrow-array/src/array/string_array.rs

@@ -216,8 +216,22 @@ impl<OffsetSize: OffsetSizeTrait> From<GenericBinaryArray<OffsetSize>>
     for GenericStringArray<OffsetSize>
 {
     fn from(v: GenericBinaryArray<OffsetSize>) -> Self {
+        let offsets = v.value_offsets();
+        let values = v.value_data();
+
+        // We only need to validate that all values are valid UTF-8
+        let validated = std::str::from_utf8(values).expect("Invalid UTF-8 sequence");
+        for offset in offsets.iter() {
+            assert!(
+                validated.is_char_boundary(offset.as_usize()),
+                "Invalid UTF-8 sequence"
+            )
+        }
+
         let builder = v.into_data().into_builder().data_type(Self::DATA_TYPE);
-        Self::from(builder.build().unwrap())
+        // SAFETY:
+        // Validated UTF-8 above
+        Self::from(unsafe { builder.build_unchecked() })
     }
 }
 

arrow/benches/array_data_validate.rs

@@ -52,6 +52,12 @@ fn validate_benchmark(c: &mut Criterion) {
     c.bench_function("validate_utf8_array_data 20000", |b| {
         b.iter(|| validate_utf8_array(&str_arr))
     });
+
+    let byte_array =
+        BinaryArray::from_iter_values(std::iter::repeat(b"test").take(20000));
+    c.bench_function("byte_array_to_string_array 20000", |b| {
+        b.iter(|| StringArray::from(BinaryArray::from(byte_array.data().clone())))
+    });
 }
 
 criterion_group!(benches, validate_benchmark);

arrow/src/row/mod.rs

@@ -1425,7 +1425,7 @@ mod tests {
     }
 
     #[test]
-    #[should_panic(expected = "Invalid UTF8 sequence at string")]
+    #[should_panic(expected = "Invalid UTF-8 sequence")]
     fn test_invalid_utf8() {
         let mut converter =
             RowConverter::new(vec![SortField::new(DataType::Binary)]).unwrap();