jecsand838 commented on code in PR #7852: URL: https://github.com/apache/arrow-rs/pull/7852#discussion_r2196398061
########## arrow-avro/src/reader/mod.rs: ########## @@ -308,4 +315,129 @@ mod test { assert_eq!(read_file(&file, 3), expected); } } + + #[test] + fn test_decimal() { + let files = [ + ("avro/fixed_length_decimal.avro", 25, 2), + ("avro/fixed_length_decimal_legacy.avro", 13, 2), + ("avro/int32_decimal.avro", 4, 2), + ("avro/int64_decimal.avro", 10, 2), + ]; + let decimal_values: Vec<i128> = (1..=24).map(|n| n as i128 * 100).collect(); + for (file, precision, scale) in files { + let file_path = arrow_test_data(file); + let actual_batch = read_file(&file_path, 8); + let expected_array = Decimal128Array::from_iter_values(decimal_values.clone()) + .with_precision_and_scale(precision, scale) + .unwrap(); + let mut meta = HashMap::new(); + meta.insert("precision".to_string(), precision.to_string()); + meta.insert("scale".to_string(), scale.to_string()); + let field_with_meta = Field::new("value", DataType::Decimal128(precision, scale), true) + .with_metadata(meta); + let expected_schema = Arc::new(Schema::new(vec![field_with_meta])); + let expected_batch = + RecordBatch::try_new(expected_schema.clone(), vec![Arc::new(expected_array)]) + .expect("Failed to build expected RecordBatch"); + assert_eq!( + actual_batch, expected_batch, + "Decoded RecordBatch does not match the expected Decimal128 data for file {file}" + ); + let actual_batch_small = read_file(&file_path, 3); + assert_eq!( + actual_batch_small, + expected_batch, + "Decoded RecordBatch does not match the expected Decimal128 data for file {file} with batch size 3" + ); + } + } + + #[test] + fn test_simple() { + let tests = [ + ("avro/simple_enum.avro", 4, build_expected_enum(), 2), + ("avro/simple_fixed.avro", 2, build_expected_fixed(), 1), + ]; + + fn build_expected_enum() -> RecordBatch { Review Comment: That's a good call out. I was attempting to keep the dependencies (dev ones included) as minimal as possible, however the tests would absolutely be cleaner. I'd also have some hesitation about keeping test files in two repos in sync as well. I'll think on it some more, the upcoming PR to add the remaining integration tests maybe a good time to make this change as well. -- 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: github-unsubscr...@arrow.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org