alamb commented on code in PR #7896:
URL: https://github.com/apache/arrow-rs/pull/7896#discussion_r2200583931
##########
parquet-variant/src/builder.rs:
##########
@@ -2171,4 +2171,116 @@ mod tests {
let variant = Variant::try_new_with_metadata(metadata,
&value).unwrap();
assert_eq!(variant, Variant::Int8(2));
}
+
+ fn test_variant_object_with_count(count: i32,
expected_field_id_size_minus_one: u8) {
+ let mut builder = VariantBuilder::new();
+ let mut obj = builder.new_object();
+
+ for val in 0..count {
+ let key = format!("id_{}", val);
+ obj.insert(&key, val);
+ }
+
+ obj.finish().unwrap();
+ let (metadata, value) = builder.finish();
+ let variant = Variant::try_new(&metadata, &value).unwrap();
+
+ if let Variant::Object(obj) = variant {
+ assert_eq!(obj.len(), count as usize);
+ assert_eq!(obj.get(&format!("id_{}", 0)).unwrap(),
Variant::Int32(0));
+ assert_eq!(
+ obj.get(&format!("id_{}", count - 1)).unwrap(),
+ Variant::Int32(count - 1)
+ );
+
+ let header_byte = first_byte_from_slice(&value).unwrap();
+ let value_header = header_byte >> 2;
+ let field_id_size_minus_one = (value_header >> 2) & 0x03;
+ assert_eq!(
+ field_id_size_minus_one, expected_field_id_size_minus_one,
+ "Expected {}-byte field IDs",
+ expected_field_id_size_minus_one
+ );
+ } else {
+ panic!("Expected object variant");
+ }
+ }
+
+ #[test]
+ fn test_variant_object_257_elements() {
+ test_variant_object_with_count(2_i32.pow(8) + 1, 1); // 2^8 + 1,
expected 2-byte field IDs
+ }
+
+ #[test]
+ fn test_variant_object_65537_elements() {
+ test_variant_object_with_count(2_i32.pow(16) + 1, 2); // 2^16 + 1,
expected 3-byte field IDs
+ }
+
+ #[test]
+ #[ignore]
Review Comment:
why is this one ignored?
##########
parquet-variant/src/builder.rs:
##########
@@ -2171,4 +2171,116 @@ mod tests {
let variant = Variant::try_new_with_metadata(metadata,
&value).unwrap();
assert_eq!(variant, Variant::Int8(2));
}
+
+ fn test_variant_object_with_count(count: i32,
expected_field_id_size_minus_one: u8) {
+ let mut builder = VariantBuilder::new();
+ let mut obj = builder.new_object();
+
+ for val in 0..count {
+ let key = format!("id_{}", val);
+ obj.insert(&key, val);
+ }
+
+ obj.finish().unwrap();
+ let (metadata, value) = builder.finish();
+ let variant = Variant::try_new(&metadata, &value).unwrap();
+
+ if let Variant::Object(obj) = variant {
+ assert_eq!(obj.len(), count as usize);
+ assert_eq!(obj.get(&format!("id_{}", 0)).unwrap(),
Variant::Int32(0));
+ assert_eq!(
+ obj.get(&format!("id_{}", count - 1)).unwrap(),
+ Variant::Int32(count - 1)
+ );
+
+ let header_byte = first_byte_from_slice(&value).unwrap();
Review Comment:
Instead of extracting the field value directly, you could also use the API
in VariantHeader, like @codephage2020 did in
https://github.com/apache/arrow-rs/pull/7876
Something like
```rust
let expected_offset_size = OffsetSizeBytes::Three;
assert_eq!(obj.header.field_offset_size, expected_offset_size)
```
You would also have to put the tests into `object.rs` so the header field
was accessable
--
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]